FCOA LLC v. Foremost Title & Escrow Services LLC

57 F.4th 939

• Court: Court of Appeals for the Eleventh Circuit
• Decided: January 12, 2023
• Docket: 19-13390
• Status: Published

Opinion

USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 1 of 40

[PUBLISH] In the United States Court of Appeals For the Eleventh Circuit

____________________

No. 19-13390 ____________________

FCOA LLC, Plaintiff-Appellant, versus FOREMOST TITLE & ESCROW SERVICES LLC,

Defendant-Appellee.

Appeal from the United States District Court for the Southern District of Florida D.C. Docket No. 1:17-cv-23971-KMW ____________________ USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 2 of 40

2 Opinion of the Court 19-13390

Before BRANCH, GRANT, and TJOFLAT, Circuit Judges. TJOFLAT, Circuit Judge: In this trademark infringement case, we must decide whether the parties’ FOREMOST trademarks at issue could con- fuse consumers into thinking that a relationship exists between the parties. Here, the District Court found at summary judgment that there was no likelihood of confusion (and thus no trademark in- fringement) between the FOREMOST marks of Foremost Insur- ance Company (“FIC”), a multi-billion dollar insurance company which for 70 years has sold many different lines of insurance, and Foremost Title and Escrow (“FT&E”), a shell company set up to sell title insurance for a law firm. After reviewing the record and with the benefit of oral argument, we disagree with the District Court’s likelihood of confusion analysis and thus reverse the Court’s grant of summary judgment on FIC’s trademark infringe- ment claim. 1 I. In 1952, FIC was founded and started using FOREMOST- branded marks to market and sell its insurance products. After FIC operated independently for several decades, Farmers Insurance Group acquired FIC in 2000. Now a subsidiary of Farmers, FIC

1 FT&E also appealed the District Court’s denial of attorney’s fees and non- taxable costs to FT&E in a separate cross-appeal that is not before us. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 3 of 40

19-13390 Opinion of the Court 3

continues to sell insurance in the United States and Florida under its FOREMOST-branded marks. 2 In total, FIC owns 21 registered trademarks with the word “Foremost.” On its website, FIC displays a FOREMOST mark in the following way:

FIC also uses its FOREMOST marks on its online advertisements, social media, emails, magazines, and brochures. From 2011 to 2017, FIC spent an average of $6,765,627 per year to advertise and promote its FOREMOST marks. Moreover, the American Associ- ation of Retired Persons (“AARP”) endorsed FIC in 1989. Thus, FIC also advertises to AARP members using its FOREMOST marks through AARP’s website, email and mailing lists, and the AARP magazine. AARP has 2.7 million Florida members; of these, FIC sent FOREMOST-branded emails or mail solicitations to over 120,000 AARP members in 2016 and 2017. Additionally, FIC’s

2 One of FIC’s wholly owned subsidiaries, FCOA, owns legal title to FIC’s FOREMOST trademarks. Although FCOA filed this lawsuit, the real party in interest is FIC and we refer to these entities collectively as FIC throughout this opinion. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 4 of 40

4 Opinion of the Court 19-13390

independent insurance agents often use the FOREMOST marks in their own marketing. FIC has issued over 3 million FOREMOST-branded insur- ance policies nationwide, including homeowners’ insurance, prop- erty insurance, fire insurance, business building insurance, landlord insurance, and mobile home insurance. In Florida alone, FIC has over 95,000 customers. FIC primarily sells its insurance policies through its over 33,000 independent agents at 77,000 locations. FIC generated over $2 billion in insurance premiums nationwide in 2017, of which $80 million came from Florida policies. How- ever, one thing FIC does not offer is title insurance. Under Florida law, entities that issue title insurance are prohibited from selling any other type of insurance, and vice versa. Fla. Stat. § 627.786. 3 Enter FT&E. In 2015, two partners of the law firm Stok Folk + Kon, Robert Stok and Joshua Kon, set up FT&E as a Florida- based limited liability company. Stok and Kon created FT&E to do one thing: take over the real estate closings and title insurance4 sales that Stok Folk + Kon previously performed. FT&E shares

3We need not reach the issue of whether FIC could have created a subsidiary to sell title insurance under a FOREMOST mark if it had so desired. 4 Titleinsurance protects home purchasers and lenders from the risks associ- ated with defects in title, such as issues not discoverable in a title search and mistakes made in a title search. Fla. Stat. § 624.608 (defining title insurance as “[i]nsurance of owners of real property or others having an interest in real property . . . against loss by encumbrance, or defective titles, or invalidity, or adverse claim to title”). USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 5 of 40

19-13390 Opinion of the Court 5

with Stok Folk + Kon both a physical address in Aventura, Florida (a suburb of Miami), and a phone number. In preparing to open FT&E, Stok and Kon conducted a search for potential business names. As part of this process, they brainstormed the term “foremost,” searched the term on the Flor- ida Secretary of State’s online business list, and found no other ac- tive title insurance businesses with “foremost” in their name. So, Stok and Kon settled on the name “Foremost Title & Escrow.” FT&E adopted the following mark, which it displays on its website:

FT&E derives its clients from the Stok Folk + Kon law firm and realtor referrals. Still, FT&E markets its title insurance and closing services through online advertisements, social media, a locally dis- tributed magazine, trade shows, public events, and emails to home- owners. FT&E received its license to operate as a title insurance agency in Florida on October 13, 2015, from the Florida USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 6 of 40

6 Opinion of the Court 19-13390

Department of Financial Services (“DFS”).5 In May 2016, FT&E began conducting closings within the Tri-County area of South Florida (Miami-Dade, Broward, and Palm Beach counties) and marketing to real estate agents, bankers, mortgage brokers, and de- velopers. FT&E had completed seven closings by October 2017 (when FIC filed this lawsuit) and at least 20 closings by November 2018 (when the parties moved for summary judgment). Obtaining title insurance is an integral part of FT&E’s closing services. How- ever, FT&E does not underwrite title insurance itself. Instead, FT&E conducts a title search as a title insurance agent for Fidelity National Title Insurance Company and Old Republic Title Insur- ance Company based on agency agreements executed in May 2016. Fidelity National and Old Republic then decide whether to issue a policy insuring the purchaser’s title to real estate at closing. FT&E

5 Under Florida law, title insurance may only be sold by a “licensed and ap- pointed title insurance agent employed by a licensed and appointed title insur- ance agency.” Fla. Stat. § 626.8412(1)(a). A title insurance agency is different than a title insurer, who underwrites and issues a policy insuring title. The agent acts on the insurer’s behalf in selling the policy and running a “reasona- ble title search.” Id. §§ 627.7845, 627.796. Typically, title insurance agents must either be attorneys licensed by the state of Florida or pass a licensing exam. Id. §§ 626.8417(4) (noting that attorneys are exempt from the licensing and appointment requirements of title insurance agents), 626.241(7). Like- wise, title insurance agencies (which employ title insurance agents) must be licensed by the state and appointed as an agent by a title insurer. Id. §§ 626.8418, 626.8417(6) (providing that a title insurance agency owned by lawyers and not engaged in the practice of law must still comply with licensing and appointment requirements). USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 7 of 40

