NON-CONFIDENTIAL FINDINGS OF FACT AND NON-CONFIDENTIAL CONCLUSIONS OF LAW
This matter came before the Court for trial starting on February 10, 2010, and ending on February 24, 2010, to determine whether use of the name CRISTALINO on cava produced by J. Garcia Carrión, S.A., (Carrión) and imported by CIV USA, infringes or dilutes the CRISTAL trademarks of Champagne Louis Roederer (Roederer).1 Based on the evidence received at trial, the Court makes the following Non-Confidential Findings of Fact and Non-Confidential Conclusions of Law.2
NON-CONFIDENTIAL FINDINGS OF FACT
I.Sparkling wine
1. Still wine, or wine without bubbles, is made by a fermentation process that converts grape sugar into alcohol and carbon dioxide. The carbon dioxide in still wine is released by the wine producer after fermentation.
2. Sparkling wine, or wine with bubbles, is made using a second fermentation process that converts still wine into sparkling wine. According to the traditional method, also known as the “Method Champenoise” or “champagne method,” the second fermentation occurs in the bottle in which the sparkling wine is sold. The carbon dioxide is trapped in the bottle, dissolves into the wine, and escapes as bubbles when the sparkling wine is released from the bottle. According to the alternative “bulk” or “Charmat” method, the second fermentation takes place in a large tank or vat. After the second fermentation is complete, the contents of the tank are pumped into bottles for sale. The traditional method of producing sparkling wine results in higher-quality wine than the Charmat method.
3. Champagne and cava are both sparkling wines. When used properly, the term “champagne” denotes a sparkling wine made according to the traditional method from grapes grown in the Champagne region of France. A vintage champagne is made from grapes of the same year’s harvest. A non-vintage champagne is made from grapes of different years’ harvests. The term “cava” denotes a sparkling wine made in the Catalonia region of Spain.
II. The U.S. sparkling wine market
4. The United States has a three-tier distribution system for alcohol, including wine. According to this system, a foreign wine producer sells wine to an importer. The importer then sells the wine to retailers, which are typically hotels, restaurants, and liquor stores. The retailers sell the wine to the consumer. Each tier sets its own price for the wine. Wine producers are prohibited by law from communicating directly with or selling directly to consumers, but an importer may communicate with its distribution network, including retailers.
*8425. The size of the U.S. sparMing wine market was relatively constant between 1995 and 2006. With the exception of 1999, sparkling wine consumption between 1995 and 2006 varied between 11.8 and 13.3 million cases. In 1999, sparkling wine consumption increased to 15.7 million cases due to the millennium celebration. The sparMing wine market peaks in May, June, November, and December because large numbers of celebratory events, including graduations, weddings, and holidays, occur during those months.
6. Most consumers who purchase wine, including sparkling wine, in liquor stores have only a cursory knowledge of wine. Despite the efforts of champagne producers to educate consumers about the differences between champagne and other sparkling wines, consumers continue to receive mixed messages about the meaning of “champagne” because some sparkling wines made outside the Champagne region of France identify themselves as “champagne.” Consequently, many consumers do not distinguish between champagne and other sparkling wines.
7. Consumers generally enter liquor stores with a price point and type of wine in mind. It is unusual for a consumer to purchase a wine at a price significantly different from the original price point. When inexperienced consumers purchase a more-expensive product, they educate themselves about the product by asMng the retailers questions. Consumers who purchase less-expensive bottles of wine ask fewer questions, although some will ask for guidance or discuss the wine regardless of price.
8. Many consumers purchase sparMing wine at different price points depending on the reason for the purchase. Consumers commonly purchase a less-expensive spar-Ming wine for large gatherings and a more-expensive sparMing wine for special occasions such as anniversaries, weddings, birthdays, and New Year’s Eve. In addition, consumers purchasing sparkling wine for a wedding reception may economize by purchasing a more-expensive sparkling wine for the wedding party and a less-expensive sparMing wine for the guests.
9. When purchasing wine in liquor stores, consumers are often hurried and make impulse purchases. They look for a wine they are familiar with based on prior consumption, advertising, newspaper or magazine articles, or word-of-mouth. Consumers rely on several cues when selecting a wine, including the name of the wine, the color of and fonts used on the labels, and the shape of the bottles.
10. Because consumers rely on labels, it is a common practice for a prestigious French winery to inform consumers of an affiliation or association between a less-expensive product and the winery by including the winery’s name or a portion of it on the front label of the less-expensive product. Moet & Chandon, which produces DOM PERIGNON champagne, labels its California sparkling wines with the name “CHANDON.” G.H. Mumm, which also produces a high-quality champagne, labels its California sparMing wines with “MUMM NAPA.” Chateau Mouton-Rothschild, which produces a Bordeaux wine that sells for approximately $200 a bottle, labels its less-expensive wines MOUTON-CADET. The similarity in the names informs consumers of the connection between the less-expensive product and the French winery.
11. The vast majority of sparkling wine is sold in green bottles that have concave bases. The front labels on most bottles of sparkling wine do not have gold as the background color.3 Rather, the back*843ground color can be one of any number of colors, including white, cream, black, yellow, red, brown, and blue. Many front labels include a distinguishing feature, such as a ribbon, stripe, or picture, in a contrasting color. Most front labels prominently display the winery’s name. It is common for rosé sparkling wines to have a pink front label, and a few rosé sparkling wines have a pink-hued copper front label. Gold neck labels with a medallion on the front of the label are extremely common in the sparkling wine market, and many of the neck labels extend downwardly to accommodate the medallion. The medallions vary from an elaborate shield or crest to a circle identifying the winery or the sweetness of the sparkling wine. It is common for the neck labels to have a stripe or ribbon in a different color at the base. “Shoulder labels,” or labels located between the front and neck labels, are rare.
III. The parties and their products
A. Champagne Louis Roederer
12.Roederer is a French company headquartered in Reims, which is in the Champagne region of France. Roederer is a family-owned business that was founded in 1776 by Nicolas Henri Schreider, the uncle of Louis Roederer. Louis Roederer took ownership of the company in 1826, at which time the company’s name changed to Champagne Louis Roederer. Jean-Claude Rouzaud, a member of the Roederer family, is the president of Roederer’s board of directors and its former chief executive officer. Members of the Roederer family make up the remainder of the board. In January 2006, Frederic Rouzaud, Jean-Claude Rouzaud’s son, became the chief executive officer and general manager of Roederer. Roederer employs between 140 and 150 people on a year-round basis and an additional 600 people during the two-week harvest season.
13. Roederer owns a winery located in Reims where it produces vintage champagnes under the names CRISTAL and CRISTAL ROSÉ. CRISTAL champagne was first produced in 1876 at the request of Tsar Alexander II of Russia. According to Roederer tradition, the champagne was named “CRISTAL” because it was sent to Tsar Alexander II in crystal bottles.
14. The size of Roederer’s vineyards and the quality of each year’s grape harvest limits Roederer’s production of CRISTAL champagne. Roederer only uses grapes from its own vineyards, and if the grapes in a year’s harvest are below the standard required for CRISTAL champagne, Roederer does not produce CRISTAL champagne that year. The quality of CRISTAL champagne is enhanced by various wine-making practices, including harvesting the grapes by hand and aging the wine for six years in oak casks.
15. * * *. In 1995, Roederer sold * * * bottles of CRISTAL champagne in the United States. Roederer sold * * * bottles and * * * bottles in the United States in 2000 and 2004, respectively. With the exception of 2009, the demand for CRISTAL champagne has always exceeded its supply. An economic downturn caused the decreased demand for CRISTAL champagne in 2009.
*84416. Liquor stores typically sell CRIS-TAL champagne for between $200 and $280 per bottle. Restaurants, hotels, and nightclubs generally sell CRISTAL champagne for between $300 and $700 per bottle.
17. Roederer produces other champagnes, including BRUT PREMIER, BRUT VINTAGE, ROSÉ VINTAGE, and BLANC DE BLANCS champagne. CRISTAL and BRUT PREMIER champagne make up about 20% and 75% of Roederer’s champagne production, respectively. Roederer’s other champagne products account for the remainder. BRUT PREMIER champagne sells for between $40 and $50 per bottle in liquor stores and BRUT VINTAGE champagne sells for about $75 per bottle in liquor stores.
18. Roederer also owns the Roederer Estate winery in California’s Anderson Valley, which it purchased in 1981 because there was no room to expand in the Champagne region of France. Roederer selected the Roederer Estate vineyards in the Anderson Valley because the “terroir,” or microclimate, resembled that of the Champagne region. It presents Roederer Estate as “the California out-post of Champagne Louis Roederer” and chose the name “Roederer Estate” to increase the success of its California sparkling wines by linking them to the Roederer name. Roederer considered using the name “CRISTALINO” for its Roederer Estate products, but decided against it because Jean-Claude Rouzaud believed it would dilute the image of CRISTAL champagne.
19. Roederer Estate produced its first wines in 1988. The Roederer Estate winery produces ROEDERER ESTATE BRUT and ROEDERER ESTATE BRUT ROSÉ sparkling wine, along with ROEDERER ESTATE L’ERMITAGE and ROEDERER ESTATE L’ERMITAGE ROSÉ “prestige” sparkling wine. Roederer does not identify its Roederer Estate sparkling wines as “champagne” because they are made from grapes grown in California, not the Champagne region of France. A bottle of ROEDERER ESTATE sparkling wine sells for about $20 in liquor stores. Roederer sold * * * cases of ROEDERER ESTATE sparkling wine in 2007 and expects sales to increase to * * * cases by 2012.
20. About six years ago, Roederer acquired the Seharffenberger Cellars winery, which is also located in the Anderson Valley. Seharffenberger Cellars sparkling wines are made using a different method and have a different “style” than Roederer and Roederer Estate wines. Roederer does not promote any connection between Seharffenberger Cellars wines and Roederer.
21. Maisons, Marques & Domaines (MMD) is the importer and distributor for Roederer in the United States. Roederer owns approximately 95% of MMD and MMD management owns the remainder. MMD represents twenty-six wine producers, half of which Roederer owns.
B. Defendants
22. Carrión is a Spanish corporation that produces and markets wines and fruit juices. In 2009, Carrión made 800 million liters of juice and wine, about 60% of which were juice and 40% were wine. Priesca, S.A., owns Carrion’s stock. The Garcia-Carrión family owns 75% of Pries-ca’s stock and local banks own the remaining 25%.
23. In April 1997, Priesca purchased the stock of Jaume Serra, S.A., a Spanish winery founded in 1943. In August 1998, Jaume Serra merged with Carrión. The merged entity operates under the name “J. Garcia Carrión.”
24. The Jaume Serra winery, which is located in Vilanova y la Geltru, Spain, pro*845duces still wines and cava. Jaume Serra produced a still wine under the name CRISTALINO in the 1950’s and another still wine under the name CRISTALINO in 1979. Jaume Serra began selling cava under the name CRISTALINO some time before 1987. Carrion currently sells cava under the name CRISTALINO and the name JAUME SERRA. CRISTALINO cava is made according to the traditional method and is aged at least eighteen months before sale. Bottles of Carrion’s cava are not labeled until shortly before they leave the winery.
25. In connection with the 1997 merger, Carrion upgraded the Jaume Serra winery to expand its capacity and decrease production costs. The upgrade included installing a new wine cellar, new lines, and new production robots. This upgrade did not change the manufacturing process for CRISTALINO cava.
26. Jaume Serra began importing CRISTALINO cava into the United States in 1989. In 1992, Jaume Serra sold just over 10,000 bottles of CRISTALINO cava in the United States. In 1997, 384,864 bottles of CRISTALINO cava were sold in the United States, which increased to 689,-076 bottles in 2000. In 2004, Carrion sold just over 2 million bottles of CRISTALINO cava in the United States, and by 2009, Carrion’s U.S. sales had increased to slightly over 4 million bottles. Carrion’s sales of CRISTALINO cava in the United States make up 90% of its U.S. wine sales. Between 75% and 80% of CRISTALINO cava sold in the United States is CRIS-TALINO BRUT. In 2002, Carrion began selling significant amounts of CRISTALINO ROSÉ cava in the United States.4 CRISTALINO cava typically sells for between $8 and $10 per bottle in liquor stores and between $19 and $32 per bottle in restaurants. Defendants have maintained the price of CRISTALINO cava at no more than $10 per bottle in liquor stores for several years, despite inflation and an unfavorable exchange rate.
27. CIV USA, Inc., is a consortium of Spanish wineries that promotes and distributes Spanish wines in the United States. Jaume Serra joined CIV USA in 1991. Also in 1991, Friend Wine Marketing d/b/a CIV USA was incorporated to facilitate the process of obtaining an import license from the U.S. Bureau of Alcohol, Tobacco, and Firearms (BATF). Vince Friend is the president and sole shareholder of Friend Wine Marketing. Carrion took Jaume Serra’s place in CIV USA after the 1997 merger.
IV. Trademarks
A. Roederer’s marks
28. Roederer owns U.S. Trademark Registration No. 662,343 for CRISTAL CHAMPAGNE & Design for use in connection with champagne. The '343 mark was registered on the Principal Register on May 27, 1958. Roederer disclaimed the words “CRISTAL CHAMPAGNE” apart from the mark as shown in the registration.
29. Roederer also owns U.S. Trademark Registration No. 1,163,998 for the word mark CRISTAL CHAMPAGNE for use in connection with champagne. The '998 mark was registered on the Principal Register on August 4, 1981 under Section 2(f) of the Lanham Act, 15 U.S.C. § 1052 (2006). Roederer disclaimed the word “CHAMPAGNE” separate and apart from the mark as shown. Roederer filed an affidavit of incontestability for the '998 *846mark in 1987, which the U.S. Patent & Trademark Office accepted in 1989.
30. The '998 and '343 marks have dates of first use in commerce of March 25, 1937. In 1948, Roederer sold * * * bottles of CRISTAL champagne in the United States, and sales have occurred continuously since that date. Sales of CRISTAL champagne in the United States increased from * * * bottles in 1960 to * * * bottles in 1969. Roederer sold * * * bottles of CRISTAL champagne in the United States in 1979 and * * * bottles in the United States in 1989. In 2004, Roederer sold * * * bottles of CRISTAL champagne in the United States. Roederer has continuously used in commerce the CRISTAL mark in connection with champagne in the United States since at least 1948.
31. The appearance of bottles of CRISTAL champagne has not significantly changed since CRISTAL champagne’s creation in 1876. The following image depicts a present-day bottle of CRISTAL champagne.
32. The bottle is clear and, unlike most other sparkling wine bottles, has a flat base. The front label is rectangular, gold, and features a large white “LR” medallion on its center with a stylized design of white curved lines underneath the medallion. The vintage year is printed in maroon over the LR medallion. “CRISTAL®” appears on the upper left side of the LR medallion and “CHAMPAGNE” appears on the upper right side of the LR medallion, both printed in a maroon Roman serif font. “CRISTAL®” is printed in white beneath the LR medallion and “LOUIS ROEDERER” is printed in maroon above the LR medallion. “BRUT” is printed in maroon on the bottom left of the LR medallion, and the volume and percent alcohol by volume is printed in maroon beneath “BRUT.” “REIMS” and “PRODUCT OF FRANCE” are printed in maroon on the bottom right of the LR medallion.