19-13390 Opinion of the Court 7

earns revenue by charging a fee for closing services and collecting a portion of the insurance premium for the policies it procures. Seven months after FT&E started conducting closings and obtaining title insurance, FIC sent FT&E a cease-and-desist letter claiming that FT&E’s use of the term “foremost” infringed on FIC’s FOREMOST marks. FT&E insists that this letter was the first time it learned of FIC and its numerous lines of insurance in Flor- ida. FT&E responded to FIC’s letter by disputing the allegations of trademark infringement in both a phone call and a letter to FIC. On October 4, 2017, FIC filed a five-count complaint against FT&E. FIC alleged trademark infringement under the Lanham Act, 15 U.S.C. §§ 1114 and 1116 (Count I); false designation of origin, a form of unfair competition under the Lanham Act, 15 U.S.C. § 1125(a) (Count II); dilution under the Lanham Act, 15 U.S.C. § 1125(c) (Count III); unfair competition under Florida com- mon law (Count IV); and antidilution under Fla. Stat. § 495.151 (Count V). Following discovery, both parties moved for summary judgment on November 2, 2018. In August 2019, the District Court issued its order denying FIC’s motion for summary judgment and granting FT&E’s motion for summary judgment. FCOA, LLC v. Foremost Title & Escrow Servs., LLC, 416 F. Supp. 3d 1381, 1395 (S.D. Fla. 2019). Because the District Court believed that each count of FIC’s complaint re- quired a showing of likelihood of confusion between FT&E’s and USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 8 of 40

8 Opinion of the Court 19-13390

FIC’s marks,6 it began (and ended) its analysis there. Id. at 1387– 88. The District Court held that the two marks did not create a likelihood of confusion by consumers that a relationship exists be- tween the parties as a matter of law. Id. at 1394. Accordingly, the Court ruled that FIC had no cognizable claim and granted FT&E’s motion for summary judgment. Id. at 1394–95. FIC then timely appealed the Court’s final judgment as to Count I, its trademark infringement claim.7

6 This belief was only partially correct. The District Court correctly recog- nized that FIC’s trademark infringement and unfair competition claims (Counts I, II, and IV) do all require a showing of likelihood of confusion. FCOA, 416 F. Supp. 3d at 1388 & n.4; Fla. Int’l Univ. Bd. of Trs. v. Fla. Nat’l Univ., Inc. (“FIU ”), 830 F.3d 1242, 1265 (11th Cir. 2016) (citing Suntree Techs., Inc. v. Ecosense Int’l, Inc., 693 F.3d 1338, 1346 (11th Cir. 2012)); Custom Mfg. & Eng’g, Inc. v. Midway Servs., Inc., 508 F.3d 641, 652–53 (11th Cir. 2007) (holding that Florida unfair competition and trademark infringement use the same likelihood of confusion analysis as the federal Lanham Act test). How- ever, dilution under federal and Florida law (Counts III and V) do not. Mose- ley v. V Secret Catalogue, Inc., 537 U.S. 418, 429, 123 S. Ct. 1115, 1122 (2003); 15 U.S.C. § 1125(c)(1); Great S. Bank v. First S. Bank, 625 So. 2d 463, 470 (Fla. 1993). As far as we can tell, the District Court never mentioned FIC’s dilution claims in its order at all despite purporting to decide FIC’s “five claims.” See FCOA, 416 F. Supp. 3d at 1386. Since FIC only appealed its trademark in- fringement claim, we need not address this matter further. 7FIC failed to mention the unfair competition and dilution claims (Counts II through V) in its initial appellate brief beyond noting that its complaint alleged these claims in the facts section of FIC’s brief. Appellant Br. at 3. By failing to provide any argument or citation that the District Court erred in deciding these claims in its initial appellate brief, FIC forfeited them. United States v. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 9 of 40

19-13390 Opinion of the Court 9

II. On summary judgment, we review district court decisions de novo using the same standard as the district courts. Tana v. Dantanna’s, 611 F.3d 767, 772 (11th Cir. 2010). We view all the evidence and draw all reasonable inferences in favor of the non- moving party. Id. A grant of summary judgment is proper where there is “no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). 8 We have recognized that district courts deciding summary judgment motions may occasionally draw inferences against the

Campbell, 26 F.4th 860, 873 (11th Cir. 2022) (en banc) (holding that the “failure to raise an issue in an initial brief on direct appeal” results in forfeiture). 8 FT&E urges us to treat the proceedings below as a bench trial because the District Court implicitly (and wrongly) decided questions of fact as a matter of law. In other words, FT&E seeks to shift our standard of review from the more exacting summary judgment standard—whether any issues of material fact remain—to the more deferential standard for factual determinations in a bench trial—whether the District Court’s determinations were clearly errone- ous. We refuse to do so. Cross-motions for summary judgment may be treated on appeal as a bench trial only in rare “limited circumstances,” consid- ering whether (1) there was a hearing on the merits of the motions where the facts were fully developed; (2) the parties “expressly stipulated to an agreed set of facts”; and (3) the record shows that the parties “in effect submitted the case to the court for trial on an agreed statement of facts embodied in a limited written record, which would have enabled the [district] court to decide all is- sues and resolve all factual disputes.” FIU, 830 F.3d at 1252–53 (alterations in original); see also Ga. State Conf. of NAACP v. Fayette Cnty. Bd. of Comm’rs, 775 F.3d 1336, 1345–46 (11th Cir. 2015). None of these circumstances are pre- sent here. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 10 of 40

10 Opinion of the Court 19-13390

non-movant when “there are no genuine issues of material fact,” “no issues of witness credibility,” and the district court must decide the motion based on a cold record consisting of “affidavits, deposi- tions, and stipulations.” Useden v. Acker, 947 F.2d 1563, 1572 (11th Cir. 1991) (quoting Nunez v. Superior Oil Co., 572 F.2d 1119, 1123– 24 (5th Cir. 1978)). However, our standard of review on appeal is “unaffected by any inferential conclusions reached below” under the Nunez standard of review. Id. at 1573 & n.14. III. Trademark infringement under the Lanham Act occurs when a defendant, without consent, uses “in commerce any repro- duction, counterfeit, copy, or colorable imitation of a registered mark” that “is likely to cause confusion” that a relationship exists between the parties. 15 U.S.C. § 1114(1). 9 For a trademark in- fringement claim, a plaintiff must demonstrate (1) that it owns a valid mark with priority, and (2) that the defendant’s mark is likely to cause consumer confusion with the plaintiff’s mark. See Frehling Enters., Inc. v. Int’l Select Grp., Inc., 192 F.3d 1330, 1335 (11th Cir. 1999). The parties agree that at least some of FIC’s

9That relationship can take two forms. First, a consumer could be confused about the source of the marks, thinking that the goods or services associated with a second mark are produced by the original mark holder. Second, a con- sumer may be confused as to the existence of an affiliation, connection, or sponsorship between the parties. Univ. of Ga. Athletic Ass’n v. Laite, 756 F.2d 1535, 1546–47 (11th Cir. 1985) (citing Bos. Pro. Hockey Ass’n v. Dallas Cap & Emblem Mfg., Inc., 510 F.2d 1004, 1012–13 (5th Cir. 1975)). USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 11 of 40

19-13390 Opinion of the Court 11

FOREMOST marks have priority over FT&E’s. So, because the District Court granted summary judgment to FT&E, the sole is- sue 10 before us is whether a reasonable jury could find that FT&E’s FOREMOST mark is likely to cause confusion with FIC’s marks.11 The likelihood of confusion analysis involves two steps. At step one, the court considers several factors which can provide cir- cumstantial evidence of likelihood of confusion. See Fla. Int’l Univ. Bd. of Trs. v. Fla. Nat’l Univ., Inc. (“FIU ”), 830 F.3d 1242, 1255 (11th Cir. 2016). Or, to put it another way, the court conducts sev- eral separate inquiries on the factors which yield “circumstantial facts” that shed light on the likelihood of confusion as a whole.12

10 The number of FIC’s marks that have priority over FT&E’s mark is insignif- icant to our analysis because only one mark needs priority to reverse the grant of summary judgment. 11 FT&E also provided our panel with its application to register its mark, FOREMOST TITLE & ESCROW, with the Patent and Trademark Office (“PTO”), and with the record of the PTO’s subsequent actions. However, the PTO’s ex parte decision is not entitled to even persuasive weight in the likeli- hood of confusion analysis, so it is irrelevant to our decision. PlayNation Play Sys., Inc. v. Velex Corp., 924 F.3d 1159, 1169 (11th Cir. 2019). 12This Court has usually discussed the likelihood of confusion test in terms of evaluating factors, not as separate inquiries yielding “circumstantial facts.” See, e.g., FIU, 830 F.3d at 1255; Tana, 611 F.3d at 774–75; Frehling, 192 F.3d at 1335. In this opinion, we discuss the multifactor likelihood of confusion test in these terms to reinforce the idea that each of these factors is analytically a separate factual inquiry relevant to, but ultimately independent of, likelihood of confusion. Therefore, our use of the terms separate inquiries and “circum- stantial facts” throughout this opinion is meant as a stylistic choice only and does not substantively change our caselaw. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 12 of 40

12 Opinion of the Court 19-13390

Of course, on a motion for summary judgment, these separate in- quiries must view all the evidence and draw all reasonable infer- ences in favor of the non-moving party. This Court has recognized seven factors as relevant: (1) the strength of the allegedly infringed mark; (2) the similarity of the infringed and infringing marks; (3) the similarity of the goods and services the marks represent; (4) the similarity of the parties’ trade chan- nels and customers; (5) the similarity of advertising media used by the parties; (6) the intent of the alleged infringer to misappropriate the proprietor’s good will; and (7) the existence and extent of actual confu- sion in the consuming public.

Id. Additionally, this Court has also analyzed consumer sophistica- tion as a separate factor or circumstantial fact relevant to determin- ing likelihood of confusion, see id. at 1256, and we analyze it as such infra Part III.A.viii. At step two, the court weighs each of the relevant circum- stantial facts—independently and then together—to determine whether the ultimate fact, likelihood of confusion, can reasonably be inferred. See Frehling, 192 F.3d at 1335. This inference is also a factual inquiry. Id. In drawing the ultimate inference about likeli- hood of confusion, the two most important circumstantial facts are respectively actual confusion and the strength of the mark. FIU, 830 F.3d at 1255. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 13 of 40

19-13390 Opinion of the Court 13

This two-step analysis is the same whether the court is de- ciding likelihood of confusion at a bench trial or entertaining a mo- tion for summary judgment. See Frehling, 192 F.3d at 1335; Tana, 611 F.3d at 774–82. On summary judgment, the court conducts the step one separate inquiries with all the relevant evidence and rea- sonable inferences cast in the light most favorable to the non-mo- vant. See J-B Weld Co., LLC v. Gorilla Glue Co., 978 F.3d 778, 789 (11th Cir. 2020). At step two, the court weighs the relative im- portance of the circumstantial facts and determines whether these facts, taken in the light most favorable to the non-movant, would permit a reasonable fact finder to infer likelihood of confusion. Id. Courts may grant summary judgment on likelihood of confusion even if some circumstantial facts favor the non-movant because the two-step analysis “presupposes that [the] various [circumstantial facts may] point in opposing directions.” Tana, 611 F.3d at 775 n.7. Before it decided the parties’ motions for summary judg- ment in this case, the District Court conducted the step one inquir- ies and found the following circumstantial facts as a matter of law: (1) FIC’s marks were “relatively weak;” (2) FIC’s marks were not sufficiently similar to FT&E’s mark; (3) both FIC and FT&E’s marks represented similar goods or services; (4) both FIC and FT&E “advertise their services using online advertising, websites and social media;” (5) FT&E did not intend to cause consumer con- fusion by infringing on FIC’s marks; (6) no evidence existed of ac- tual consumer confusion about FIC and FT&E’s marks; and (7) FT&E’s client base was “sophisticated and unlikely to be USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 14 of 40

14 Opinion of the Court 19-13390

confused.” FCOA, 416 F. Supp. 3d at 1388–95. For factor four, similarity of trade channels and customers, the Court found evi- dence that favored both parties and, instead of viewing the evi- dence in FIC’s favor, concluded the factor was “neutral.” Id. at 1392. In fact, throughout its opinion the Court consistently drew inferences against FIC by misapplying the Nunez standard, a mat- ter we discuss more infra Part IV. Id. at 1387 (citing Nunez, 572 F.2d at 1123–24). At step two, the District Court mechanically added up these findings and held, without any further analysis, that FT&E’s mark did not create a likelihood of confusion with FIC’s marks as a matter of law. Id. at 1394–95. On appeal, FIC argues that the District Court incorrectly conducted the inquiries as to the first, second, fourth, and fifth cir- cumstantial facts. FIC also argues that the District Court improp- erly weighed the circumstantial facts at step two. Accordingly, we proceed to review the Court’s determinations de novo. A. i. Strength of FIC’s Marks In trademark law, the strength of a mark is the second most important circumstantial fact and determines the scope of the mark’s protection. Frehling, 192 F.3d at 1335. Strength or “distinc- tiveness” describes a mark’s ability to allow consumers to identify the source of a good or service. Id.; John H. Harland Co. v. Clarke Checks, Inc., 711 F.2d 966, 973–74 (11th Cir. 1983). So, strength or distinctiveness is just another way of talking about consumer USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 15 of 40

19-13390 Opinion of the Court 15

recognition. When Joe Consumer goes to his local grocer and buys Coca-Cola branded products, he can rest assured that he has bought something with Coca-Cola’s quality standards. As a result, the stronger Coca-Cola’s mark, the easier it is for consumers to rec- ognize the product and its source, and thus the more likely it is that consumers will associate a similar mark with the same source as Coca-Cola-branded products. 4 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition (“McCarthy”) § 24:49, Westlaw (5th ed. database updated Dec. 2022). The stronger the mark, then, the greater the likelihood of confusion and the greater the protection given to the mark. Welding Servs., Inc. v. Forman, 509 F.3d 1351, 1361 (11th Cir. 2007). We have described two steps in assessing the strength of a mark: conceptual strength and commercial strength. The first step in assessing strength is to determine the “con- ceptual strength” of the mark. FIU, 830 F.3d at 1258; 2 McCarthy § 11:80. Conceptual strength describes the potential of a mark to aid consumer recognition, which we evaluate through an abstract linguistic analysis. Courts determine this potential by placing a mark on the sliding scale of trademark strength, from weakest to strongest: (1) generic, (2) descriptive, (3) suggestive, and (4) fanciful or arbitrary. Frehling, 192 F.3d at 1335; Two Pesos, Inc. v. Taco Cabana, Inc., 505 U.S. 763, 768, 112 S. Ct. 2753, 2757 (1992). As we have previously explained, [Generic marks] refer to a class of which an individual service is a member (e.g., “liquor store” used in USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 16 of 40

16 Opinion of the Court 19-13390

connection with the sale of liquor). Descriptive marks describe a characteristic or quality of an article or service (e.g., “vision center” denoting a place where glasses are sold). “Suggestive terms suggest characteristics of the goods and services and require an effort of the imagination by the consumer in order to be understood as descriptive.” For instance, “pen- guin” would be suggestive of refrigerators. An arbi- trary mark is a word or phrase that bears no relation- ship to the product (e.g., “Sun Bank” is arbitrary when applied to banking services).