33. A gold shoulder label depicts the Imperial coat of arms flanked on both sides by the initials “L.O.R.” in maroon. Gold engraving circumscribes the shield and the initials.
34. The neck label is gold. A smaller white LR medallion is located on the front and “CHAMPAGNE” is printed above the small LR medallion in white on a maroon ribbon. “LOUIS ROEDERER” is printed underneath the small LR medallion, also in white on a maroon ribbon. The smaller LR medallion and LOUIS ROEDERER are located in a downwardly-extending portion of the neck label. “CRISTAL” is horizontally embossed in maroon above the smaller LR medallion. An LR monogram is embossed in maroon above the word “CRISTAL.” The word “CRISTAL®” is also printed in maroon on both sides of the neck label approximately parallel to the *847smaller LR medallion and printed in white beneath the smaller LR medallion. An LR monogram is embossed in maroon on the back label. “MAISON FONDÉE EN 1776” and a very small LR monogram are printed in maroon on the back of the neck label directly opposite the smaller LR medallion. Gold engraving circumscribes the neck label.
35. The back label depicts the Imperial coat of arms flanked by the initials “L.O.R.” on either side in maroon. “CRISTAL®” appears on the left side of the back label, and the remainder of the label provides a government warning, lists the address of Roederer’s website, and states that the product was imported by MMD.
36. The use of the CRISTAL marks on bottles of CRISTAL ROSÉ champagne is very similar to their use on bottles of CRISTAL champagne. The CRISTAL ROSÉ champagne labels are a pink-hued copper color rather than the gold of CRISTAL champagne labels. The words “BRUT ROSÉ,” rather than “BRUT,” are printed in maroon on the bottom left of the LR medallion, and the words “MARQUE DÉPOSÉE” are printed in maroon at the top of the front label.
37. Bottles of CRISTAL and CRIS-TAL ROSÉ champagne are wrapped in yellow cellophane and packaged in gold boxes bearing a large LR medallion on their front. When displayed in liquor stores, the bottles are kept in the yellow cellophane because it prevents ultraviolet light from penetrating the clear bottle and degrading the quality of the champagne.
38. In 2008, Roederer redesigned the labels of its non-CRISTAL champagnes. The following image depicts the redesigned BRUT PREMIER bottle and label.
39. The BRUT PREMIER bottle is green with a concave bottom. The front label is light gold and bordered by maroon and gold lines. A gold LR monogram is centered on the front label, over which “LOUIS ROEDERER” is printed in large, all-capital maroon letters. “CHAMPAGNE” is centered beneath “LOUIS ROEDERER” in smaller all-capital maroon letters. As with the bottle of CRIS-TAL champagne, “BRUT” is printed in the lower left-hand corner and “REIMS” is printed in the lower right-hand corner of the front label. “BRUT PREMIER” is printed above the LR monogram on the front label, and “MAISON FONDÉE EN 1776” is printed in very small gold letters on the top of the maroon border.
40. The BRUT PREMIER bottle does not have a shoulder label. The neck label states “ROEDERER” in vertical maroon letters. The lower part of the neck label is virtually identical to the lower part of the CRISTAL neck label, including the white medallion with the gold LR monogram and trim. The neck label states “LOUIS *848ROEDERER,” rather than “CRISTAL,” on either side of the white medallion.
41. The labels of Roederer’s other champagnes are similar to the BRUT PREMIER label. They include an LR monogram on the front label and a white LR medallion on the front of the neck label. The use of the LR monogram on the front labels and the white LR medallion on the neck labels of Roederer’s nonCRISTAL champagnes links those champagnes with CRISTAL champagne.
42. Prior to 2008, the bottles of Roederer’s non-CRISTAL champagnes had a maroon, rather than white, LR medallion on the neck label. The medallion was larger and more ornate. The pre-2008 labels did include elements common to the CRISTAL champagne label, including the LR monogram on the front label and the prominent display of the LOUIS ROEDERER name on the front label.
43. As seen in the following image, ROEDERER ESTATE sparkling wine is sold in a green bottle. The bottle has a concave base.
44.The ROEDERER ESTATE label is brown and has a gold border. It is generally rectangular with an arch at the top. “ROEDERER ESTATE” is printed in all-capital yellow letters on the center of the front label. “ANDERSON VALLEY BRUT” and “ESTATE BOTTLED SPARKLING WINE” are printed in smaller gold letters beneath “ROEDERER ESTATE.”
45. The neck label extends downwardly at the front. An “RE” monogram is printed in maroon and gold on the downwardly-extending portion of the neck label in a location corresponding to the location of the white LR medallion on the neck labels of Roederer champagnes. The RE monogram is topped by an eagle and surrounded by grapevines, both printed in gold. A gold ribbon stating “ROEDERER ESTATE” is located beneath the RE monogram. “ROEDERER ESTATE” is vertically printed in gold on the front of the neck label and horizontally printed in maroon on either side of the RE monogram.
46. Another RE monogram is printed on the back of the neck label, encircled by the words “ROEDERER ESTATE ANDERSON VALLEY.” “ROEDERER ESTATE” is vertically printed in gold letters on the back of the neck label.
47. The bottles and labeling of ROEDERER ESTATE ROSÉ and ROEDERER ESTATE L’ERMITAGE sparkling wine are similar to those of ROEDERER ESTATE sparkling wine. The ROEDERER ESTATE L’ERMITAGE label is dark gray with the word “L’ERMITAGE” centered above the name ROEDERER ESTATE. An RE monogram and shield are formed in the shoulder of the bottle. The labels on bottles of ROEDERER ESTATE ROSÉ and ROEDERER ESTATE L’ERMITAGE ROSÉ sparkling wine are a pink-hued dark copper.
B. Carrion’s marks
48. In 2000, Carrion sought registration of the CRISTALINO mark in connection with alcoholic beverages (excluding beer) in International Class 33.5 Carrion *849did not conduct any formal or informal searches or investigations before seeking registration of the CRISTALINO mark. In June 2002, after the mark was published for opposition, Roederer sent a letter to Carrion informing it of Roederer’s rights in the '343 and '998 marks and asking Carrion to withdraw its application for registration and agree not to use the CRISTALINO mark for wine, sparkling wine, and champagne. Carrion did not respond to the letter or inform Friend of its existence. Roederer subsequently opposed registration of the CRISTALINO mark.6
50.The bottle is green and has a concave bottom. The front label is a gold rectangle with a black bottom edge. “CRISTALINO” is printed in black Roman serif type on the center of the front label. A black line is located underneath “CRISTALINO” and a gold medallion with an image of a bunch of grapes is located immediately above “CRISTALINO.” “METODO TRADICIONAL” is printed in smaller burgundy letters beneath the word “CRISTALINO.” “BRUT” is printed in larger letters in black beneath the words “METODO TRADICIONAL.” “CAVA” is printed in burgundy on the bottom right-hand corner of the label. The bottom left-hand corner of the label states in burgundy “SPARKLING WINE FERMENTED IN THIS BOTTLE.” Two borders extend around the perimeter of the front label. The burgundy inner border is comprised of grape leaves and grape bunches. A second border of two gold lines encloses the border of grape leaves and grape bunches. The name “CRISTALINO” is the dominant element of the front label.
51. On the black bottom edge, which is outside the borders, the label states in very small letters “Produced by: Jaume Serra / PRODUCE OF SPAIN,” provides the address of the Jaume Serra winery, and states “IMPORTED BY: CIV (USA), SACRAMENTO, CA.” The alcohol by volume is printed on the left side of the black edge and the volume of the bottle is printed on the right side of the black edge.
52. A circular sticker having a black center and gold edging is located on the shoulder of the bottle. “Wine & Spirits Magazine” is printed in gold on the sticker’s center and “Value Brand of the Year 3 Consecutive Years” is printed in black on its edging.
53. The neck label is gold and extends downwardly in the front. A bunch-of-grapes medallion is located in the downwardly-extending region of the neck label. “CRISTALINO” is vertically printed in *850black on the front of the neck label. The bottom of the neck label is edged with highly-reflective gold. The back of the neck label is blank.
54. Another label is located on the back of the CRISTALINO bottle. Within a two-line gold border, the name “CRISTALINO” is printed in all-capital black letters. The word “BRUT” appears beneath “CRISTALINO,” separated by a black line. Underneath the word “BRUT,” the label states that the second fermentation of the cava took place in the bottle, advertises the eava’s “soft scent of toast and dry, fingering citrus qualities on the palate,” and describes CRISTALINO cava as “a clean bubbly that is sophisticated enough for just about any meal — as well as your next celebration.” The surgeon general’s warning and a bar code are located on the bottom of the back label, outside of the gold border.
55. The CRISTALINO ROSÉ label is virtually identical to the CRISTALINO label, except that it is.,a pink-hued copper color. The CRISTALINO ROSÉ bottle is clear and has a concave bottom. “ROSÉ BRUT” appears in black beneath the “CRISTALINO” name on the front label, and the center of the Wine & Spirits sticker is pink instead of black.
56.Although the bottles used for CRISTALINO cava have always been green, the CRISTALINO label has changed significantly since its introduction in 1989. The 1989 version of the CRIS-TALINO label is shown in the following black-and-white copy of the 1989 BATF application.
57.In 1989, the front label of CRIS-TALINO BRUT was gold. “CRISTALINO” was centered on the front label in large all-capital letters. “JAUME SERRA” was printed in all-capital letters immediately below “CRISTALINO.” The JAUME SERRA was smaller than the *851CRISTALINO. A design of curly lines surrounded the words “CRISTALINO JAUME SERRA,” and the words “BRUT” and “SPARKLING WINE PRODUCT OF SPAIN” were vertically stacked above the word “CRISTALINO.” “METHODE CHAMPENOISE” and “METHODE TRADITIONNELLE” were printed beneath JAUME SERRA, and CAVA was printed in large letters beneath those phrases.
58. “CRISTALINO JAUME SERRA” was centered at the top of the back label, which included a description of the in-the-bottle second fermentation and grapes used to produce the cava. “CRISTALINO” was printed on the center of the neck label, with “JAUME SERRA” on the right and “CRISTALINO” on the left. “BRUT” was printed on the neck label above “CRISTALINO.” The back of the neck label included a crest surrounded by the words “CRISTALINO JAUME SERRA.” The 1989 CRISTALINO bottle prominently featured the JAUME SERRA name on the front, back, and neck labels.
59. Jaume Serra updated the CRIS-TALINO labels in 1991. A copy of the BATF application for the 1991 label is shown below.
60. The name “CRISTALINO” was centered on the front label and the name “JAUME SERRA” printed beneath it. The phrases “METHODE CHAMPENOISE” and “METHODE TRADITIONNELLE” were located in smaller font between “CRISTALINO” and “JAUME SERRA.” The bunch-of-grapes medallion found on the label today was added to the label above the name “CRISTALINO.” The back label continued to prominently feature the name “CRISTALINO JAUME SERRA.”
61. The bunch-of-grapes medallion was centered on the neck label with a ribbon identifying the sweetness of the cava beneath it and “CRISTALINO” above it. The right and left sides of the neck label both read “JAUME SERRA.” A medallion including the name “JAUME SERRA” was located on the back of the neck label. The CRISTALINO BRUT labels were black. The CRISTALINO DRY and SEMI-DRY labels were gold. Again, the 1991 labels emphasized the JAUME SERRA name.
*85262.In 1993, Carrion again changed the CRISTALINO cava labels, as shown in the following reproduction of the BATF label application.
63. The 1993 CRISTALINO BRUT labels were very similar to the current labels, except that 1993 labels were black instead of gold. The front label included the CRISTALINO name, bunch-of-grapes medallion, border, and words “BRUT” and “CAVA” found on the front label today. “METHODE TRADITIONNELLE” and “METHOD CHAMPENOISE” were printed underneath “CRISTALINO” in the location where “METODO TRADICIONAL” is printed today. “SPARKLING WINE” was printed above the medallion. The winery and import company information was located at the bottom edge of the front label and outside the borders, just as it is today.
64. With the exception of the language describing the cava, the back label is virtually identical to the current back label. The neck label included the bunch-of-grapes medallion encircled by “CRISTALINO.” The word “BRUT” appeared on a ribbon beneath the bunch-of-grapes medallion, and the Jaume Serra crest appeared, including the name “JAUME SERRA,” on the back of the neck label.
65. The 1993 labels de-emphasized the “JAUME SERRA” name by removing it from the main portion of the front label. Friend credibly testified that “JAUME SERRA” was removed because the pre1993 labels were “busy” and “confusing” and the name “Jaume Serra” was difficult for U.S. customers to pronounce.
66. Friend further testified that a buyer at the Wine Max store in San Francisco indicated in 1994 that he would purchase five cases of CRISTALINO BRUT cava if the label was changed from black to the gold label used by CRISTALINO EXTRA-DRY cava because there was no market for CRISTALINO EXTRA-DRY cava but he could sell CRISTALINO BRUT cava if the label was gold. Friend believed that other retailers would agree with the Wine Max buyer’s opinion. He thought the change was worthwhile because a purchase of five cases represented a commitment to the product on the part of the retailer and he could use the Wine Max buyer’s purchase of five cases to persuade other retailers to stock CRISTALINO cava. As a result, the CRISTALINO BRUT label was changed from black to gold.
67. In the mid-1990’s, Jaume Serra changed the phrases “METHODE TRA-DITIONNELLE” and “METHODE CHAMPENOISE” to “METODO TRADI-*853CIONAL” to conform to European Community regulations.
C. The meaning of the marks
68. “CRISTAL” is not a common word in English. The English translation of “CRISTAL” is “crystal,” which means “quartz that is transparent or nearly so and that is either colorless or only slightly tinged.” Webster’s Third New International Dictionary 548 (2002). When used in the United States, the word “CRIS-TAL” refers to a proper name, a geographic location, or Roederer’s CRISTAL champagne. U.S. consumers pronounce “CRISTAL” with a long “e” in the first syllable and the accent on the second syllable.
69. “CRISTALINO” is not a common word in English. The English translation of “CRISTALINO” is “crystalline,” which means “made of crystal” or “resembling crystal as transparent, pure, pellucid.” Id. at 549. When spoken by U.S. consumers, the first syllable is pronounced with a short “i” and the accent is on the third syllable.
70. “CRISTAL” is a generally meaningful word, or morpheme. English speakers recognize the suffix “INO” as conveying a diminutive, i.e., a smaller or lesser, meaning or a “related to” meaning.7 The addition of the suffix “INO” to the root word “CRISTAL” to form “CRIS-TALINO” suggests to English speakers that CRISTALINO is a diminutive of or somehow associated with the root word CRISTAL.