Frehling, 192 F.3d at 1335 (citations omitted). Arbitrary, fanciful, and suggestive marks are generally strong. Freedom Sav. & Loan Ass’n v. Way, 757 F.2d 1176, 1182 & n.5 (11th Cir. 1985). Generic and descriptive marks are so weak that they are not valid trade- marks. 15 U.S.C. §§ 1115(b)(4), 1065(4). However, if a descriptive mark, like FOREMOST, acquires “secondary meaning,” then the descriptive mark is strong enough to be valid under the Lanham Act. Royal Palm Props., LLC v. Pink Palm Props., LLC, 950 F.3d 776, 782–83 (11th Cir. 2020). Descriptive marks could (in the ab- stract) refer to many entities. So, a mark has secondary meaning when consumers view the mark as synonymous with the mark holder’s goods or services. For example, American Airlines could theoretically refer to any airline based in North or South America. See id. at 783. But with the mark holder’s time and effort, Ameri- can Airlines now calls to mind a specific airline through its second- ary meaning. See id. Incontestable descriptive marks, like the USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 17 of 40

19-13390 Opinion of the Court 17

FOREMOST marks, are statutorily presumed to be valid and thus must have some degree of secondary meaning. Dieter v. B & H Indus. of Sw. Fla., Inc., 880 F.2d 322, 328 (11th Cir. 1989). Other- wise, they would not be valid marks at all. Id. Incontestable descriptive marks are also presumed, in our circuit, to be “relatively strong mark[s].” Id. at 329; see also Sover- eign Mil. Hospitaller Ord. of Saint John of Jerusalem of Rhodes & of Malta v. Fla. Priory of the Knights Hospitallers of the Sovereign Ord. of Saint John of Jerusalem, Knights of Malta, the Ecumenical Ord. (“Sovereign Mil.”), 809 F.3d 1171, 1183 (11th Cir. 2015). 13 This Dieter presumption can be rebutted by looking to the second step: commercial strength. FIU, 830 F.3d at 1256–60. Commercial strength refers to the real-world consumer recognition of a mark, most often created by the efforts and work

13Marks that are registered start off as contestable. Marks become incontest- able once they have been registered on the Principal Register with the PTO for at least five years, among other statutory formalities. Frehling, 192 F.3d at 1336; 15 U.S.C. § 1065 (listing the requirements for incontestability). FT&E argues that this Court is “an outlier” insofar as we recognize a connection between incontestable status and mark strength. We have openly admitted as much. Sovereign Mil., 809 F.3d at 1183 (indicating that Dieter is arguably incorrect because whether a mark is registered says nothing about consumer perceptions). Although Dieter may rest on faulty ground, a deci- sion being wrong does not mean that it lacks legal force. See id. at 1184. In our circuit, prior precedent (even if erroneous) continues to bind us until over- turned en banc or by an opinion of the Supreme Court. Id. We are still bound to follow Dieter. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 18 of 40

18 Opinion of the Court 19-13390

of the mark holder. See id. at 1258 (“It is surely true that focusing solely on conceptual strength is an ‘incomplete’ method of analysis . . . .”); John H. Harland Co., 711 F.2d at 974 n.13. We have held that “[d]etermining the strength of any mark requires weighing ei- ther or both circumstantial evidence of advertising and promotion and direct evidence of consumer recognition, such as by a survey.” FIU, 830 F.3d at 1259 (quoting 2 J. Thomas McCarthy, McCarthy on Trademarks and Unfair Competition § 11:83 (4th ed. 2016)). Commonly used evidence of commercial strength includes third party use; advertising and promotion; sales and number and types of customers; recognition by trade, media, and customers; and sur- vey of likely customers. 2 McCarthy § 11:81. As relevant here, the Dieter presumption can be rebutted by a strong showing of third-party use of the mark that significantly impacts consumer recognition of the original mark. See FIU, 830 F.3d at 1257; Univ. of Ga. Athletic Ass’n v. Laite, 756 F.2d 1535, 1545 n.27 (11th Cir. 1985) (stating that third-party use matters in determining “whether the unauthorized third-party uses signifi- cantly diminish the public’s perception that the mark identifies items connected with the owner of the mark”). In assessing the third-party use, we consider: (1) the frequency of third-party use, (2) the full names that the third-party uses, and (3) “the kind of busi- ness in which the user[s] [are] engaged.” FIU, 830 F.3d at 1257. Though the number of third-party uses is important, “there is no hard-and-fast rule establishing a single number that suffices to weaken a mark.” Savannah Coll. of Art & Design, Inc. v. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 19 of 40

19-13390 Opinion of the Court 19

Sportswear, Inc., 983 F.3d 1273, 1283 (11th Cir. 2020) (quoting FIU, 830 F.3d at 1257). Moreover, third-party uses in the same market diminish a mark’s strength more than uses in other markets.14 See Amstar Corp. v. Domino’s Pizza, Inc., 615 F.2d 252, 259–60 (5th Cir. 1980); Sun Banks of Fla., Inc. v. Sun Fed. Sav. & Loan Ass’n, 651 F.2d 311, 316 (5th Cir. July 1981). 15 And, because the consum- ing public is unlikely to be aware of mere federal registrations of third-party marks, such evidence is not probative of the diminished

14We have not been entirely clear about whether third-party uses in other markets diminishes a mark’s strength. At times we have appeared to say they don’t. See PlayNation, 924 F.3d at 1166 (“[S]imilar marks used by third parties in unrelated businesses or markets do not diminish the strength of a mark in a particular market.”); Safeway Stores, Inc. v. Safeway Disc. Drugs, Inc., 675 F.2d 1160, 1165 (11th Cir. 1982). At other times we have said they do. See Amstar Corp. v. Domino’s Pizza, Inc., 615 F.2d 252, 259–60 (5th Cir. 1980) (“We do not believe that such extensive third-party use and registration of ‘Domino’ can be so readily dismissed. The impact of such evidence is not dispelled merely because ‘Domino’ cigarettes and matches are not leading brands, or because some uses of the mark ‘Domino’ by third parties have not been related to food products.”); Sun Banks of Fla., Inc. v. Sun Fed. Sav. & Loan Ass’n, 651 F.2d 311, 316 (5th Cir. July 1981). The best way to synthesize the caselaw is to say that other-market uses can diminish a mark’s strength, but not always to a significant extent—certainly not always to the point of making a mark weak. 15 Bonner v. City of Prichard, 661 F.2d 1206, 1207 (11th Cir. 1981) (en banc) (all Fifth Circuit decisions handed down before October 1, 1981, are binding precedent in the Eleventh Circuit). USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 20 of 40

20 Opinion of the Court 19-13390

distinctiveness of the original mark.16 Turner v. HMH Publ’g Co., 380 F.2d 224, 228 & n.2 (5th Cir. 1967). Here, “Foremost” is a descriptive mark, a self-laudatory term meaning the best. Platinum Home Mortg. Corp. v. Platinum Fin. Grp., Inc., 149 F.3d 722, 728 (7th Cir. 1998) (describing self- laudatory marks as descriptive). Because it is incontestable, this mark is presumed to be a relatively strong mark. Dieter, 880 F.2d at 329. The District Court held that the existence of 62 registered trademarks and 541 registered business names in various states us- ing the term “foremost” provided evidence of third-party use

16 To be fair, we have been unclear on this point as well. We have stated that “mere registrations” of similar marks will not weaken a mark. Turner v. HMH Publ’g Co., 380 F.2d 224, 228 & n.2 (5th Cir. 1967). At other times, however, we’ve seemingly relied on mark registrations and business lists as a proxy for distinctiveness without demanding any evidence that they have affected pub- lic perception in any way. See AmBrit, Inc. v. Kraft, Inc., 812 F.2d 1531, 1539 (11th Cir. 1986); Am. Heritage Life Ins. Co. v. Heritage Life Ins. Co., 494 F.2d 3, 13 (5th Cir. 1974), abrogated by B & B Hardware, Inc. v. Hargis Indus., Inc., 575 U.S. 138, 135 S. Ct. 1293 (2015). However, our oldest case in this arena directly addressing this ques- tion, Turner, controls. The District Court below used an even older case as support for the proposition that registrations are evidence of third-party use. See El Chico, Inc. v. El Chico Cafe, 214 F.2d 721, 725 (5th Cir. 1954). In El Chico, the Former Fifth Circuit considered trademark registrations alongside other evidence of third-party use to determine that a mark was weak. Id. Still, El Chico did not directly address this issue, and so it did not foreclose the hold- ing in Turner. Thus, Turner controls. See Scott v. United States, 890 F.3d 1239, 1257 (11th Cir. 2018) (“The prior-panel-precedent rule requires subse- quent panels of the court to follow the precedent of the first panel to address the relevant issue . . . . ”). USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 21 of 40

19-13390 Opinion of the Court 21

sufficient to rebut the Dieter presumption. FCOA, 416 F. Supp. 3d at 1390 (citing El Chico, Inc. v. El Chico Cafe, 214 F.2d 721, 725 (5th Cir. 1954) and AmBrit, Inc. v. Kraft, Inc., 812 F.2d 1531, 1539 (11th Cir. 1986)). The Court did so without analyzing the industries or names of the marks or businesses presented. Id. On appeal, FIC first argues that the District Court erred by relying solely on trade- mark registrations and business lists to rebut the Dieter presump- tion. FIC argues in addition that the District Court erred when it failed to analyze the additional “commercial strength” evidence FIC proffered of its marketing and promotional efforts. Conducting the analysis de novo, we conclude that the Di- eter presumption remains unrebutted. As explained above, the mere fact that a mark has been registered or that a business is named in a registry is not evidence of third-party use. Turner, 380 F.2d at 228. FT&E simply provided a list of businesses printed from Secretary of States’ webpages and trademark registrations. In re- sponse, FIC has shown that none of the “active” entities on the business lists (other than FIC) contain the word “insurance” in their name, the business lists do not show which industries the compa- nies operate in, and no registered trademarks with the term “Fore- most” are registered for use in the insurance industry (other than FIC). Our review of the evidence in the light most favorable to FIC shows that there is no reliable evidence that these businesses or marks are active. Inactive businesses and marks are not relevant to our analysis. See id. Given this (lack of) evidence, we hold that the USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 22 of 40

22 Opinion of the Court 19-13390

Dieter presumption remains unrebutted, and FIC’s marks are still “relatively strong.” Dieter, 880 F.2d at 329. FIC has also provided evidence of consumer recognition that bolsters its already presumptively strong mark. While FT&E argues FIC only introduced evidence of its $7,000,000-a-year adver- tising budget, 17 FIC has also provided evidence about the size and scope of its agent class, its over $2.4 billion in annual insurance pre- miums, its recognition in independent publications, an AARP en- dorsement, and a survey showing that a majority of South Florida respondents had heard of FIC. Viewing the evidence in FIC’s favor, a reasonable factfinder conducting a separate inquiry on the strength of FIC’s marks could find the marks strong based on both

17 We have previously doubted the probative value of “raw advertising fig- ures,” that is, the dollar amount that a company spends to advertise its mark. In FIU, we stated that a district court under clear error review was permitted to discount evidence of raw advertising figures, standing alone, as it impacts the strength of the mark. 830 F.3d at 1259 (“There simply was not sufficient evidence of commercial strength in the record to [require] the district court to ignore the substantial third-party usage.”). We reasoned that raw advertising figures alone tell us little about the efficacy of those efforts in the mind of con- sumers. Id. We stated that this evidence was far more probative if there was comparative spending evidence with others in the industry or direct evidence of consumer recognition. Id. In a later case, we held that it was also not clear error for a court to find, with additional evidence beyond raw advertising fig- ures, that advertising expenditures contribute to the strength of a mark. Play- Nation, 924 F.3d at 1166 n.3. Because it was not clear error to consider that evidence in PlayNation, we can consider that same type of evidence on sum- mary judgment. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 23 of 40

19-13390 Opinion of the Court 23

the Dieter presumption and the additional evidence of commercial strength. ii. Similarity of the Marks The second inquiry requires us to examine the similarity be- tween the parties’ marks. FIU, 830 F.3d at 1260. “The greater the similarity, the greater the likelihood of confusion.” Id. Of course, the marks don’t need to be identical to support a finding of similar- ity, because the key is to determine if the similarities are sufficient to deceive the public. Id. We “consider[] the overall impressions that the marks create, including the sound, appearance, and man- ner in which they are used,” rather than comparing isolated fea- tures. Frehling, 192 F.3d at 1337; Sovereign Mil., 809 F.3d at 1186. Because of its malleability, we have described this analysis as a “subjective eyeball test.” AmBrit, 812 F.2d at 1540. The District Court stated that both marks started with “Foremost” and looked similar at first glance. FCOA, 416 F. Supp. 3d at 1390–91. However, the District Court focused on the differ- ences in color, fonts, and logos, and the words that followed the FOREMOST marks. Id. Specifically, it found that the words “title” and “escrow” separated FT&E’s mark from FIC’s FOREMOST marks, because it is not obvious to the public that “title” refers to title insurance. Id. Thus, the District Court concluded that the marks were dissimilar. Id. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 24 of 40

24 Opinion of the Court 19-13390

FIC argues that the District Court did not fully consider the commercial impression of the marks and that its side-by-side anal- ysis of the marks was insufficient. Considering similarity anew, we believe a reasonable factfinder could find that the parties’ marks are similar in sound, appearance, meaning and commercial impres- sion:

Our analysis focuses on the distinctive parts of marks. See AmBrit, 812 F.2d at 1541; John H. Harland Co., 711 F.2d at 976. “Foremost” is the most distinctive part of both parties’ marks, and far more important than generic words like title and escrow. See PlayNation Play Sys., Inc. v. Velex Corp., 924 F.3d 1159, 1168 (11th Cir. 2019); John H. Harland Co., 711 F.2d at 976. With that frame of reference, the marks are similar in sight, sound, and meaning. The logos create a similar overall effect and accentuate the marks’ similarities, because both feature two lines of text, with “Fore- most” in bold, sans-serif type above smaller letters detailing the ge- neric parts of the marks, to the right of a stylized “F.” See Safeway Stores, Inc. v. Safeway Disc. Drugs, Inc., 675 F.2d 1160, 1165 (11th Cir. 1982) (focusing on overall effect of a mark and ignoring its non- distinctive parts); Exxon Corp. v. Tex. Motor Exch. of Hous., Inc., 628 F.2d 500, 505 (5th Cir. 1980) (noting that two marks were sim- ilar when they were both in block letters on an all-white back- ground with blue underneath). This conclusion is bolstered by considering how the marks are used in the actual world, something USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 25 of 40