71. In recent years, the media has recognized the verbal association between “CRISTALINO” and “CRISTAL” by suggesting that those who cannot afford CRISTAL champagne should instead purchase CRISTALINO cava and describing CRISTALINO cava as “the other CRIS-TAL.” In addition, a number of Internet postings indicate that consumers purchase CRISTALINO cava because the name sounds like “CRISTAL.” Some of those posts describe CRISTALINO cava as CRISTAL’s “younger brother.”
D. Similar marks
72. Marks including the terms “cristal,” “crystal,” “krystal,” or some variant (CRISTAL or similar terms) have been registered for use in connection with a number of products. A 1988 search of U.S. trademark registrations indicates that there were about sixty registrations or applications for registrations for marks using CRISTAL or similar terms in connection with alcoholic and non-alcoholic beverages. Some of the applications for registration were abandoned. Also in 1988, trade names including CRISTAL or similar terms were listed in trade directories in connection with alcoholic and nonalcoholic beverages. Defendants introduced no evidence of the geographical extent or volume of sales of the products sold under those marks or trade names.
73. In 1995, CRYSTAL LAKE wine (including “California champagne”), CRYSTAL GEYSER mineral water, AUGUARDIENTE CRISTAL liquor, STOLICHNAYA CRISTALL vodka, *854CRYSTAL LIGHT soft drinks, and CRISTALINO sparkling wine were available for purchase in California.
74. A 2009 search of U.S. trademark registrations identified registrations for marks using CRISTAL or similar terms for a number of products, about forty of which were alcoholic beverages. Many of those registrations were related. For example, there were registrations for over fifteen variants of the “CRISTALL” mark for use in connection with vodka. In addition, an Internet search located approximately forty products using CRISTAL or similar terms, including rum, vodka, salt, and bottled water. Defendants introduced no evidence of the geographical extent or volume of sales of the products sold under the registered marks or names. Several of the products located on the Internet were not available for sale from the website on which they were found. Ten of the products found on the Internet were wines, but none of them had labels that were gold or otherwise resembled the CRISTAL labels. Based on the evidence received at trial, some of the wines did not display the CRISTAL or similar terms on their labels or bottles.
75. Roederer has enforced its trademark rights against those using CRISTAL or similar terms in connection with champagne and sparkling wine, but does not enforce its trademark rights against every product having CRISTAL or similar terms in its name. In 1988, Roederer sent a cease-and-desist letter to an entity marketing an American sparkling wine under the name “California Crystal.” In 1998, Roederer unsuccessfully opposed the registration of “CRYSTAL CREEK” for use in connection with wine. The United States Court of Appeals for the Federal Circuit affirmed the Trademark Trial and Appeal Board’s finding of no likelihood of confusion based on “the dissimilarity of the marks with respect to appearance, sound, significance, and commercial impression.” Champagne Louis Roederer, S.A. v. Delicato Vineyards, 148 F.3d 1373, 1374-75 (Fed.Cir.1998). In 2004, Roederer halted the sale in the United States of cava marketed under the name “CRISTAL CAVA CASTELLBLANCH” by Castellblanch S.A. and Victore Imports Company.
V. Marketing and trade channels
A. Recognition, advertising, and promotion of CRISTAL champagne
76. Roederer is considered one of the finest champagne houses in the world. CRISTAL champagne was known in the United States as one of the best champagnes by wine connoisseurs and those knowledgeable about wine at least as early as the 1970’s. The quality of CRISTAL champagne continues to be recognized today. For example, in December 2009, CRISTAL champagne received a 100-point rating from Wine & Spirits magazine. The recognition of CRISTAL champagne as one of the best champagnes by those knowledgeable about wine is unquestionable.
77. CRISTAL champagne is made with a sophisticated consumer or wine connoisseur in mind. However, not all consumers of CRISTAL champagne are sophisticated or wine connoisseurs. Some consumers purchase CRISTAL champagne because they appreciate its quality. Others, including celebrities, purchase CRISTAL champagne for the social recognition it brings.
78. CRISTAL is Roederer’s “flagship” brand. The reputation of CRISTAL champagne helps persuade distributors and retailers to carry other Roederer and Roederer Estate products and permits Roederer to sell its other products at a premium price. Displaying CRISTAL champagne in a liquor store adds to the prestige of the store, and retailers actively *855market CRISTAL champagne for this reason. However, only a small percent of consumers know that Roederer produces CRISTAL champagne. Consequently, while CRISTAL’s reputation provides certain benefits within the industry to Roederer and its other products, its influence on consumer perception of the Roederer brand is attenuated.
79. “Aspirational” brands convey a certain image that people aspire to share. By their very nature, aspirational brands are known by those who do not actually purchase or consume the branded product. For example, the ROLEX brand projects an image of success that consumers seek to display by wearing a ROLEX watch. Some who cannot afford a ROLEX watch aspire to do so in the future in the hopes of projecting a successful image. Consumers purchase CRISTAL champagne because it projects a sense of quality, elegance, prestige, glamour, and status. To maintain the image of the CRISTAL brand, Roederer prices CRISTAL champagne higher than DOM PERIGNON champagne and portrays CRISTAL champagne as a work of art. The CRISTAL brand is an aspirational brand and its audience is larger than those who consume CRISTAL champagne.
80. In 1998, Roederer adopted the slogan “Without Compromise” for its products, which it utilized in two to three campaigns over a period of several years. The campaign featured BRUT PREMIER champagne and was intended to advertise the Roederer brand by focusing on the LR monogram. The Without Compromise advertisements were published in Roederer’s L’Officiel newsletter in 2001, 2002, and 2003. Roederer publishes about 50,000 copies of L’Officiel annually in several different languages. Copies of the newsletter are sent to Roederer’s distributors, including Roederer’s U.S. distributors, who then provide them to customers, journalists, and liquor stores. In 2006, Roederer placed brightly-colored Without Compromise advertisements inspired by Andy Warhol and an advertisement highlighting the name Roederer and its founding year (1776) in the New York Times, Wine News, Quarterly Review of Wine, Wine & Spirits, Wine Enthusiast, and Wine Spectator.
81. In 2009, Roederer adopted an “Artistry of Champagne” slogan for its advertising. The five Artistry of Champagne advertisements feature an abstract depiction of a champagne bottle, the front label of the champagne, and an LR monogram on the neck in the same location as that of the smaller white LR medallion on the neck of an actual bottle. The labels shown include those of CRISTAL, BRUT VINTAGE, BRUT PREMIER, and BLANC DE BLANCS champagne. The 2009 Artistry of Champagne advertisements showing CRISTAL champagne appeared in the New Yorker, Pebble Beach publications, a catalog for a classic motorcar show, a catalog for a horse show, and the December 2009 issue of Wine & Spirits.
82. Prior to 2009, Roederer’s policy was not to promote CRISTAL champagne. Roederer had two reasons for this policy. First, advertising was unnecessary because Roederer sold every bottle of CRISTAL champagne it produced. Second, Roederer believed advertising CRISTAL champagne would make it seem like an ordinary product. On occasion, however, distributors promote the CRISTAL brand with Roederer’s permission, and retailers prominently feature CRISTAL champagne in their advertisements. Roederer donates CRISTAL champagne to about twenty charitable events and festivals in the United States on an annual basis. Roederer also sends samples of CRISTAL champagne to journalists at Cooking Light, Quarterly Review of Wine, Wine & *856Spirits, Gourmet, Bon Appétit, and similar publications, as well as to newspaper journalists in major metropolitan areas.
83. CRISTAL champagne has enjoyed unsolicited publicity for several decades. Between 1950 and 1989, CRISTAL champagne was mentioned in 140 newspaper articles, and between 1990 and 1993, it was mentioned in an additional 67 newspaper articles. Newspapers that published articles mentioning CRISTAL champagne during that time include the New York Times, San Francisco Globe, San Francisco Chronicle, Los Angeles Times, and Miami Herald. The vast majority of the newspaper articles mentioned CRISTAL champagne either in connection with a special event at which CRISTAL champagne was served, as one of several champagnes in an article about sparkling wine, or in a description of the expensive or extravagant lifestyle of a particular celebrity.
84. For example, one article published in several newspapers in December 1989 recommended serving champagne when entertaining over the holidays. That article characterized “Louis Roederer” as one of “America’s favorite French champagnes” and then briefly described Roederer’s CRISTAL and BRUT PREMIER champagne. The remainder of the article was an appetizer recipe. On March 5, 1989, the New York Times mentioned that CRISTAL ROSÉ champagne was served as the dessert wine to President Ronald Reagan at Le Cirque, and on January 20, 1983, the Palm Beach News published an article describing a lunch with President Richard Nixon and mentioning that CRISTAL champagne was served
85. Only a few of the articles were directed to Roederer or CRISTAL champagne. One such article was published in the San Francisco Globe in 1954. This article highlighted the visit of the chief executive of Roederer to introduce the 1949 vintage of CRISTAL champagne.
86. In 1979, the Los Angeles Times published an article on champagne describing Roederer’s products, including CRIS-TAL champagne, and an article in the New York Times Magazine in 1979 described CRISTAL champagne as the most expensive champagne available. The Los Angeles Times published the results of the Wine & Cheese Festival rankings in 1970, in which CRISTAL champagne was rated the Grand Champion. In 1979, the Fresno Bee described a champagne dinner at which several Roederer champagnes were served, including CRISTAL champagne. Edward McCarthy, the author of Champagne for Dummies and a wine expert, testified that CRISTAL champagne’s reputation in the field of wine and champagne was “singular” and had been since at least the early 1970’s.
87. Between 1950 and 1989, CRISTAL champagne was mentioned in seventeen magazines, and between 1990 and 1993, in another six magazines. Many of those magazines were special-interest magazines such as Food & Wine and Wine Spectator. When CRISTAL champagne was mentioned or featured in a general-interest magazine such as Vogue, W, or Bazaar, it was generally mentioned in connection with entertaining or travel and as one of several products. In 1985, CRISTAL champagne was displayed on the cover of Bride’s in a wine cooler next to two champagne flutes. Only the neck label was visible. Elle featured a bottle of CRISTAL champagne on its cover along with a number of other luxury goods in May 1989. Only the neck and shoulder labels were visible. In 1973, Time published an article describing President Nixon’s dinner with Leonid Brezhnev, the General Secretary of the Communist Party of the Soviet Union. The first line of the article read: “The fear has been buried in champagne toasts (Roederer Cristal) and broad presidential smiles and the haunting strains of Tchai*857kovsky by the Marine Band.” The other nine paragraphs described the dinner and its political significance.
88. CRISTAL champagne was featured as a product placement in the movies Heartburn, Star 80, and In the Money. All of those movies were released before 1990.
89. In 1987, Andrew Jones, the “Flying Wine Man,” hosted local radio shows on 135 stations in 20 states explaining champagne and featuring ten brands. During his show, Jones described Roederer’s history and the history of CRISTAL champagne. The predicted listening audience for the tour was 40 million.
90. In the 1980’s and earlier decades, celebrities such as Sophia Loren, Julio Iglesias and Ivana Trump were unofficial “ambassadors” of the CRISTAL brand. Pictures showing the “ambassadors” and a bottle of CRISTAL champagne were distributed in various publications, but no evidence indicates the extent of their distribution. In April 1992, Frank Sinatra described CRISTAL champagne as his “favorite” in front of a 1000-person audience. President Reagan was photographed with CRISTAL champagne, although the publicity this photograph received is unknown.
91. Between 1997 and 2009, CRISTAL champagne was mentioned in about 35 magazines and almost 900 newspaper articles. Those articles were published in the New York Times, Miami Herald, Washington Post, Wall Street Journal, USA Today, InStyle, Entertainment Weekly, Rolling Stone, and Forbes, as well as on cnn.com and msnbc.com.
92. Between 1997 and 2009, CRISTAL champagne was mentioned in just under 300 books. A significant number of the books were written by bestselling authors such as Nora Roberts, Candace Bushnell, Richard North Patterson, John Grisham, and Danielle Steele. Typically, CRISTAL champagne was mentioned in connection with the consumption of or a request for CRISTAL champagne by a character.
93. Today, CRISTAL champagne is frequently used as a product placement in movies and television shows. Roederer and MMD manage the use of the CRIS-TAL mark in movies and on television by reviewing the scripts to determine how CRISTAL champagne will be used. If Roederer considers the proposed use inappropriate, Roederer will reject the request to use CRISTAL champagne. In some cases, producers alter the use of CRIS-TAL champagne in response to a rejection by Roederer. As of 2004, MMD was receiving about forty requests a year to place CRISTAL champagne in movies, television shows, and music videos. Roederer and MMD do not pay for any of those product placements.
94. CRISTAL champagne has appeared in several movies since 1990, including The Hours, Austin Powers Gold-member, Lost in Translation, Something’s Gotta Give, The First Wives Club, Eyes Wide Shut, and The Terminal. The references to CRISTAL champagne are typically oral or a screen shot of a CRISTAL champagne bottle that is often in the background or partially obscured. CRISTAL champagne has also appeared in several television shows, including Entourage, ER, and 30 Rock. In many of those television shows, CRISTAL champagne was shown in the background or partially obscured. In other television shows, CRISTAL champagne was mentioned by name. For example, a character complained about the high price of CRISTAL champagne in one episode of The Sopranos.
95. In some of the television shows, CRISTAL champagne was highlighted as an important part of the show. In a show celebrating Oprah Winfrey’s 50th birthday, John Travolta proposed a toast to Winfrey, *858mentioned CRISTAL ROSÉ champagne by name twice, and described CRISTAL ROSÉ champagne as Winfrey’s “favorite.” A large bottle of CRISTAL ROSÉ champagne was then presented on-stage and the champagne consumed. On The Tonight Show with Jay Leño, a guest brandished a bottle of CRISTAL champagne during his entrance, mentioned the champagne by name, and presented it to Jay Leno with much fanfare over a two-minute period.
96. The number of viewers of television shows and movies and the number of readers of magazines and newspapers referring to CRISTAL champagne is significant. On average, OprahTs audience numbers 8 million, The Sopranos’ audience numbers 11 million, and Entourage’s audience numbers 2.5 million. The average circulation of the New York Times is about 1.5 million, the San Francisco Chronicle a little under 1 million, and InStyle about 1.7 million.
97. Beginning in the late 1990’s, mentions of CRISTAL champagne became prevalent in music lyrics and videos by artists such as Jay-Z, Lil’ Kim, Mariah Carey, Snoop Dog, and 50 Cent. Women’s Wear Daily reported that CRISTAL was the ninth-most mentioned brand in Billboard’s Top 20 singles in 2005, and the CRISTAL name was the eighth most-mentioned brand in rap music the same year. The Washington Post described CRISTAL as a “household name” as a result of this publicity. Media recognition of the CRIS-TAL brand includes statements such as “CRISTAL, everybody knows that kind of champagne.” The brand recognition resulting from the references to CRISTAL champagne in rap music was commented on in the San Francisco Chronicle in 2004 and the Wall Street Journal in 2006.
98. In addition, over fifty videos, blogs, and pictures are available on the Internet where the poster displays a picture of a CRISTALINO bottle with the “INO” obliterated or obscured and describes CRIS-TALINO cava as CRISTAL’s “younger brother” or otherwise jokes about the similarity between the words “CRISTAL” and “CRISTALINO.” In making the joke, the consumers demonstrate their awareness of CRISTAL champagne.
B. Recognition, advertising, and promotion of CRISTALINO cava
99. Since 1993, CRISTALINO cava has received favorable reviews or been recognized as a good value in a number of publications, including the San Francisco Independent, Atlanta Journal, and Philadelphia Enquirer. CRISTALINO cava received a silver medal in its category at the 1994 International Wine Competition in Atlanta. Wine Spectator recognized CRISTALINO cava as a “Best Buy” in 1996, and Wine & Spirits recognized CRISTALINO cava as the “Value Brand of the Year” in 2002, 2003, and 2004. The December 2003 issue of Bon Appétit described CRISTALINO cava as giving “excellent effervescent flavor at a moderate price.” Carrion promotes CRISTALINO cava as a “best buy” based on its combination of price and quality, and its marketing strategy is based in part on the assumption that consumers will purchase a less-expensive sparkling wine when large quantities are required.
100. Carrion maintains a “bank” for the promotion of CRISTALINO cava of $* * * for each case of CRISTALINO cava sold to a distributor. The bank funds pricing promotions, revisions to wine lists, and point-of-sale promotional devices. Typically, the amount expensed against the bank by CIV USA on an annual basis is between $* * * and $* * *. Between 1999 and 2000, CIV USA retained TKO Advertising, Inc., to assist with the promotion of CRISTALINO cava. During *859that time frame, CIV USA spent about $* * * on advertising CRISTALINO cava. Between 2000 and 2006, excluding the TKO Advertising expenditures, CIV USA spent approximately $* * * on media advertising and $* * * on promoting and marketing CRISTALINO cava.
C. Trade channels
101. There are around * * * U.S. distributors of CRISTAL champagne, which is sold in approximately 4500 liquor stores in the United States. About 70% of those liquor stores also carry BRUT PREMIER champagne. Sales at liquor stores account for half of the sales of CRISTAL champagne and sales at restaurants, hotels, and nightclubs account for the other half. At least twelve liquor distributors carry both CRISTAL champagne and CRISTALINO cava.
102. CRISTAL champagne is sold at a variety of liquor stores, including Surdyk’s Liquor in Minneapolis and Sherry Lehmann in New York. At times, CRISTAL champagne is sold at mass merchandisers such as Safeway and CostCo, although such sales are inconsistent with the desired image of CRISTAL champagne. Liquor stores display CRISTAL champagne on the top shelf, behind the check-out counter, or in a “lockbox.”
103. CRISTALINO cava is sold at mass merchandisers such as CostCo, Safeway, and Sam’s Club. CRISTALINO cava is also sold in a variety of liquor stores, including Surdyk’s in Minneapolis; Beltramo, K & L, and The Wine Exchange in California; and Sherry Lehmann in New York. When sold in liquor stores, CRIS-TALINO cava is typically displayed on lower shelves or in bins located in or beside the aisles.
104. Although CRISTAL champagne is not sold at every liquor store that sells CRISTALINO cava, the products are frequently sold at the same liquor stores. CRISTAL champagne and CRISTALINO cava are displayed in the same area, usually identified as the “champagne and sparkling wine” area, which may be subdivided by country or region of origin.
105. In flyers and brochures for sales at liquor stores, sparkling wines are generally listed together under a “champagne and sparkling wine” category, which is then divided into subcategories of “French Champagne,” “European Sparkling,” and “California Sparkling” or “American Sparkling.” CRISTAL champagne is listed under the French Champagne subcategory and CRISTALINO cava under the European Sparkling subcategory.
106. Although CRISTAL champagne and CRISTALINO cava do not frequently appear on the same wine list, such listings exist. When they are displayed on the same wine lists, CRISTAL champagne and CRISTALINO cava are typically listed together in close proximity in a “champagne and sparkling wine” section. Many wine lists identify the country of origin of CRISTAL champagne and CRISTALINO cava. Almost all wine lists identify CRIS-TAL champagne as a Roederer product; fewer identify CRISTALINO cava as a product of the Jaume Serra winery.
107. Neither CRISTAL champagne nor CRISTALINO cava is heavily advertised by its producer. Advertisements for and reviews of both products have appeared in Wine & Spirits, Wine Enthusiast, and Wine Spectator. The audiences of those publications have a median age in the late 40’s and a median household income of $125,000 or higher. Over 80% of the readers have attended college.
VI. Actual confusion
A. Testimony
108. Fabrice Rossett began working at Roederer in 1974 as the export director. *860He worked for Roederer in various capacities until 1996, at which time he was the president and chief executive officer of MMD. Rossett testified that no retailer or consumer ever indicated to him that there was confusion between CRISTALINO cava and CRISTAL champagne.
109. McCarthy, the wine expert, worked in wine shops between 1976 and 1998. McCarthy testified that during the last nine years he worked in wine shops, after the introduction of CRISTALINO cava in 1989, consumers asked him if there was a relationship between CRISTALINO cava and CRISTAL champagne. McCarthy did not testify as to the frequency of those questions.
110. Xavier Barlier, the vice-president of marketing and communications at MMD, regularly visits the champagne and sparkling wine section of liquor stores as part of his duties. While working the Roederer table and pouring CRISTAL champagne at a Toast of the Town event, Barlier was asked: “Do you know that there’s a Cristalino table?” He testified that, beginning in 2001 and continuing to the present, consumers asked him if there was a relationship between CRISTALINO cava and CRISTAL champagne at store tastings, particularly during the winter holidays.
111. In December 2004, a liquor store owner in Michigan purchased bottles of CRISTALINO cava believing they were bottles of CRISTAL champagne because stickers placed on the bottle concealed the “INO” portion of “CRISTALINO” and proclaimed “1995 Louis Roederer.” Once notified of the sale, the district manager for CRISTALINO cava’s distributor in the region informed the store owner that the bottles were not CRISTAL champagne. No other action was taken with respect to this incident.
B. Likelihood of confusion survey
112. Roederer retained Dr. Leon B. Kaplan, the president and chief executive officer of Princeton Research and Consulting Center, to design and implement a survey intended to measure the likelihood of consumer confusion between CRISTAL and CRISTALINO. Dr. Kaplan holds a Ph.D. in Industrial Consumer Psychology from Purdue University and an MBA from the Wharton School. He has worked in the field of market research for over thirty years. Dr. Kaplan has designed about twenty surveys for intellectual property litigation and has been involved with over 200 other surveys.
113. The Kaplan survey was conducted on a nationwide basis during the fall of 2007 by a professional interviewing service. Neither the survey interviewers nor the survey respondents knew the purpose of the survey. The interviewers surveyed 261 people in shopping malls in eight different regional U.S. markets. Two of the malls were located in the Northeast, two in the Midwest, two in the South, and two in the West. The survey participants were randomly divided into a control group and a test group.
114. Dr. Kaplan defined the survey universe as people who were 21 or older, had purchased in the past six months or were likely to purchase in the next six months imported sparkling wine under $35, and were aware of CRISTAL champagne before the interview. Dr. Kaplan characterized this universe as “purchasers and potential purchasers of CRISTALINO who are aware of CRISTAL.” He selected the $35 figure because he wanted to include those who had purchased CRIS-TALINO cava in a restaurant or were likely to do so as well as those who had purchased CRISTALINO cava in a liquor store or were likely to do so.
*861115. The interviewer first determined the potential respondents’ gender and age. Next, the interviewers determined whether the potential respondents had purchased in a store or restaurant in the past six months an imported sparkling wine for under $35 per bottle or were likely to do so in the next six months. After screening for characteristics that would render a potential respondent ineligible, such as working for a law firm, the interviewers brought potential respondents to an interview area. Once in that area, the potential respondents were shown a bottle of CRIS-TAL champagne still wrapped in yellow cellophane and asked: “To the best of your knowledge, have you ever heard of or seen this brand of wine before today, or don’t you have an opinion about that?” Only potential respondents who indicated that they had heard of or seen CRISTAL champagne proceeded to the rest of the survey. Dr. Kaplan estimated that the interviewers at each survey site would need to bring 90 people to the interview room to reach the desired number of 30 respondents per site. He had no knowledge of how many people were actually brought to the interview room.
116. Next, the interviewers asked the respondents a series of questions intended to serve as a “mental change of pace.” The respondents were then shown four bottles of imported sparkling wine displayed together in a line-up. The test group was shown bottles of VEUVE DU VERNAY sparkling wine, WILLM sparkling wine, PAUL CHENEAU sparkling wine, and CRISTALINO cava. The control group was shown VEUVE DU VER-NAY sparkling wine, WILLM sparkling wine, PAUL CHENEAU sparkling wine, and a bottle of “COURTALINO” cava created as a control for the survey. The COURTALINO bottle was virtually identical to a CRISTALINO bottle except that the front and neck labels read “COUR-TALINO” rather than “CRISTALINO.” Dr. Kaplan chose the name “COURTALINO” because he thought it was as similar to “CRISTALINO” as possible without infringing the CRISTAL marks.
117. The respondent was given as much time as desired to look at the bottles as the respondent “normally would if you were thinking about what you might buy [in a store] or order [at a restaurant].” The respondent was then asked a series of questions intended to determine confusion as to brand; source; affiliation, connection, or sponsorship; and permission. The questions asked for both positive and negative responses and included the possibility that the respondent might have no opinion.8 If a respondent answered “yes” to a question, the interviewer inquired which bottle, noted the response, and asked for and noted the respondent’s reason for responding in the affirmative. Once a respondent identified a bottle, it was removed. Respondents were permitted to identify more than one bottle in response to each question.
118. When analyzing the data, Dr. Kaplan classified a respondent as contused if the respondent identified the CRISTALINO bottle and mentioned the name similarity as the reason. With respect to CRISTALINO, Dr. Kaplan calculated that 6.9% of respondents were confused as to brand, 9.2% were confused as to source, 4.6% were confused as to affiliation, connection, or sponsorship, and 2.3% were confused as to permission for a total confusion of 23.1%. No respondents identified the COURTALINO bottle based on its name.
*862119. Of the respondents who identified the CRISTALINO bottle based on the name, 57% selected only the CRISTALINO bottle, 80% selected the CRISTALINO bottle when all four bottles were present, and 20% selected the CRISTALINO bottle when three bottles were present. The CRISTALINO bottle was never the third or fourth bottle selected. In addition to identifying the name as the reason for an affirmative answer, some respondents gave the similarity of the labels as a reason and stated that the CRISTALINO bottle could be a “less expensive version,” “an American knock-off of the first bottle,” or “a cheaper version of the original” bottle.
120. Nearly 30% of the respondents identified WILLM as being related to CRISTAL. The WILLM bottle is clear and its labels are primarily gold. The gold neck label on the WILLM bottle includes a medallion in the same location as the LR medallion on bottles of CRISTAL champagne.
121. Defendants retained Dr. Itamar Simonson, a professor of marketing at Stanford University, to review Dr. Kaplan’s report. Dr. Simonson studies consumer behavior and has conducted numerous mall intercept surveys. He testified that the survey universe was underinclusive because it should have included the potential respondents who were unaware of CRISTAL champagne. Based on Dr. Kaplan’s quota of 90 people to reach 30 respondents, Dr. Simonson calculated that the total confusion was one-third (30/90) of 23.1%, or 7.7%, of what he considered the proper survey universe. Dr. Simonson also testified that the survey universe was overinclusive because the inclusion of people who had purchased wine for less than $35 would include those who had purchased wine in a liquor store in the $30 range. He believed those who purchased sparkling wine in a liquor store for $30 were more likely to have heard of CRISTAL champagne because $30 is closer to the price of CRISTAL champagne than $7.
122.Dr. Simonson further testified that the survey format did not approximate market conditions because it was unlikely that a consumer would inspect a bottle of CRISTAL champagne immediately before looking at a bottle of CRISTALINO cava. He opined that the failure to approximate market conditions turned the survey into a “matching game” where respondents looked for the “right” answer. In other words, he opined that the survey created “demand effects.” Dr. Simonson testified that the identification of the WILLM bottle as related to the CRISTAL bottle by about 30% of the respondents is evidence of those demand effects because there was “no obvious reason” why the respondents would give a positive response with respect to confusion for the WILLM bottle. Dr. Simonson later testified that the primary reason respondents identified the WILLM bottle was because it was clear, as was the CRISTAL bottle. Finally, Simonson opined that “COURTALINO” was an inappropriate control and that a control should be “as similar to the junior mark, in this case, CRISTALINO, as possible, without infringing on the senior mark.”
NON-CONFIDENTIAL CONCLUSIONS OF LAW
I. Jurisdiction
1. Roederer asserts claims of trademark infringement, unfair competition, and trademark dilution under the Lanham Act, 15 U.S.C. §§ 1051-1141n(2006). Roederer also asserts a common-law unfair competition claim and a claim under the Minnesota Deceptive Trade Practices Act, Minn.Stat. §§ 325D.43-.48 (2008).
*8632. The Court has jurisdiction over the Lanham Act claims pursuant to 15 U.S.C. § 1121 and 28 U.S.C. § 1388 (2006). The Court has supplemental jurisdiction over the state-law claims pursuant to 28 U.S.C. § 1367 (2006).
II. Trademark infringement and unfair competition
3. Roederer asserts claims of trademark infringement and unfair competition under §§ 32 and 43(a) of the Lanham Act. To succeed on those claims, Roederer must show that it has a valid, protectable mark and that there is a likelihood of confusion between its mark and the CRISTALINO mark. See B & B Hardware, Inc. v. Hargis Indus., Inc., 569 F.3d 383, 389 (8th Cir.2009); see also Eniva Corp. v. Global Water Solutions, Inc., 440 F.Supp.2d 1042, 1049 n. 3 (D.Minn.2006) (noting that claims under §§32 and 43(a) of the Lanham Act “are measured by identical ‘likely to cause confusion’ standards”). The Court does not independently analyze Roederer’s state-law claims because the parties agree they are coextensive with the Lanham Act claims. See DaimlerChrysler AG v. Bloom, 315 F.3d 932, 935 n. 3 (8th Cir. 2003).
A. Validity of the CRISTAL marks
4. Roederer’s federal registration for the '343 mark for CRISTAL CHAMPAGNE & Design for use in connection with champagne is prima facie evidence of the validity of the '343 mark. See 15 U.S.C. § 1115(a); see also Aromatique, Inc. v. Gold Seal, Inc., 28 F.3d 863, 869 (8th Cir.1994) (“[Registration of a mark creates a rebuttable presumption that the mark is valid.”). Defendants have offered no evidence to rebut that presumption. The '343 mark is valid and protectable.