19-13390 Opinion of the Court 25

the District Court did not do. See Conagra, Inc. v. Singleton, 743 F.2d 1508, 1514 (11th Cir. 1984) (noting that “Singleton,” the most distinct part of the mark at issue, was emphasized on materials in the marketplace). FT&E often refers to itself simply as “Foremost” on its website, which is precisely FIC’s trademark and the most crit- ical part of other marks. Admittedly, there are some differences in the marks because the parties use different fonts and colors. FT&E relies on a green and gold color scheme, whereas FIC relies on black and blue colors. We do not find these minor differences to be significant on sum- mary judgment, given that “Foremost” is the dominant part of both marks and what consumers would focus on. Conagra, 743 F.2d at 1514; see also Carnival Brand Seafood Co. v. Carnival Brands, Inc., 187 F.3d 1307, 1312 & n.7 (11th Cir. 1999); Safeway, 675 F.2d at 1165. Those differences are even less important because consumers are unlikely to confront them side-by-side in the real- world where they could be discerning about those differences. See Sun-Fun Prods., Inc. v. Suntan Rsch. & Dev. Inc., 656 F.2d 186, 192 (5th Cir. Unit B Sept. 1981) (stating that the likelihood of confusion may be increased when consumers are unable “to compare the products side by side and observe the precise differences in appear- ance”). Drawing all inferences in FIC’s favor, a reasonable fact- finder could determine that these two marks are similar. iii. Similarity of the Products The third circumstantial fact, the similarity of the products, concerns whether the products are of a kind the public could think USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 26 of 40

26 Opinion of the Court 19-13390

originate from a single source. Frehling, 192 F.3d at 1338. Here, both parties sell insurance. FT&E argues that there can be no con- fusion because a consumer could not purchase a policy of FIC that overlaps with a policy of FT&E, that the policy itself would not bear the FOREMOST marks because FT&E is merely an agent of larger national title insurance companies, and that FIC cannot con- duct closings or issue title insurance under Florida law. While title insurance is a monoline industry in Florida, consumers are unlikely to know that and, even if they did, could potentially assume that FT&E was a subsidiary or affiliate of FIC. Moreover, the logo or trademark on the insurance policy does not somehow transform insurance into the mere work of an agent separate from the insur- ance itself and, even if it did, by the time FT&E’s customers actu- ally see FT&E’s policies, they almost certainly have already been exposed to FT&E’s “foremost” mark. Accordingly, a reasonable factfinder could find that the parties’ products are similar. iv. Similarity of Trade Channels and Customers The fourth circumstantial fact, similarity of trade channels and customers, focuses on “where, how, and [with] whom” the parties transact with their actual and potential customers. Sover- eign Mil., 809 F.3d at 1187–88 (“Dissimilarities between the retail outlets for and the predominant consumers of plaintiff’s and de- fendants’ goods lessen the possibility of confusion . . . .” (quoting Amstar Corp., 615 F.2d at 262)); Safeway, 675 F.2d at 1166; see also Freedom Sav. & Loan Ass’n, 757 F.2d at 1184–85, 1184 n.7 (holding for this circumstantial fact that it is enough if a plaintiff “show[s] USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 27 of 40

19-13390 Opinion of the Court 27

that the same customers are likely to use both services”). The pri- mary focus in this inquiry is on the overlap of the customer bases, because the greater the overlap, the greater the likelihood that con- sumers will be exposed to both marks and become confused. See Savannah Coll. of Art & Design, Inc., 983 F.3d at 1284. Therefore, direct competition or identity of sales is not required; we look to whether the companies “cater to the same general kinds of individ- uals.” Sovereign Mil., 809 F.3d at 1188; PlayNation, 924 F.3d at 1168. Likewise, the similarity of trade channels analysis focuses on whether the medium (e.g., stores, agents, online, mail, etc.) that customers frequent would expose them to both marks, not on whether the products or services are sold in the same location or manner. Frehling, 192 F.3d at 1339; see also Century 21 Real Est. Corp. v. Century Life of Am., 970 F.2d 874, 877 (Fed. Cir. 1992). Here, the District Court correctly recognized that FIC pro- duced evidence tending to show that FIC and FT&E’s customer bases overlapped, i.e., both targeted homeowners seeking home insurance-related products. FCOA, 416 F. Supp. 3d at 1391–93. However, the District Court also credited FT&E’s argument that FIC and FT&E “differ[ed]” because most of FT&E’s customers came from referrals by Stok Folk + Kon, realtors, brokers, and other agents. Id. at 1392. Faced with contradictory evidence, the Court found this factor to be neutral instead of applying the sum- mary judgment standard. Id. Conducting the analysis de novo, we agree with the District Court that both parties targeted the same type of individuals, i.e., USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 28 of 40

28 Opinion of the Court 19-13390

“real property purchasers” or sellers who need insurance. Accord- ingly, FIC has put forth evidence tending to show that its customer base overlapped with FT&E’s. However, we disagree with the Dis- trict Court and FT&E that FT&E has introduced any evidence tending to show that its customer base did not overlap with FIC’s. FT&E primarily gets customers from its referral system, yes, but the persons buying title insurance are the same type of people (home buyers and sellers) who will likely need homeowners’ insur- ance and thereby could be exposed to both marks. See Freedom Sav. & Loan Ass’n, 757 F.2d at 1184 n.7 (noting that it is sufficient for plaintiffs “to show that the same customers are likely to use both services”); Safeway, 675 F.2d at 1166 (focusing on the overlap of actual or potential customers). Nothing prevents a potential home buyer or seller from purchasing homeowner’s insurance from FIC, noting the FOREMOST mark, and then being referred to FT&E for closing services and title insurance. Indeed, consider- ing the evidence FIC has introduced about its presence in Florida, this scenario is quite plausible. On appeal, FT&E defends the Court’s finding that FT&E’s customer base differed from FIC with three additional arguments: (1) FT&E only sells through its physical location in Aventura, not through other trade channels or in other geographic regions; (2) FT&E sells directly to consumers at its office, as opposed to inde- pendent agents; and (3) title insurance is a monoline industry in Florida. We find all three unpersuasive. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 29 of 40

19-13390 Opinion of the Court 29

As an established multi-billion dollar nationwide insurance company, FIC naturally sells insurance through many more medi- ums and in many more locations than a small start-up like FT&E. The question, however, is not whether one party’s trade channels and customer base exceeds the other’s, but whether the parties’ trade channels and customer bases overlap. FIC’s insurance prod- ucts are sold by its independent agents in physical locations throughout Florida. In other words, FIC sells insurance in Florida in the exact same manner as FT&E. Under our precedent, how- ever, FIC would not even need a physical presence in Florida to show overlap. In Safeway, we held that even though Safeway the grocer had no physical stores in Florida, its in-state food purchases provided a presence that allowed consumers to potentially mistake Safeway Discount Centers for Safeway the grocer. 675 F.2d at 1166. Likewise, FIC’s over 95,000 customers in Florida establish its presence in Florida. Nor does title insurance being a monoline in- dustry in Florida prevent the parties’ customer bases from overlap- ping; as explained above, home buyers and sellers are likely to buy both homeowner’s insurance and title insurance. Drawing all rea- sonable inferences in favor of FIC, as we must, a reasonable fact- finder could find that FIC’s actual and potential consumer base overlaps with FT&E’s. v. Similarity of Advertising The fifth circumstantial fact, similarity of advertising, fo- cuses on the audience reached by the advertisements of the parties. Sovereign Mil., 809 F.3d at 1187–88. Like with similarity of trade USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 30 of 40

30 Opinion of the Court 19-13390

channels and customer bases, the greater the overlap or similarity of the audiences, the greater the likelihood of confusion. Id. Iden- tity of advertising methods is not required, but instead we assess whether the overlap in readership of the parties’ advertisements is “significant enough” that “a possibility of confusion could result” in a fashion very similar to the previous factor. PlayNation, 924 F.3d at 1168–69; Sovereign Mil., 809 F.3d at 1188 (quoting Frehling, 192 F.3d at 1340). As the District Court found, both parties advertise through similar mediums, i.e., online advertising, magazines, brochures, emails, social media, and their websites. FCOA, 416 F. Supp. 3d at 1392. On appeal, FT&E argues that it directs its advertisements to a different “universe of consumers” than FIC, namely “real estate developers, purchasers, sellers, lenders, and borrowers,” and so there is no audience overlap. Appellant Br. at 41–42. But FT&E’s argument fails on its face. Even if FT&E focuses its advertisements on real estate professionals, it admits it also advertises to “purchas- ers, sellers, lenders, and borrowers,” the exact sort of people who may be interested in buying FIC’s homeowner’s insurance. For ex- ample, FT&E’s website—which bears FT&E’s logo on each of its pages—lists “homeowners” as among the customers served by USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 31 of 40