5. The U.S. Patent & Trademark Office accepted Roederer’s certificate of incontestability for the '998 mark for the words CRISTAL CHAMPAGNE for use in connection with champagne. Accordingly, the '998 mark is incontestable and valid. See 15 U.S.C. § 1065; see also B & B Hardware, 569 F.3d at 389 (“Once a mark becomes incontestable, it ‘cannot be challenged ... for mere descriptiveness, or on the basis that the mark lacks secondary meaning.’ ”) (quoting Sunrise Jewelry Mfg. Corp. v. Fred S.A., 175 F.3d 1322, 1324 (Fed.Cir.1999)).
6. Roederer also asserts rights in a common-law CRISTAL mark used in connection with champagne. “[A] common-law trademark arises from the adoption and actual use of a word, phrase, logo, or other device to identify goods or services with a particular party.” First Bank v. First Bank Sys., Inc., 84 F.3d 1040, 1044 (8th Cir.1996). Roederer owns the common-law CRISTAL mark for use in connection with champagne by virtue of its continuous use in the United States of the word “CRISTAL” to identify its champagne for over six decades.
7. A trademark is entitled to trademark protection if it is distinctive. Schwan’s IP, LLC v. Kraft Pizza Co., 460 F.3d 971, 974 (8th Cir.2006). Roederer bears the burden of establishing that its unregistered CRISTAL mark is protectable under trademark law. Frosty Treats, Inc. v. Sony Computer Entm’t Am., Inc., 426 F.3d 1001, 1003 (8th Cir.2005). Although Defendants assert the CRISTAL marks are weak, they do not dispute the validity of the common-law CRISTAL mark. Moreover, for the reasons set forth below with respect to the conceptual strength of the CRISTAL marks, the common-law CRISTAL mark is suggestive. Because it is suggestive, the common-law CRISTAL mark is valid and protectable. See Co-Rect Prods., Inc. v. Marvy! Adver. Photography, Inc., 780 F.2d 1324, 1329 (8th Cir.1985).
*864 B. Likelihood of confusion
8. Roederer asserts a likelihood of confusion as to association or sponsorship between the use of “CRISTALINO” on cava and its CRISTAL marks. Accordingly, the question is whether Defendants’ use of the CRISTALINO mark so resembles Roederer’s use of the CRISTAL marks that it is likely to cause confusion among consumers as to whether the source of CRISTAL champagne has sponsored, endorsed, or is otherwise affiliated with CRISTALINO cava. See Mut. of Omaha Ins. Co. v. Novak, 836 F.2d 397, 398 (8th Cir.1987); see also Insty*Bit, Inc. v. Poly-Tech Indus., Inc., 95 F.3d 663, 671 (8th Cir.1996) (explaining that a finding of trade dress infringement requires consumers to “purchase the alleged infringer’s products after associating the trade dress of those products with the trade dress of a single, albeit anonymous source”); Anheuser-Busch, Inc. v. Balducci Publ’ns, 28 F.3d 769, 774 (8th Cir.1994) (explaining that likelihood of confusion protects against use of a plaintiffs mark on any product that would reasonably be thought by the buying public to be affiliated with, connected with, or sponsored by, the trademark owner). The likelihood of confusion must be “among an appreciable number of consumers.” Gateway Inc. v. Companion Prods., Inc., 384 F.3d 503, 509 (8th Cir.2004).
9. Likelihood of confusion is a question of fact. SquirtCo v. Seven-Up Co., 628 F.2d 1086, 1090-91 (8th Cir.1980). Several factors are pertinent to the likelihood of confusion analysis, including (1) the strength of the trademark; (2) the similarity between the plaintiffs and defendant’s marks; (3) the competitive proximity of the parties’ products; (4) the alleged infringer’s intent to confuse the public; (5) evidence of any actual confusion; and (6) whether the kind of product, its cost, and the conditions of purchase can eliminate the likelihood of confusion that would otherwise exist. Id. at 1090-91. These factors do not operate as a precise test, but instead represent the type of considerations a court should examine in determining whether likelihood of confusion exists. See also Duluth News-Tribune v. Mesabi Publ’g Co., 84 F.3d 1093, 1096 (8th Cir.1996). “When balancing the interests in a famous, established mark against the interests of a newcomer, [a court is] compelled to resolve doubts on this point against the newcomer.” Specialty Brands, Inc. v. Coffee Bean Distribs., Inc., 748 F.2d 669, 674 (Fed.Cir.1984).
1. Strength of Roederer’s trademarks
10. A mark’s strength consists of both conceptual strength and commercial strength in the marketplace. George & Co. v. Imagination Entm’t Ltd., 575 F.3d 383, 393 (4th Cir.2009); Mars Musical Adventures, Inc. v. Mars, Inc., 159 F.Supp.2d 1146, 1150 (D.Minn.2001); see also 2 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 11:83 (4th ed. 2010) (“While some courts have made the strong-weak evaluation solely upon the place of a term on the spectrum of marks, such an approach is incomplete. One must in addition look at the marketplace strength of the mark at the time of litigation or at the time registration is sought.” (footnote omitted)).
a. Conceptual strength
11. The conceptual strength of a trademark depends on its classification as (1) arbitrary or fanciful, (2) suggestive, (3) descriptive, or (4) generic. See Insty*Bit, 95 F.3d at 672-73. An arbitrary mark is one that uses a common word to describe a product in an unfamiliar way, while a fanciful mark is one invented solely for use as a trademark. Id. *865at 673 n. 10. A suggestive mark requires some measure of imagination to reach a conclusion regarding the nature of the product. Id. at 673 n. 11. “An arbitrary, fanciful, or suggestive mark is deemed inherently distinctive, and therefore entitled to protection, because its ‘intrinsic nature serves to identify a particular source of a product.’ ” Id. at 673 (quoting Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 768, 112 S.Ct. 2753, 120 L.Ed.2d 615 (1992)).
12. “A term is descriptive if it conveys an ‘immediate idea of the ingredients, qualities or characteristics of the goods.’ ” Frosty Treats, 426 F.3d at 1005 (quoting Stuart Hall Co., Inc. v. Ampad Corp., 51 F.3d 780, 785-86 (8th Cir.1995)). A descriptive mark is protectable only if it has become distinctive by acquiring secondary meaning. Id. A trademark has acquired secondary meaning if it has become so associated in the public mind with certain goods that it serves to identify them and distinguish them from other goods. Id. A generic term, which is merely used by the general public to identify a category of goods, is not protectable. Id.
13. An arbitrary or fanciful mark is the strongest possible trademark and is entitled to the “maximum degree of legal protection.” First Bank, 84 F.3d at 1045. A descriptive trademark is the weakest protectable mark. Id. The strength of a suggestive mark falls between that of an arbitrary or a fanciful mark and that of a descriptive mark. See id. A strong and distinctive trademark is entitled to greater protection than a weak or common mark. Frosty Treats, 426 F.3d at 1008. If a mark is weak, consumer confusion is unlikely because the mark’s components are so widely used that the public can easily distinguish slight differences in the marks, even if the goods are related. Gen. Mills, Inc. v. Kellogg Co., 824 F.2d 622, 626 (8th Cir.1987).
14. The incontestability of the '998 mark does not prevent an inquiry into its strength. See Lone Star Steakhouse & Saloon, Inc. v. Alpha of Va., Inc., 43 F.3d 922, 934-35 (4th Cir.1995). Roederer obtained its registration for the '998 mark for CRISTAL CHAMPAGNE by claiming the mark had acquired distinctiveness under Section 2(f) of the Lanham Act. Consequently, Roederer has conceded that the '998 mark is descriptive and not inherently distinctive. See Aromatique, 28 F.3d at 869. The '998 mark is the weakest protectable mark with respect to conceptual strength.
15. Defendants assert that Roederer’s concession of descriptiveness in its registration of the '998 mark extends to the '343 mark and the common-law CRISTAL mark. In Aromatique, the court held that “[t]he submission of evidence under Section 2(f) to show secondary meaning ... amounts to a concession that the mark sought to be registered is not inherently distinctive.” Id. (emphasis added). Moreover, it is well-established that federal registration of a mark does not affect the registrant’s common-law rights in a mark because those rights arise from use, not registration. See Gilbert/Robinson, Inc. v. Carrie Beverage-Mo., Inc., 989 F.2d 985, 991-92 (8th Cir.1993); 4A Louis Altman & Malla Pollack, Callmann on Unfair Competition, Trademarks & Monopolies § 26:3 (4th ed. 2010) (explaining that federal registration does not affect common-law trademark rights because rights in a trademark are acquired by use); 3 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 19:3 (same); cf. 15 U.S.C. § 1056(b) (limiting effect of disclaimers to mark for which registration was sought). Roederer’s registration of the '998 mark under Section 2(f) does not affect the conceptual strength of the '343 mark or the common-law CRISTAL mark.
*86616. Defendants also contend that Roederer’s disclaimer of the words “CRISTAL CHAMPAGNE” in its registration for the '343 mark and of the word “CHAMPAGNE” in its registration for the '998 mark are admissions that the term “CRIS-TAL” is descriptive. The disclaimers in the registrations for the '343 and '998 marks disclaim only a claim that the registrations give Roederer “an exclusive right in those disclaimed words or symbols per se.” 3 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 19:63. “No disclaimer ... shall prejudice or affect the applicant’s or registrant’s rights then existing or thereafter arising in the disclaimed matter, or his right of registration on another application if the disclaimed matter be or shall have become distinctive of his goods or services.” 15 U.S.C. § 1056(b); see also Official Airline Guides, Inc. v. Goss, 856 F.2d 85, 87 (9th Cir.1988) (“OAG’s disclaimer of the phrase ‘Travel Planner’ in its registration does not deprive it of any common law rights it may have in the disclaimed matter.”). The '343 mark’s disclaimer does not affect Roederer’s rights in the '998 mark, nor does either disclaimer affect Roederer’s rights in the common-law CRISTAL mark.
17. With respect to the conceptual strength of the '343 and common-law marks, Roederer contends that “CRIS-TAL” is arbitrary as applied to champagne. Relying on the English translation of “CRISTAL” as “crystal,” Defendants maintain that “CRISTAL” is descriptive because it describes the material from which bottles of CRISTAL champagne are made.
18. Roederer does not claim rights in a CRISTAL mark for use in connection with bottles. The term “CRISTAL” identifies the source of the champagne, not of the bottle in which CRISTAL champagne is sold. “CRISTAL” is not descriptive because it does not immediately convey the ingredients, qualities, or characteristics of the champagne it identifies. See Frosty Treats, 426 F.3d at 1005. Champagne is neither crystal nor quartz. Moreover, “CRISTAL” is not descriptive because a competitor does not need to use the word “CRISTAL” (or “crystal”) to describe its sparkling wine. See 2 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 11:68 (“If, however, the message conveyed by the mark about the goods or services is so direct and clear that competing sellers would be likely to need to use the term in describing or advertising their goods, then this indicates that the mark is descriptive.”).
19. “Arbitrary marks are comprised of words in common usage, but, because they do not suggest or describe any quality, ingredient, or characteristic of the goods they serve, are said to have been arbitrarily assigned.” Sara Lee Corp. v. Kayser-Roth Corp., 81 F.3d 455, 464 (4th Cir.1996). The term “CRISTAL” is not arbitrary because it suggests the sparkling quality of the champagne once it is released from the bottle. However, it requires imagination and reasoning by consumers to make the connection between the sparkle of crystal and the sparkle of champagne. Consequently, “CRISTAL” is suggestive when used in connection with champagne. See Insty*Bit, 95 F.3d at 673 n. 11; 2 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 11:67 (“[I]f one must exercise mature thought or follow a multi-stage reasoning process to determine attributes of the product or service, the term is suggestive, not descriptive.” (quotation marks omitted)). Because the '343 and common-law CRISTAL marks are suggestive, they are conceptually strong.
b. Commercial strength
20. In the likelihood of confusion context, a mark’s commercial strength or *867“fame” is determined based on the “public recognition and renown” of the mark as evidenced by the extent of advertising, sales volume, features and reviews in publications, and survey evidence. See, e.g., Palm Bay Imports, Inc. v. Veuve Clicquot Ponsardin Maison Fondee En 1772, 396 F.3d 1369, 1374-76 (Fed.Cir.2005) (considering evidence of advertising expenditures, sales volume, articles and reviews in general — and special-interest magazines in commercial strength analysis); Insty*Bit, 95 F.3d at 670 (finding favorable reviews in special-interest magazines and survey evidence demonstrated the strength of plaintiffs trade dress); ConAgra, Inc. v. George A. Hormel & Co., 990 F.2d 368, 369 (8th Cir.1993) (including “marketplace recognition value” in assessment of trademark’s strength). For purposes of the likelihood of confusion analysis, the relevant market for evaluating commercial strength or fame is “the class of customers and potential customers of a product or service, and not the general public.” Palm Bay Imports, 396 F.3d at 1375. This determination is made at “the time the mark is asserted in litigation.” Renaissance Greeting Cards, Inc. v. Dollar Tree Stores, Inc., 405 F.Supp.2d 680, 690 (E.D.Va. 2005); ConAgra, Inc. v. George & Hormel & Co., 784 F.Supp. 700, 707 (D.Neb.1992).
21. With respect to the recognition and renown of the mark, Roederer has used “CRISTAL” in the United States to identify its champagne since at least 1937. CRISTAL champagne has been known by connoisseurs and those in the wine industry as one of the best champagnes since at least the 1970’s. Numerous favorable reviews in wine publications recognize the quality of CRISTAL champagne. Retailers feature CRISTAL champagne in then-stores and advertisements and use it to promote their businesses. Although the sales volume of CRISTAL champagne is relatively small, it is sold in 4500 liquor stores across the United States and, with the exception of 2009, demand for CRISTAL champagne has always exceeded supply. All of these facts support a finding of a high level of commercial strength.
22. In addition, the CRISTAL brand has achieved significant recognition by the general public over the past two decades. Cuisine, wine, and general-interest magazines have reviewed CRISTAL champagne in articles and featured it on their covers. For the past two decades, CRISTAL champagne has been frequently used as a product placement in television shows and movies without Roederer paying for such placement. CRISTAL champagne has been commonly mentioned in popular novels. Although Defendants characterize them as “fleeting,” the quantity and continuity of the media mentions and product placements indicate that today’s general public recognizes CRISTAL champagne as a high-status product. Moreover, CRISTAL champagne played more than a peripheral role in some of the television shows. Finally, the general media has commented on the extensive publicization and recognition of the CRISTAL brand due to its prominence in rap music. The general renown of CRISTAL champagne is further evidence of the commercial strength of the CRISTAL mark.