19-13390 Opinion of the Court 31

FT&E. On summary judgment, this is sufficient to show overlap in advertisement audience.18 vi. Defendant’s Intent With respect to the sixth circumstantial fact, the defendant’s intent to confuse consumers, the District Court found that FT&E had no intent to “capitalize on . . . [FIC]’s business reputation.” FCOA, 416 F. Supp. 3d at 1393 (quoting FIU, 830 F.3d at 1263). Be- cause FIC did not take issue with this finding nor the weight as- cribed to it, FIC has forfeited the analysis applicable to this circum- stantial fact and we decline to address it further. See United States v. Campbell, 26 F.4th 860, 873 (11th Cir. 2022) (en banc) (explaining that issues not briefed on appeal are forfeited and thus may only be addressed in extraordinary circumstances). vii. Actual Confusion Actual confusion asks whether there is evidence in fact of confusion. Frehling, 192 F.3d at 1340. Though it is the most im- portant circumstantial fact, it is not a requirement for finding a like- lihood of confusion. Id. The District Court found no evidence of actual confusion. FCOA, 416 F. Supp. 3d at 1393–95. While FIC produced two expert witnesses who used internet surveys as the bases of their opinions that there was the potential for actual

18 We note, however, that as the internet ages and becomes ubiquitous, having

a website or advertising online informs the court precious little about expo- sure and confusion. See Network Automation, Inc. v. Advanced Sys. Con- cepts, Inc., 638 F.3d 1137, 1151 (9th Cir. 2011). USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 32 of 40

32 Opinion of the Court 19-13390

confusion, the District Court discounted this evidence. Id. at 1394. The District Court stated that survey evidence was of slight weight and viewed unfavorably in the Eleventh Circuit, and FIC’s surveys were even less probative because they did not distinguish between confusion caused by a word (that is, “Foremost”) and the whole mark. Id. On appeal, FIC does not dispute the lack of evidence of actual confusion. Instead, FIC argues that the lack of confusion should be given no weight because there was no time for actual confusion to develop. We agree with FIC. While no hard and fast rules set how much evidence is necessary to show actual confusion, we have said that courts must take into account the circumstances of each case. J-B Weld Co., 978 F.3d at 793 (citing Lone Star Steakhouse & Sa- loon, Inc. v. Longhorn Steaks, Inc., 122 F.3d 1379, 1382 (11th Cir. 1997)). Those circumstances include the extent of advertising, the length of time for which an infringing product has been advertised, and any other factors that might influence the reporting of actual confusion. The District Court did not consider these circum- stances when assessing the lack of evidence of actual confusion. In Hard Candy, LLC v. Anastasia Beverly Hills, Inc., we stated that a lack of evidence showing actual confusion can be dis- counted when there is not an “adequate period of time” for actual confusion to develop among consumers. 921 F.3d 1343, 1362–63 (11th Cir. 2019). In that case, cosmetic maker Anastasia had sold nearly 250,000 makeup kits over a period of eight months, contain- ing a mark allegedly similar to Hard Candy’s trademark. Id. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 33 of 40

19-13390 Opinion of the Court 33

Despite that, Hard Candy could not point to a single instance of actual confusion. Id. We stated that the district court did not clearly err in finding that the lack of evidence of actual confusion was probative. Id. However, on summary judgment, we must take every inference in FIC’s favor, something the District Court did not do here. Even assuming it was correct to reject the survey data that FIC provided, we think the lack of evidence of actual con- fusion is not particularly probative. A reasonable inference for FIC is that actual confusion did not have adequate time to develop, given that FT&E started in May 2015, and then conducted at least 20 closings between May 2016 and November 2018, when motions for summary judgment had been filed. This is different from Hard Candy, where we held it was not clear error to state that the lack of actual confusion was probative where potentially millions of consumers were exposed to the infringing mark, hundreds of thou- sands of consumers bought the makeup palette, and not a single instance of actual confusion arose. See id.; see also Tana, 611 F.3d at 779–80 (finding that no reasonable jury could find actual confu- sion where there were two instances of actual confusion where a company served over a million customers in five years, and affirm- ing a grant of summary judgment for the defendant). Thus, a rea- sonable factfinder could discount the importance of the evidence of actual confusion. viii. Consumer Sophistication Typically, we analyze likelihood of confusion using only seven factors or, as we put it earlier, seven separate inquiries. See, USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 34 of 40

34 Opinion of the Court 19-13390

e.g., Wreal, LLC v. Amazon.com, Inc., 38 F.4th 114, 127 (11th Cir. 2022) (“In determining the likelihood of confusion, we consider the following seven factors . . .”); Sovereign Mil., 809 F.3d at 1181 (same); Tana, 611 F.3d at 774–75 (same); Frehling, 192 F.3d at 1335 (same). Indeed, courts in this circuit are required to consider the seven factors analyzed above when conducting a likelihood of con- fusion analysis. J-B Weld Co., 978 F.3d at 794. However, we have recognized that consumer sophistication may also be relevant to assessing likelihood of confusion. See, e.g., FIU, 830 F.3d at 1256, 1265 (analyzing consumer sophistication separately from the seven factors); Welding Servs., 509 F.3d at 1361 (“[S]ophisticated consum- ers [of complex goods or services] . . . are less likely to be confused than casual purchasers of small items.”); Freedom Sav. & Loan Ass’n, 757 F.2d at 1185 (“[S]ince most of these customers are mak- ing a major investment, they are likely to be especially well-in- formed buyers. The sophistication of a buyer certainly bears on the possibility that he or she will become confused by similar marks.”). After all, consumers that either have special knowledge of the industry through education or experience, see Welding Servs., 509 F.3d at 1361, or have invested significant time into be- coming well-informed due to the nature of the purchase, see Free- dom Sav. & Loan Ass’n, 757 F.2d at 1184–85, are more likely to distinguish between similar marks and thereby avoid becoming confused. As both the parties and the District Court considered consumer sophistication, we find it appropriate to do so as well. In so doing, we recognize only that consumer sophistication may im- pact likelihood of confusion and do not require that the factor be USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 35 of 40

19-13390 Opinion of the Court 35

considered in every, or even most, likelihood of confusion anal- yses. Cf. J-B Weld Co., 978 F.3d at 794; see also Star Indus., Inc. v. Bacardi & Co. Ltd., 412 F.3d 373, 389–90 (2d Cir. 2005) (likewise treating consumer sophistication as a separate factor). The District Court accepted FT&E’s argument that its “cli- ent base is largely comprised of real estate developers, referring realtors, sellers, lenders and mortgage brokers who are sophisti- cated and unlikely to be confused.” FCOA, 416 F. Supp. 3d at 1392. We agree with FT&E that real estate developers and realtors are sophisticated in real estate purchases and that individual home buy- ers and sellers are likely to be sophisticated due to the importance and size of the transaction. Appellee Br. at 19. However, as FT&E points out, “consumers generally have minimal involvement in the selection of a title insurance provider and rely entirely on the rec- ommendations of their agents or professionals.” Id. at 19–20. Or, in other words, individual home buyers and sellers may be well- informed when buying or selling a home, but that sophistication does not transfer to buying title insurance. This makes sense; title insurance is a complex legal matter that, for the consumer, only involves a single, relatively small payment. 19 Thus, home buyers