23. Defendants contend that the evidence of third-party registrations of marks and trade names including CRISTAL or similar terms is evidence of the commercial weakness of the CRISTAL marks. “[EJvidence of third party usage of similar marks on similar goods is admissible and relevant to show that the mark is relatively weak and entitled to a narrower scope of protection.” Gen. Mills, 824 F.2d at 626-27. Such evidence indicates a mark is entitled to a narrower scope of protection because consumers are already “conditioned” to look for minor distinctions in marks and are less likely to be confused. *868See In re Mighty Leaf Tea, 601 F.3d 1342, 1346 (Fed.Cir.2010); Palm Bay Imports, 396 F.3d at 1374. However, the probative value of such third party registrations and trade names “depends entirely upon their usage.” Palm Bay Imports, 396 F.3d at 1373. “Third party registrations are not evidence of use so as to have conditioned the mind of prospective purchasers.” 2 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 11:89 (quotation marks omitted). In the absence of evidence about the extent of the uses, the probative value of such third-party registrations and trade names is minimal. Han Beauty, Inc. v. Alberto-Culver Co., 236 F.3d 1333, 1338 (Fed.Cir.2001); see also 2 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 11:88 (4th ed. 2010) (“To present a more compelling case, [a] defendant should go further to show how extensive these uses are and how long they have continued.”).
24.There is no evidence of the extent of the sales or publicity of the products sold under the third-party registrations and trade names. In the absence of such evidence, the third-party registrations for and Internet availability of products including CRISTAL or similar terms in their names do not indicate that U.S. consumers are aware of those marks, much less that they have become conditioned to distinguish between CRISTAL champagne and other alcoholic beverages having CRIS-TAL or similar terms in their names. Cf. CareFirst of Md., Inc. v. First Care, P.C., 434 F.3d 263, 270 (4th Cir.2006) (relying in part on investigator’s report confirming that businesses identified based on trademark search reports and web page printouts from the businesses were currently active when concluding that plaintiffs mark was not conceptually strong). Moreover, Roederer’s successful enforcement of its rights in the CRISTAL marks against CRISTAL CAVA CASTELLBLANCH cava enhances the commercial strength of the CRISTAL marks. See Morningside Group Ltd. v. Morningside Cap. Group, L.L.C., 182 F.3d 133, 139 (2d Cir.1999). Consequently, the evidence of third-party registrations and Internet listings does not alter the Court’s conclusion that the CRIS-TAL marks are commercially strong.
25. In summary, the '998 mark does not enjoy great conceptual strength because it is descriptive, but the other two CRISTAL marks are suggestive and therefore conceptually strong. Based on the general renown of CRISTAL champagne and recognition of the CRISTAL marks, all three of the CRISTAL marks are commercially strong. When the CRISTAL marks’ conceptual and commercial strength are considered in combination, all three marks are “strong” for likelihood of confusion purposes. The first SquirtCo factor therefore weighs in favor of a finding of likelihood of confusion.
2. Competitive proximity
26. The second SquirtCo factor requires consideration of whether the products are in competitive proximity. 628 F.2d at 1091. Although CRISTAL champagne and CRISTALINO cava do not compete directly for sales, trademark infringement may be found in the absence of direct competition. See id. Thus, the competitive proximity factor requires a broader examination of the products’ relationship in the market. Kemp v. Bumble Bee Seafoods, Inc., 398 F.3d 1049, 1056 (8th Cir.2005).
27. Products are in competitive proximity when there is similarity or overlap in their sales outlets, trade channels, and customers. See Xtreme Lashes, LLC v. Xtended Beauty, Inc., 576 F.3d 221, 234 (5th Cir.2009) (finding “outlet and purchaser identity” weighing in favor of likelihood of confusion where products were sold to the same class of professional beauticians and their clients and advertised through *869similar media channels); In re Chatam Int’l Inc., 380 F.3d 1340, 1344-45 (Fed.Cir. 2004) (finding a close relationship weighing in favor of likelihood of confusion with respect to beer and tequila because they were sold through many of the same channels of trade to many of the same consumers); Star Indus., Inc. v. Bacardi & Co., 412 F.3d 373, 386 (2d Cir.2005) (finding that vodka and rum were in competitive proximity because they were sold in the same locations which were frequented by the same consumers); J & B Wholesale Distrib., Inc. v. Redux Beverages, LLC, 621 F.Supp.2d 678, 685-86 (D.Minn.2007) (finding competitive proximity based on overlapping geographic distribution and similarity of trade channels). This factor favors a finding of likelihood of confusion when it is reasonable for consumers to think the products came from the same source or are somehow affiliated. See Kemp, 398 F.3d at 1056; J & B Wholesale Distrib., 621 F.Supp.2d at 686; 4 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 24:52 (“Even if the goods are vastly differently priced and are clearly not in competition, this does not mean that low-priced buyers have not at least heard of the expensive goods and may think that an expensive seller has expanded into a low-priced field.”).
28. The sales and distribution channels of CRISTAL champagne and CRISTALINO cava overlap. Moreover, the publications in which their advertising and reviews appear target the same educated, relatively affluent consumers, some of whom will purchase a more-expensive sparkling wine for one occasion and a less-expensive sparkling wine for another. This overlap favors a conclusion of competitive proximity. The impact of this competitive proximity on the likelihood of confusion as to association or sponsorship is not significantly affected by the differences in price and manner of display between CRISTAL champagne and CRISTALINO cava. See Star Indus., 412 F.3d at 387 (finding proximity factor favored plaintiff who alleged confusion as to affiliation, sponsorship, or connection even though “[defendant’s] vodka costs half as much as [plaintiffs] rum and is displayed on different shelves in the same store”). Finally, it is reasonable for consumers to think CRISTAL champagne and CRISTALINO cava are affiliated because it is common for wineries that produce prestige wines, including Roederer, to expand their product lines to include lower-priced products. The competitive proximity factor weighs in favor of a finding of likelihood of confusion.
3. Similarity of the marks
29. Similarity of the marks supports a finding of likelihood of confusion. SquirtCo, 628 F.2d at 1091. In evaluating the similarity of the marks, a court considers the overall impression created by the marks, including their trade dress and visual, aural, and definitional attributes, to determine whether “the total effect conveyed by the two marks is confusingly similar.” Luigino’s, Inc. v. Stouffer Corp., 170 F.3d 827, 830 (8th Cir.1999).
30. “CRISTAL,” as the only word in the CRISTAL marks and the first portion of the CRISTALINO mark, is a dominant term in the CRISTAL and CRISTALINO marks. This dominance supports a finding of similarity. See Palm Bay Imports, 396 F.3d at 1372-73; Kemp, 398 F.3d at 1055. This similarity is increased because “CRISTAL” does not have a meaning in English other than to identify CRISTAL champagne or as a proper name. Cf. Luigino’s, 170 F.3d at 830 (finding marks dissimilar when dominant term “lean” was generally descriptive of food and not registerable as a trademark). Moreover, “INO” is likely to suggest to consumers that “CRISTALINO” is a diminutive of the root word “CRISTAL” *870or at least associate “CRISTALINO” with “CRISTAL.” The marks have some aural similarity despite the difference in accent and pronunciation of the “CRISTAL” portions. The media mentions and Internet postings describing CRISTALINO cava as CRISTAL champagne’s younger brother, pretending that CRISTALINO cava is CRISTAL champagne, and suggesting that consumers purchase CRISTALINO cava instead of CRISTAL champagne further support a finding of association and aural and phonetic similarity.9
31. Defendants suggest that the marks are distinguishable in meaning because “CRISTAL” translates to “crystal” and “CRISTALINO” translates to “crystalline.” Under the doctrine of foreign equivalents, foreign words from common languages are translated into English to determine “similarity of connotation in order to ascertain confusing similarity with English word marks.” Palm Bay Imports, 396 F.3d at 1377. This doctrine applies when it is likely that an American consumer “will translate the foreign mark and will take it as it is.” Id. To the extent that consumers would translate “CRIS-TAL” and “CRISTALINO” to “crystal” and “crystalline,” the different translations do not weigh against a finding of definitional similarity because the definition of “crystalline” includes “made of crystal” or “resembling crystal.” The word marks have definitional similarity.
32. In addition to considering the similarity of the word marks, the Court must “evaluate the impression that each mark in its entirety is likely to have on a purchaser exercising the attention usually given by purchasers of such products.” See Duluth News-Tribune, 84 F.3d at 1097. In making this comparison, the appearance of the litigated marks side by side in the courtroom does not accurately reflect market conditions. Daddy’s Junky Music Stores, Inc. v. Big Daddy’s Family Music Center, 109 F.3d 275, 283 (6th Cir. 1997) (quotation marks omitted). “Rather, courts must determine whether a given mark would confuse the public when viewed alone, in order to account for the possibility that sufficiently similar marks may confuse consumers who do not have both marks before them but who may have a general, vague, or even hazy, impression or recollection of the other party’s mark.” Id. (quotation marks omitted).
33. The CRISTAL champagne and CRISTALINO cava bottles are the same shape. Bottles of CRISTAL champagne and CRISTALINO cava have rectangular gold front labels, gold neck labels, and a gold shoulder label. Bottles of CRISTAL ROSÉ champagne and CRISTALINO ROSÉ cava have the same elements in a pink-hued copper color. The front labels prominently display the word “CRISTAL” or “CRISTALINO” in all-capital letters in a Roman serif font, as do both neck labels. Both bottles display a medallion on the front label and in the downwardly-extending region of the neck label. These common elements support a finding of similarity with respect to the overall commercial impression. In addition, Carrion’s failure to prominently display the “Jaume Serra” name (or any other distinguishing feature) on bottles of CRISTALINO cava further supports a finding of similarity. Cf. Luigino’s, 170 F.3d at 831 (finding that prominent display of house marks conveyed “perceptible distinctions” between products).
*87134. Defendants contend that the differences in the color of the bottles, the language used on the labels, the geometry of the bottles’ bases, the display of CRIS-TAL champagne in yellow cellophane, and the details of certain design elements, as well as the overall more-expensive appearance of the CRISTAL bottle, mean the products’ commercial impressions are dissimilar. Where the products are closely related, less similarity is required to support a finding of infringement. ConAgra, 990 F.2d at 371; SquirtCo, 628 F.2d at 1091. Moreover, the differences identified by Defendants do not weigh against the likelihood of the confusion alleged here— confusion as to association or sponsorship — because those differences exist between Roederer’s CRISTAL champagne and Roederer’s less-expensive sparkling wines. Indeed, the verbatim responses describing the CRISTALINO bottle as “less expensive version,” an “American knock-off,” and a “cheaper version” of the CRISTAL bottle support this conclusion. Similarly, inclusion of the “Value Brand of the Year” sticker on bottles of CRISTALINO cava is not inconsistent with the belief that the producer of CRISTAL champagne has expanded its product offering to include a less-expensive, yet relatively high-quality, sparkling wine.10 This factor weighs in favor of a finding of likelihood of confusion.
4. Intent
35. Although not an element of a claim for trademark infringement, intent on the part of the alleged infringer to pass off its goods as the product of another creates an inference of likelihood of confusion. SquirtCo, 628 F.2d at 1091. The intent factor is relevant because it demonstrates the junior user’s true opinion as to the dispositive issue of whether confusion is likely. Kemp, 398 F.3d at 1057.
36. Jaume Serra’s use of the CRISTALINO name on still wines before introducing CRISTALINO cava to the United States in 1989 weighs against a finding of intent. Roederer asserts, however, that the changes in the CRISTALINO labels between 1989 and 1993 demonstrate Jaume Serra’s intent to trade off the CRISTAL name. Defendants have offered credible testimony that “JAUME SERRA” was removed from the CRIS-TALINO label to make it less cluttered. The Court also credits Friend’s testimony that the color of the CRISTALINO BRUT label was changed from black to gold based on the Wine Max buyer’s recommendation. In addition, gold labels, while not comprising the majority of sparkling wine labels, are not so rare as to require the conclusion that the color of the label was changed with the intent to copy the labels of CRISTAL champagne. Similarly, pink-hued labels for rosé sparkling wines are not uncommon. The Court concludes that the changes to the CRISTALINO labels made between 1989 and 1993 and the selection of pink-hued copper labels for CRISTALINO ROSÉ cava do not support a conclusion of intent to infringe.
37. Roederer maintains that Defendants’ promotion of CRISTALINO cava as a higher-quality sparkling wine made using the traditional method with yeast from the Champagne region of France supports a finding of intent. Nothing in Defendants’ promotional materials indicates any intent to link CRISTALINO cava to Roederer or CRISTAL champagne. Defen*872dants’ statements in its advertisements suggesting that consumers can avoid embarrassment by bringing CRISTALINO cava to a party do not indicate an intent to position CRISTALINO as an aspirational or French brand rather than a cava superi- or to other comparably-priced sparkling wines. Defendants’ promotion of CRIS-TALINO cava does not support a finding of intent to benefit from an association with CRISTAL champagne or Roederer.
38.Roederer also asserts that Defendants are willfully indifferent to its trademark rights because Carrion did not conduct any trademark searches before using the CRISTALINO mark on cava in the United States or respond to a cease-and-desist letter Roederer sent Carrion in 2002. Carrion’s failure to conduct a trademark search does not demonstrate an intent to infringe or willful indifference to Roederer’s trademark rights. See George & Co., 575 F.3d at 398 (“Finally, the failure to conduct a trademark search or contact counsel shows carelessness at most, but is in any event irrelevant because knowledge of another’s goods is not the same as an intent to mislead and to cause consumer confusion.” (quotation marks omitted)); see also King of the Mountain Sports, Inc. v. Chrysler Corp., 185 F.3d 1084, 1091-92 (10th Cir.1999) (declining to infer intent based on failure to conduct trademark search). Similarly, Carrion’s “refusal to abandon [the CRISTALINO] mark in the face of a cease and desist letter cannot demonstrate bad faith standing alone.” O’Keefe v. Ogilvy & Mather Worldwide, Inc., 590 F.Supp.2d 500, 525 (S.D.N.Y. 2008); see also Straus v. Notaseme Hosiery Co., 240 U.S. 179, 181, 36 S.Ct. 288, 60 L.Ed. 590 (1916) (“[T]he defendants’ persistence in their use of the design after notice proves little or nothing against them.”). No evidence indicates Defendants willfully disregarded Roederer’s trademark rights.
39. Finally, Roederer contends that Defendants did “nothing” after receiving notice of counterfeiting in the marketplace. Roederer bases this argument on the 2004 sale of CRISTALINO cava that had been altered to resemble CRISTAL champagne to a Michigan liquor store owner. Rather than failing to respond, the distributor of CRISTALINO cava immediately informed the liquor store owner of his mistake. Defendants’ failure to do more in response to this isolated instance of counterfeiting by an unknown person does not permit an inference of intent on the part of Defendants to trade off the CRISTAL mark.
40. In conclusion, the evidence does not support a finding that Defendants intended to confuse the public. This factor weighs against a finding of likelihood of confusion.
5. Actual confusion
41. “[A]ctual confusion is not essential to a finding of trademark infringement, although it is positive proof of likelihood of confusion.” SquirtCo, 628 F.2d at 1091. Evidence of actual confusion may be presented in the form of testimony about incidents of confusion or survey evidence. See Frosty Treats, 426 F.3d at 1009.
a. Earlier’s and McCarthy’s testimony
42. Roederer maintains that Barker’s and McCarthy’s testimony about incidents of actual confusion weighs in favor of a likelihood of confusion. When determining whether a likelihood of confusion exists based on instances of actual confusion, weight is given to the number and extent of the instances. Id. Barker and McCarthy provided very few specifics about the details of the conversations and the number of instances. Consequently, the Court gives their testimony little weight with respect to the issue of actual confusion. See id. at 1009-10.
43. Defendants maintain that the lack of testimony from consumers about *873actual confusion, in light of the twenty-year coexistence in the marketplace of CRISTAL champagne and CRISTALINO cava, weighs against a finding of likelihood of confusion. The absence of testimony from consumers about incidents of actual confusion, particularly when the products have coexisted in the marketplace for an extended period of time, can indicate that confusion is not likely. Hubbard Feeds, Inc. v. Animal Feed Supplement, Inc., 182 F.3d 598, 603 (8th Cir.1999). However, in light of CRISTALINO cava’s inexpensive price and the prohibition on direct communication between consumers and Roederer and MMD, the absence of consumer testimony about actual confusion is not significant. See Au-Tomotive Gold, Inc. v. Volkswagen of Am., Inc., 457 F.3d 1062, 1077 (9th Cir.2006) (“In this case, which involves a national market and a low degree of consumer care, nothing suggests that the lack of evidence of confusion should be particularly noteworthy.”). Barber’s and McCarthy’s testimony and the absence of testimony from actual consumers does not weigh in favor of or against a finding of likelihood of confusion.
b. Survey evidence
44. Consumer surveys are an appropriate method for producing evidence of actual confusion. Stuart Hall, 51 F.3d at 790. “Because manifestations of actual confusion serve as strong evidence of a likelihood of confusion, and may, in fact, be the best such evidence, [surveys] should be given substantial weight unless seriously flawed.” Novak, 836 F.2d at 400 (citations omitted). “The probative value of a consumer survey is a highly fact-specific determination and a court may place such weight on survey evidence as it deems appropriate.” Johnson & Johnson-Merck Consumer Pharms. Co. v. Rhone-Poulenc Rarer Pharms., Inc., 19 F.3d 125, 134 (3d Cir.1994) (quotation marks omitted).
45. Dr. Kaplan calculated a 23.1% level of confusion based on his survey results. Survey results indicating that level of confusion generally would weigh in favor of a finding of likelihood of confusion. See Sara Lee Corp., 81 F.3d at 467 (“[B]ut even if the true figure were only half of the [thirty to forty percent estimated by the survey], actual confusion would, in our view, nevertheless exist to a significant degree.”); Novak, 836 F.2d at 400-01 (finding a 10% result from a credible survey supported a finding of likelihood of confusion); James Burrough, Ltd. v. Sign of Beefeater, Inc., 540 F.2d 266, 279 (7th Cir.1976) (“We cannot agree that 15% is ‘small.’ Though the percentage of likely confusion required may vary from case to case, we cannot consider 15%, in the context of this case, involving the entire restaurant-going community, to be de mini-mus.”). Defendants, however, argue that the Kaplan survey is flawed for several reasons.11
*87446. Defendants first contend that the Kaplan survey improperly defined the survey universe. The survey universe is “that segment of the population whose perceptions and state of mind are relevant to the issues in the case.” Citizens Fin. Group, Inc. v. Citizens Nat’l Bank of Evans City, 383 F.3d 110, 118-19 (3d Cir.2004) (quotation marks omitted). A survey of the wrong universe has little probative value. Id. at 119. In a “forward confusion” case such as this, the alleged confusion “occurs when customers mistakenly think that the junior user’s goods or services are from the same source as or are connected with the senior user’s goods or services.” Sands, Taylor & Wood Co. v. Quaker Oats Co., 978 F.2d 947, 957 (7th Cir.1992) (quotation marks omitted). In a forward confusion ease, “the proper universe to survey is the potential buyers of the junior user’s goods or services.” 6 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 32:159.
47. Defendants first contend that the survey was underinclusive because respondents were limited to those already aware of CRISTAL champagne. Using Dr. Kaplan’s estimate that the interviewers at each survey site would need to bring ninety people to the interview room to reach the desired number of thirty respondents per site, Defendants contend that the level of actual confusion was only one-third of 23.1%, or 7.7%. To some extent, the restriction of respondents to those already aware of CRISTAL champagne artificially inflated the confusion level measured by the survey. See Paco Sport, Ltd. v. Paco Rabanne Parfums, 86 F.Supp.2d 305, 322 n. 17 (S.D.N.Y.2000) (“The Court finds Paco Rabanne’s approach, limiting its universe to consumers aware of its products, inappropriate because even a weak brand could demonstrate a high degree of confusion because of the limited nature of the universe being surveyed. The Court therefore agrees with Paco Sport’s expert, Mr. McCullough, that the more appropriate practice is to survey all prospective purchasers.”).
48.Roederer’s reliance on IDV North America, Inc. v. S & M Brands, Inc., 26 F.Supp.2d 815, 830 (E.D.Va.1998), and similar cases in support of its exclusion of those unaware of CRISTAL champagne is unpersuasive. In those cases, courts reasoned that excluding those in the market for the senior product could increase the likelihood of confusion because those persons were more likely to know the source of the senior product did not sponsor the product identified by the junior mark. The issue here, however, is not whether those in the market for CRISTAL champagne were properly included, but rather whether those unaware of the CRISTAL mark were properly excluded. Roederer offers no plausible explanation of how persons unaware of CRISTAL champagne could be confused as to whether CRIS-TALINO cava was sponsored by the source of the product of which they were unaware — CRISTAL champagne. Although the exclusion of those unaware of CRISTAL champagne inflated the likelihood of confusion estimate, there is no evidence indicating how many people it actually took to reach the desired number of thirty respondents per site. Conse*875quently, while the Court decreases the weight given to Dr. Kaplan’s survey due to the underinclusive universe, it declines to conclude that the actual confusion was 7.7%.
49. Second, Defendants assert that the $35 price point for respondents is too high because CRISTALINO cava sells for less than $10 in liquor stores. The screening interview determined if potential respondents had purchased or intended to purchase an imported sparkling wine for under $35 per bottle in a liquor store or restaurant, but did not ask the potential respondents in which location they had purchased or intended to purchase the sparkling wine. Insofar as the $35 price point captured respondents who had purchased CRISTALINO cava at restaurants, it was reasonable since CRISTALINO cava sells for between $19 and $32 in restaurants. The survey universe, however, was overinclusive to the extent it included respondents who had purchased an imported sparkling wine in the $30 range at a liquor store. Relying on Dr. Simon-son’s testimony, Defendants contend that this overinclusiveness meant the respondents were more likely to have heard of CRISTAL champagne. To the extent this flaw increased the number of respondents who had heard of CRISTAL champagne, its effect is subsumed in the effect of restricting survey respondents to those aware of CRISTAL champagne. The Court does not further alter the weight given the survey based on this flaw.
50. Defendants next contend that the use of “COURTALINO” as a control was inadequate because the only similarity between “CRISTAL” and “COURT” is that both words begin with the letter “C.” The proper comparison when determining the validity of a control is between the junior mark and the control, not between the allegedly infringing portion of the junior mark and the replacement portion of the control. See Gov’t Employees Ins. Co. v. Google, Inc., No. 1:04CV507, 2005 WL 1903128, at *5 (E.D.Va. Aug. 8, 2005) (criticizing control for not removing allegedly infringing elements for which GEICO wanted to measure confusion while keeping the other elements as constant as possible); 6 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 32:187 (“The control question or questions should use a mark similar enough to the actual mark that it provides an accurate measure of the confusion created by the accused mark, not by some other similarity.”); Federal Judicial Center, Reference Manual on Scientific Evidence 258 (2d ed. 2000) (“In designing a control group study, the expert should select a stimulus for the control group that shares as many characteristics with the experimental stimulus as possible, with the key exception of the characteristic whose influence is being assessed.”). Rather than disagreeing with the appropriateness of this comparison, Defendants’ expert Dr. Simonson testified that a control should be “as similar to the junior mark, in this case, CRISTALINO, as possible, without infringing on the senior mark.”
51.With the exception of the term “CRISTAL” — the characteristic whose influence was being .assessed — the names “COURTALINO” and “CRISTALINO” are very similar. Both words end in “INO,” have ten letters, and start with a “C.” In addition, “COURTALINO” shares a “TALINO” with “CRISTALINO” and shares an “R” with the “CRISTAL” portion of “CRISTALINO.” “COURTALINO” and “CRISTALINO” are very similar, and Defendants identify no alternative name that meets the criteria for a control.12 Consequently, the Court concludes *876that Dr. Kaplan’s selection of “COURTALINO” as a control was appropriate.
52. Defendants also contend that the sequential survey methodology employed by the Kaplan survey created demand effects because it did not replicate market conditions. The Kaplan survey methodology consisted of showing potential respondents a bottle of CRISTAL champagne and, if the potential respondent was previously aware of CRISTAL champagne, removing the CRISTAL champagne bottle, creating a line-up of the other bottles, and asking a series of questions intended to determine whether confusion existed. This format has been used in cases where the products at issue are displayed in close proximity and the products directly compete. See, e.g., Storck USA, L.P. v. Farley Candy Co., 797 F.Supp. 1399, 1408-09 (N.D.Ill.1992). Because CRISTALINO cava and CRISTAL champagne do not directly compete, the sequential survey methodology did not fully reflect marketplace conditions. However, because CRISTAL champagne and CRISTALINO cava are both sparkling wines located in the same area of liquor stores and listed in the same category in brochures and wine lists, the Court concludes that the sequential survey methodology was not entirely inappropriate. Cf. Kargo Global, Inc. v. Advance Magazine Publishers, Inc., No. 06 Civ. 550, 2007 WL 2258688, at *5-7 (S.D.N.Y. Aug. 6, 2007) (finding sequential survey methodology inappropriate where no evidence indicated that prospective consumers were likely to encounter the KAR-GO trademark for wireless services a short time after seeing CARGO magazine).
53. Defendants contend that the demand effects are evidenced by the fact that almost 30% of the respondents said that WILLM sparkling wine was related to CRISTAL champagne because there was no reason for this level of confusion. See Kargo Global, 2007 WL 2258688, at *9-10 (finding that survey created demand effects where 80% of respondents in the control group believed there was a connection between the companies that produced the control product CARRY magazine and the completely unrelated KARGO wireless services). As acknowledged by Dr. Simon-son, however, the WILLM bottle is not properly considered a control in the Kaplan survey and the respondents selecting the WILLM bottle identified the clear col- or of the WILLM and CRISTAL bottles as the reason for their selection. In addition, the WILLM bottle has labels that are primarily gold and a medallion on the neck label in the same location as the LR medallion on the CRISTAL bottle. Consequently, the respondents’ belief that the WILLM bottle was related to the CRIS-TAL bottle was not unreasonable and does not indicate that the respondents were, as Dr. Simonson suggested, “looking for anything” upon which to base a relationship.
54.In addition, almost half of the respondents did not identify any bottle as related to or associated with the producer of CRISTAL champagne. Of those respondents identifying the CRISTALINO bottle based on the name, 57% of respondents selected only the CRISTALINO bottle. Eighty percent selected the CRIS-TALINO bottle as their first choice, while the CRISTALINO bottle was never the third or fourth bottle selected. Those statistics indicate that demand effects alone do not account for the level of confusion *877measured with respect to CRISTALINO cava and CRISTAL champagne.
55. In summary, the Court assigns diminished weight to the Kaplan survey based on the flaws in the survey universe and methodology. “Sometimes, [however,] the most illuminating and probative parts of a survey are not the numbers and percentages generated by the responses, but the verbatim accounts of the responses. The respondents’ verbatim responses to [a] ‘why’ question may provide a window into consumer thought processes in a way that mere statistical data cannot.” 6 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 32:178; see also AM Gen. Corp. v. DaimlerChrysler Corp., 311 F.3d 796, 826 (7th Cir.2002) (relying on verbatim responses when concluding survey did not support a finding of likelihood of confusion). The verbatim responses to the Kaplan survey indicate that a significant percentage of consumers aware of CRISTAL champagne were confused as to brand, source, sponsorship, or permission with respect to CRISTALINO cava and CRISTAL champagne based on the products’ names and labeling. Therefore, despite its flaws, the Court concludes that the Kaplan survey slightly favors a finding of likelihood of confusion.
6. Degree of care
56. The likelihood of confusion determination also takes into account the degree of care exercised by the purchasers, which requires consideration of the type of product, its cost, and conditions of purchase. SquirtCo, 628 F.2d at 1091. The Court must “stand in the shoes of the ordinary purchaser, buying under the normally prevalent conditions of the market and giving the attention such purchasers usually give in buying that class of goods.” Luigino’s, 170 F.3d at 832 (quotation marks omitted).
57. In general, purchasers of wine and sparkling wine are unsophisticated and rely on familiarity with brands and the information conveyed by the labels when purchasing a less-expensive product. To the extent distinguishing characteristics exist between bottles of CRISTALINO cava and CRISTAL champagne, such as the use of French or Spanish on the label and the difference in price, those characteristics are unlikely to prevent confusion as to association or sponsorship given the similarity of the products’ names and the commonplace practice of marketing sparkling wines at a variety of price points by premier champagne houses, including Roederer. See Palm Bay Imports, 396 F.3d at 1376 (“And even more sophisticated purchasers might be aware that champagne houses offer both types of products under similar marks, and could easily conclude that VEUVE ROYALE was Veuve Clicquot’s sparkling wine.”); see also Daddy’s Junky Music Stores, 109 F.3d at 286 (“[C]onfusingly similar marks may lead a purchaser who is extremely careful and knowledgeable about the instrument that he is buying to assume nonetheless that the seller is affiliated with or identical to the other party.”). Consequently, the degree of care factor favors a finding of likelihood of confusion.
7. Conclusion as to likelihood of confusion
58. In summary, the CRISTAL marks are strong, the names “CRISTAL” and “CRISTALINO” are similar, and the addition of “INO” to “CRISTAL” to form “CRISTALINO” suggests a connection between the two products. CRISTAL champagne and CRISTALINO cava have similarly-colored labels on which their names are displayed in similar fonts. The CRIS-TALINO bottle does not prominently display “JAUME SERRA,” a house mark, or any other distinguishing characteristic. CRISTALINO cava and CRISTAL champagne are in competitive proximity, and consumers exercise a low degree of care when purchasing CRISTALINO cava. All *878of these factors weigh in favor of a likelihood of confusion that, because it is reasonable for consumers to believe that the source of CRISTAL champagne would expand into a lower-priced sparkling wine category, is not offset by the differences in price, packaging, and country of origin.