19 In Florida, Office of Insurance Regulation sets title insurance premiums.Fla. Stat. § 627.7711; Fla. Admin. Code Ann. r. 69O-186.003(1). Currently, the rates are $5.75 of premium per $1,000 of the purchase price for the first $100,000 of liability written and $5.00 per $1,000 of liability written from $100,000 to $1 million. Fla. Admin. Code Ann. r. 69O-186.003(1). So, for a policy that insures the new owner for $500,000, the title insurance premium would be at most a one-time payment of $2,575. USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 36 of 40

36 Opinion of the Court 19-13390

and sellers are likely to be unsophisticated when buying title insur- ance. FT&E argues that the unsophistication of these consumers weighs in its favor because their unsophistication makes them more likely to rely on sophisticated agents and professionals who would be more likely to be able to distinguish between similar marks. Appellant Br. at 18–19. FT&E may well be able to establish this tendency at trial. But on summary judgment, we must draw inferences in the non-movant’s favor, and we see an alternative in- ference that may be drawn from consumers’ unsophistication about title insurance. While unsophisticated consumers might rely on the advice of sophisticated professionals, they might also rely on their earlier research into home purchases and insurance. If they came across and trusted FIC’s FOREMOST brand, then they may decide to trust another “foremost”-branded insurance company due to its possible affiliation with FIC. In fact, this possibility is precisely what companies seek to avoid when they file trademark infringement lawsuits. So, without sufficient evidence in the rec- ord to definitively decide this issue one way or the other and being on summary judgment, we make this inference and hold that a rea- sonable factfinder could find FT&E’s customer base to be unso- phisticated and thus more likely to be confused by the similarity between the parties’ marks. B. To recap, our separate inquiries on the evidence have yielded the following circumstantial facts under the summary USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 37 of 40

19-13390 Opinion of the Court 37

judgment standard: (1) FIC’s mark is strong; (2) FIC and FT&E’s marks are similar; (3) FIC and FT&E sell similar products; (4) FIC and FT&E’s trade channels and customers overlap; (5) FIC and FT&E’s advertising audiences overlap; (6) FT&E did not intend to cause consumer confusion about the existence of a relationship be- tween the parties; (7) there is no evidence that consumers have ac- tually confused FT&E with FIC; and (8) title insurance purchasers are unsophisticated and thus may be confused by similar marks. Now, at the second step, we must weigh these circumstantial facts in the light most favorable to FIC to determine whether a reasona- ble factfinder could infer the ultimate fact, likelihood of confusion. Here, similar unsophisticated consumers would see similar marks, displayed in similar advertising media directed at similar audiences, selling similar insurance products through similar mediums in the state of Florida. While there is no evidence of actual confusion amongst consumers about the parties’ relationship, this lack of ev- idence has relatively little weight under these facts because there simply was not enough time for actual confusion to develop. See Hard Candy, 921 F.3d at 1362–63. Similarly, while evidence of an intent to cause consumer confusion weighs heavily in favor of find- ing a likelihood of confusion amongst consumers, 20 absence of that evidence in no way prevents consumers from likely becoming con- fused. Frehling, 192 F.3d at 1340. Consequently, a reasonable

20 See Freedom Sav. & Loan Ass’n, 757 F.2d at 1185 (“If a person intends to induce confusion among customers, he or she is likely to succeed”). USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 38 of 40

38 Opinion of the Court 19-13390

factfinder could find a likelihood that consumers would be con- fused by the marks. IV. In closing, two thoughts. First, on cross-motions for summary judgment and espe- cially when applying the Nunez framework, courts should be very careful in their analysis to ensure that the proper party receives the benefit of the summary judgment standard. When parties jointly move for summary judgment, the court has three options: granting summary judgment for the plaintiff under the defendant’s best case, granting summary judgment for the defendant under the plaintiff’s best case, or denying both motions for summary judg- ment and proceeding to trial. Before granting summary judgment for a party, the court must consider the evidence in the light most favorable to the non-movant and, unless “there are no genuine is- sues of material fact,” i.e., all material facts have “been incontro- vertibly proved,” and the trial judge is the finder of fact, the court must also draw all inferences in the non-movant’s favor. Nunez, 572 F.2d at 1123–24. Only once this is done may a court determine if a party is entitled to judgment as a matter of law; should any ma- terial questions of fact remain that may cause a reasonable fact- finder to rule in the non-movant’s favor, summary judgment must be denied. Our review of the decision in this case revealed that the Dis- trict Court never discussed the summary judgment standard in its USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 39 of 40

19-13390 Opinion of the Court 39

analysis after citing Nunez in the “Legal Standard” section for its decision. See generally FCOA, 416 F. Supp. 3d at 1387–95. The closest the Court came to applying the summary judgment stand- ard was quoting a footnote from Tana stating that “[a]lthough like- lihood of confusion is a question of fact, it may be decided as a mat- ter of law.” Id. at 1388 (quoting Tana, 611 F.3d at 775 n.7). True— but only after viewing the evidence in the light most favorable to the non-movant. While the Court implicitly decided this case un- der the Nunez framework, it never actually decided whether all the material facts had been “incontrovertibly proved.” Nunez, 572 F.2d at 1124; see generally FCOA, 416 F. Supp. 3d at 1387–95. A district court may not ignore the traditional summary judgment standard merely by invoking the specter of Nunez. Nunez itself acknowledged that summary judgment can be “a ‘lethal weapon’ capable of ‘overkill’” and that situations where the Nunez standard is appropriate “may be rare.” 572 F.2d at 1223– 24 (quoting Brunswick Corp. v. Vineberg, 370 F.2d 605, 612 (5th Cir. 1967)). Indeed, the Nunez standard can easily become a trap for unwary district courts and litigants; wise lawyers and judges would do well to remember that this Court’s de novo standard of review is “unaffected by any inferential conclusions reached be- low.” Useden, 947 F.2d at 1573 n.14. Accordingly, the Nunez standard should be reserved for those rare cases where it is justified. Second, cross-motions for summary judgment in non-jury cases are a very inefficient method by which to decide a case. When parties move for summary judgment, they tell the court that USCA11 Case: 19-13390 Document: 39-1 Date Filed: 01/12/2023 Page: 40 of 40

40 Opinion of the Court 19-13390

they are ready for a determination on the merits and that no mate- rial fact issues remain for a factfinder to decide. If both parties move for summary judgment after conducting extensive discov- ery, then the parties effectively agree that no fact issues exist. But when the court grants summary judgment in these situations, the loser will invariably take an appeal and argue that there was a fact dispute or credibility issue that prevented the court from deciding the case, despite arguing the opposite below. The winner will ar- gue that no fact issues remain. If a fact issue does remain, the case must be reversed and remanded to hold a bench trial, which should have occurred in the first place. And, of course, the loser of the bench trial will then appeal. In this case, for efficiency’s sake, the parties should have es- chewed moving for summary judgment, informed the court that discovery was complete and that the case was ready for trial, and then held a bench trial. 21 Because we hold that a reasonable factfinder could deter- mine that a likelihood of confusion exists, we reverse the District Court’s grant of summary judgment as to Count I of FIC’s com- plaint and remand the case for trial on the merits. REVERSED AND REMANDED.

21The case’s docket sheet indicates that on May 16, 2019, the case had been scheduled for a bench trial on September 3, 2019. The order granting FT&E summary judgment was entered on August 1, 2019.