59.The United States Court of Appeals for the Ninth Circuit explained this type of confusion as follows:
[TJhere may not be one in a hundred buyers of this whisky who knows that it is made [and wholesaled by the plaintiffs]. Probably all that such buyers know is that Black & White Scotch whisky has satisfied them in the past or that they have heard of it and the average purchaser would no doubt select for the use of his guests something with which he was familiar and thus purchase Black & White whisky.... It is our view, and we so hold, that the average purchaser, as the courts have described him, would be likely to believe, as he noted the Black & White beer in [defendant’s] stores, that the maker of the beer had some connection with the concern which had produced the well known Black & White Scotch whisky.
Fleischmann Distilling Corp. v. Maier Brewing Co., 314 F.2d 149, 155 (9th Cir. 1963). Similarly, while purchasers and potential purchasers of CRISTALINO cava may not know that Roederer produces CRISTAL champagne, the Court’s analysis of the SquirtCo factors indicates that an appreciable number of those purchasers are likely to be confused as to whether the producer of CRISTAL champagne is associated with, sponsors, or approves of CRISTALINO cava.13 Roederer has proved its claims of trademark infringement and unfair competition under the Lanham Act and Minnesota law.
III. Trademark dilution
60. Roederer asserts a claim of trademark dilution pursuant to 15 U.S.C. § 1125(c), which provides that “the owner of a famous mark that is distinctive ... shall be entitled to an injunction against another person who, at any time after the owner’s mark has become famous, commences use of a mark or trade name in commerce that is likely to cause dilution.” 14 Accordingly, to succeed on a claim of trademark dilution, Roederer must prove the CRISTAL marks were famous before Jaume Serra began its use in commerce of the CRISTALINO mark.
61. Roederer maintains that the relevant date for the determination of fame is 1993 because sales of CRISTALINO cava in the United States before that year were “de minimus.” Roederer also argues that, although Jaume Serra’s first sales in the United States of cava bearing the CRIS-TALINO name began in 1989, the alleged diluting use did not begin until Jaume Serra removed “JAUME SERRA” from the labels of CRISTALINO cava in 1993. The caselaw does not support a “de mini-mus ” exception to federal anti-dilution law.15 See Enter. Rent-A-Car Co. v. Ad*879vantage Rent-A-Car, Inc., 830 F.3d 1333, 1342 (Fed.Cir.2003) (“[I]t is significant that there is no qualification in the statute requiring that the defendant’s prior use be substantial or cover a wide geographic area to defeat an injunction under the statute.”). In addition, “any commercial use of a famous mark in commerce is arguably a diluting use that fixes the time by which famousness is to be measured.” Nissan Motor Co. v. Nissan Computer Corp., 378 F.3d 1002, 1013 (9th Cir.2004). Nevertheless, the Court need not decide whether 1989 or 1993 is the date by which fame should be measured because, for the reasons set forth below, the Court concludes that the CRISTAL marks were not famous within the meaning of § 1125(c) in either year.
62. Fame for dilution purposes and fame for likelihood of confusion purposes are distinct concepts in that “dilution fame is an either/or proposition” while likelihood of confusion fame varies from very strong to very weak. Palm Bay Imports, 396 F.3d at 1374-75. A mark is “famous” for dilution purposes if the mark “is widely recognized by the general consuming public of the United States as a designation of source of the goods or services of the mark’s owner.” 15 U.S.C. § 1125(c)(2)(A). A mark that is famous only within a niche market does not qualify as “famous” within the meaning of § 1125(c). In determining whether a mark is famous, a court may consider all relevant factors, including (1) the duration, extent, and geographic reach of advertising and publicity of the mark, whether advertised or publicized by the owner or third parties, (2) the amount, volume, and geographic extent of sales of goods offered under the mark, (3) the extent of actual recognition of the mark, and (4) whether the mark was registered under the Act of March 3, 1881, or the Act of February 20, 1905, or on the Principal Register. Id.
63. With respect to the duration, extent, and geographic reach of advertising and publicity of the CRISTAL marks, the advertising of CRISTAL champagne before 1993 was minimal. Although CRISTAL champagne benefited from a certain level of unsolicited publicity, most of the media mentions were in special-interest publications directed to the wine industry or related industries. Similarly, most references to CRISTAL champagne in general-interest publications were peripheral to the focus of the article while depictions of CRISTAL champagne in general-interest publications showed the bottle in conjunction with other luxury goods or only a portion of the bottle. “Many products receive broad incidental media coverage. Such promotion does not lead to the conclusion that their trademarks have become a part of the collective national consciousness.” Thane Int’l Inc. v. Trek Bicycle Corp., 305 F.3d 894, 912 (9th Cir. 2002). While CRISTAL champagne had celebrity “ambassadors” prior to 1993, no evidence indicates how often those ambassadors promoted CRISTAL champagne or how many people such promotions reached. The advertising and publicity of the CRISTAL marks prior to 1993 does not support a finding of fame because it primarily reached those in the wine industry or related industries.
64. The second relevant factor is the amount, volume, and extent of sales of CRISTAL champagne. 15 U.S.C. § 1125(c)(2)(A)(ii). Roederer sold * * * bottles of CRISTAL champagne in 1989 and * * * bottles in 1993. This sales volume is small compared to that of a mass-merchandise product. However, the fact that demand for CRISTAL champagne ex*880ceeded its supply contributed to its status as a prestige champagne. See 4 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 24:106 (“[I]t should be noted that a lack of a large scale of sales need not always signal a lack of fame.”). This factor is neutral with respect to fame.
65. As to the third factor, recognition of the mark, although CRISTAL champagne was recognized as a high-quality product by those in the wine industry and related industries, the evidence does not support a finding of recognition of the CRISTAL mark by the general public pri- or to 1993. Roederer offered no evidence of any explicit recognition of the CRISTAL brand’s power in the years preceding 1993, and CRISTAL champagne was seldom the focus of an article or picture in a general-interest publication. In addition, there is no evidence of the effect CRISTAL champagne’s celebrity ambassadors had on the general public. This factor weighs against a finding of fame.
66. Finally, the registration of the '343 and '998 marks on the Principal Register weighs in favor of a finding of fame. However, “[o]ne cannot logically infer fame from the fact that a mark is one of the millions on the federal Register.” 4 J. Thomas McCarthy, McCarthy on Trademarks & Unfair Competition § 24:106.
67. The Court’s consideration of the relevant factors indicates that the fame of the CRISTAL marks before 1993 was a “niche fame.” Because Roederer has not proven that the CRISTAL marks were famous within the meaning of § 1125(c) before 1993, Roederer’s trademark dilution claim fails.
IV. Attorney fees
68. Roederer seeks attorney fees under 15 U.S.C. § 1117(a), which provides that a court may award attorney fees to the prevailing party in a trademark infringement case if the case is exceptional. A case is exceptional if the infringer’s actions were malicious, fraudulent, deliberate, or willful. Venture Tape Corp. v. McGills Glass Warehouse, 540 F.3d 56, 64 (1st Cir.2008). Roederer’s argument for attorney fees is based on the same acts it claims demonstrate Defendants’ intent to infringe.
69. The Court previously concluded that Defendants had no intent to trade off the CRISTAL marks. Nothing indicates that Defendants’ defense against Roederer’s claims of trademark infringement, unfair competition, and trademark dilution was not in good faith, and Roederer has presented no other evidence indicating that that Defendants’ conduct was malicious, fraudulent, deliberate, or willful. In such a case, attorney fees are not warranted. See Lindy Pen Co. v. Bic Pen Corp., 982 F.2d 1400, 1409-10 (9th Cir. 1993) (affirming district court’s denial of attorney fees where infringement not intentional and nothing else made case exceptional). The Court denies Roederer’s request for attorney fees under 15 U.S.C. § 1117(a).
V. Injunctive Relief
70. Roederer seeks a permanent injunction pursuant to 15 U.S.C. § 1116(a) and Minn. Stat. § 325D.45(1) barring Defendants from using the name CRISTALINO or any other mark, word, or name similar to the trademarks of Roederer that is likely to cause confusion, mistake, or deception. The Court may issue a permanent injunction if Roederer proves (1) that it has suffered an irreparable injury; (2) that remedies available at law, such as monetary damages, are inadequate to compensate for the injury; (3) that, considering the balance of hardships between the plaintiff and defendant, an equitable remedy is warranted; and (4) that the public *881interest would not be disserved by a permanent injunction. eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 391, 126 S.Ct. 1837, 164 L.Ed.2d 641 (2006).
71. Roederer’s CRISTAL marks “represente ] intangible assets such as reputation and goodwill.” Gen. Mills, 824 F.2d at 625. The reputation of CRISTAL champagne as a high-quality product is unquestionable, and Roederer has developed and protected that reputation and the goodwill associated with it since at least the 1950’s. Roederer’s inability to control the quality of CRISTALINO cava constitutes irreparable injury regardless of its quality. See Opticians Ass’n of Am. v. Indep. Opticians of Am., 920 F.2d 187, 195-96 (3d Cir.1990) (“If another uses [the plaintiffs mark], he borrows the owner’s reputation, whose quality no longer lies within his own control. This is an injury, even though the borrower does not tarnish it, or divert any sales by its use.”); Gen. Mills, 824 F.2d at 625. Monetary damages are inadequate to compensate Roederer for the harm to its goodwill and reputation. See Medicine Shoppe Int'l, Inc. v. S.B.S. Pill Dr., Inc., 336 F.3d 801, 805 (8th Cir.2003) (“Harm to reputation and goodwill is difficult, if not impossible, to quantify in terms of dollars.”). The first two requirements of eBay support entry of a permanent injunction.
72. Defendants will be harmed by entry of a permanent injunction because they have acquired significant goodwill associated with the CRISTALINO brand. However, Defendants had a duty to select a trademark that would avoid confusion. See Wesley-Jessen Div. of Schering Corp. v. Bausch & Lomb Inc., 698 F.2d 862, 867 (7th Cir.1983). Despite this duty, Jaume Serra and Defendants conducted no investigation into whether their use of CRIS-TALINO on cava could infringe another’s trademark rights, and had notice of Roederer’s objections to their use of the CRISTALINO mark in the 1990’s. See Champagne Louis Roederer, 569 F.3d at 860-61. Accordingly, Defendants’ injury is discounted because, to some extent, they brought the injury upon themselves. See Novartis Consumer Health, Inc. v. Johnson & Johnson-Merck Consumer Pharm. Co., 290 F.3d 578, 596 (3d Cir.2002). The Court concludes that the balance of the harms supports entry of a permanent injunction.
73.Finally, the public interest would not be disserved by entry of an injunction. Rather, “the public interest is served by preventing consumer confusion in the marketplace.” Davidoff & CIE, S.A. v. PLD Int’l Corp., 263 F.3d 1297, 1304 (11th Cir. 2001). Based on its analysis of the eBay factors, the Court finds that a permanent injunction is appropriate.
74. A permanent injunction in a trademark case should be narrowly tailored to fit specific legal violations and should not impose unnecessary burdens on lawful activity. Patsy’s Brand, Inc. v. I.O.B. Realty, Inc., 317 F.3d 209, 220 (2d Cir.2003); see also Jim Beam Brands Co. v. Beamish & Crawford Ltd., 937 F.2d 729, 737 (2d Cir.1991) (explaining that an injunction in a trademark infringement case must be appropriately tailored to avoid confusion in the marketplace). An injunction mandating use of a disclaimer or house mark may be an appropriate remedy. See A & H Sportswear Inc. v. Victoria’s Secret Stores, Inc., 166 F.3d 197, 208 (3d Cir.1999). In addition, an injunction specifying the use of certain fonts and font sizes and the prominence of certain phrases may be appropriate. See, e.g., Patsy’s Brand, 317 F.3d at 220-21 (modifying injunction to permit defendants to use phrase in no greater than 10-point type, in font dissimilar to the plaintiffs font, and only as a “minor component of the labeling”); Scarves by Vera, Inc. v. Todo Im*882ports Ltd., 544 F.2d 1167, 1175 (2d Cir. 1976) (permitting defendant to use its tradename “in small type, but only in conjunction with other words which prevent any likelihood of confusion”); Am. Ass’n for Advancement of Sci v. Hearst Corp., 498 F.Supp. 244, 264-65 (D.D.C.1980) (enjoining defendant from publishing its magazine Science Digest “with undue prominence given to the word ‘Science’ or with the word ‘Digest’ occupying less than seventy-five (75) percent of the area occupied by the word ‘Science’ on the cover of the magazine or anywhere else that the title is presented”).
75. Here, the total effect conveyed by the marks is confusingly similar because “CRISTAL” is the first and dominant portion of the CRISTALINO mark, the marks are printed in similar Roman serif fonts and used on similarly-colored labels on bottles of sparkling wine, and there is no prominent reference to the Jaume Serra winery (or any other distinguishing feature) on the labels of the CRISTALINO bottles. An injunction completely barring use of the word “CRISTALINO” is unnecessary to prevent consumer confusion. Rather, Defendants may use the term “CRISTALINO” in connection with the manufacture, importation, distribution, advertisement, promotion, sale, or offering for sale of cava or sparkling wine only as part of the phrase “JAUME SERRA CRISTALINO” and only in conjunction with a prominent disclaimer stating that “JAUME SERRA CRISTALINO is not affiliated with, sponsored by, approved by, endorsed by, or in any way connected to Louis Roederer’s CRISTAL® champagne or Louis Roederer.” The phrase “JAUME SERRA CRISTALINO” shall only be displayed in a font dissimilar to the Roman serif font used for the word '“CRISTAL” on bottles of CRISTAL champagne and only when the words “JAUME SERRA” are displayed in close proximity to the word “CRISTALINO,” in the same font as the word “CRISTALINO,” and in a font size at least as large as the font size of the word “CRISTALINO.” Moreover, when used on bottles of cava, the phrase “JAUME SERRA CRISTALINO” shall only be used on labels having a background color other than gold if the cava is a brut, dry, extra-dry, or semi-dry and other than a pink-hued copper if the cava is a rosé.16
76. The same restrictions apply to Defendants’ use of any mark, word, or name similar to the CRISTALINO name that is likely to cause confusion, mistake, or deception with Roederer’s CRISTAL marks in connection with the manufacture, importation, distribution, advertisement, promotion, sale, or offering for sale of cava or sparMing wine. Although Defendants may continue to sell their existing inventory of already-labeled bottles of CRISTALINO cava, they must destroy all labels, signs, prints, packages, wrappers, receptacles, and advertisements in their possession bearing the mark CRISTALINO, and all plates, molds, matrices, and other means for making the same, that do not conform with these requirements.
CONCLUSION
Based on the submissions of the parties, the arguments of counsel, and the entire file and proceedings herein, the Court *883finds in favor of Roederer on its trademark infringement and unfair competition claims (Counts I-IV) and against Roederer on its trademark dilution claim (Count V). The Court issued a separate Order and Injunction [Docket No. 328] consistent with these Non-Confidential Findings of Fact and Non-Confidential Conclusions of Law on July 27, 2010.