ORDER ON PRELIMINARY INJUNCTION
Mark E. Walker, United States District JudgeI. INTRODUCTION
You live in Gainesville and need to book a party bus, or perhaps a non-party bus or a limo to take a number of people to an event. You run a Google search for “Gainesville party bus” and, not satisfied with any of the offerings on page one of the results, turn to page two. There is a listing there for “Uber Promotions.” Thinking that perhaps this is somehow affiliated with Uber, a nationally known taxi-like service that has recently come to town, you click on the link. You are taken to a webpage with a bright green and purple “über PROMOTIONS” logo. The crux of this trademark infringement case is (roughly speaking) whether you could reasonably conclude that Uber Promotions and Uber the taxi-like service are in some way connected.
Plaintiff Uber Promotions, Inc. (“Promotions”) is a Gainesville, Florida company that provides a variety of services, including “promotional and event planning services, ... graphic, web design and print media photography services, ... modelling and talent agency services, [and] ... private venue rental services.” ECF No. 47, at 5 ¶ 12. It also provides “passenger transportation services, including through limousine and charter services.” Id. Promotions claims that it has been using “UBER,” “ÜBER,” “UBER PROMOTIONS,” and “UBER PROMOTIONS” since at least 2006. ECF No. 27-1, at 3 ¶¶ 5-6. Promotions is run by its President, Joey Friedman, ECF No. 36-1, at 2 ¶ 3, and its Vice-President, Michael Farzad, ECF No. 27-1, at 1 ¶ 3.
Defendant Uber Technologies, Inc. (“Tech”) is a San Francisco-based corporation (though it’s incorporated in Delaware) that puts out a well-known software app allowing people to “get a ride on demand from a nearby driver registered with the UBER [Technologies] service.” ECF No. 63, at 2. Tech refers to these drivers — who use their own cars to transport passengers — as “driver partners.” ECF No. 37, at 1-3 ¶¶ 2-4; 10-13 ¶¶ 13-16. Tech -launched its service in 2010 as UBERCAB and received a federal registration for that trademark on August 31 of that year. ECF No. 41-1, at 2. Later in 2010, Tech began using the mark “UBER” and registered that trademark on June 14, 2011. Id. at 10.1
At first, Tech operated in just a few select cities, but it has expanded rapidly *1260over the past few years.2 Tech first operated in Florida during the 2012 Republican National Convention in Tampa. ECF No. 37, at 16-17 ¶¶ 22-23. Tech’s first permanent entry into the Florida market came in November 2013 when it launched Uber-BLACK3 in Jacksonville. ECF No. 63, at 161. Tech expanded further into Florida throughout 2014; in August of that year, Tech started operating in Gainesville. Id. at 161-62.
Apparently alarmed by Tech’s expansion into Florida — and in particular by the use of “Uber Promotions” on one of Tech’s Twitter accounts — Promotions sent Tech a letter in April 2014 demanding that Tech “discontinue the use of the terms ‘Uber Promotions’ in connection with [Tech’s] marketing and advertising campaigns ... and undertake in writing that [Tech] will not at any time in the future use any of Uber Promotions name [sic] in any future marking [sic] or advertising campaigns, Twitter Accounts, or apply for registration of any trademarks/service mark that may be confusingly similar to Uber Promotions.” ECF No. 27-21, at 2-3. Tech responded by (1) removing the term “Uber” from the title of a page on its website that had previously been titled “Uber Promotions” and (2) responding (through counsel) to Promotions’ letter as follows:
Re: UBER PROMOTIONS
Thank you for your April 18th letter.
Uber disagrees that there is any potential for confusion regarding its use of the “Uber Promotions” heading on Twitter. However, to resolve the matter amicably, and without admission, Uber has removed the “Uber Promotions” reference.
Very truly yours,
ECF No. 40, at 3-4 ¶¶ 7-10; ECF No. 40-1. Tech apparently did not hear from Promotions again until this suit was filed. ECF No. 40, at 4 ¶ 11.
Shortly after this suit was filed, Tech unveiled a new service called UberE-VENTS. ECF No. 37, at 19 ¶ 31. This service allows customers to purchase rides for others that can be used at a particular time in the future. Id. at 19-21 ¶¶ 31-32. So, if a customer is hosting a large birthday party at a local park, that customer can purchase rides in advance with Tech driver partners for guests to use to go from their homes to the park and back again. Technically, the customer purchases a single code or a set of codes which are then sent to guests; the guests then use Tech’s app as they normally would to request rides and use the code to pay for those rides. Id. This service “has been a popular option for corporate events.” Id. A customer wishing to purchase rides for their guests can do so through Tech’s website. Id.
*1261In this suit, Promotions brings various claims against Tech under the Lanham Act and Florida state law alleging trademark4 infringement and unfair competition. ECF No. 47. Promotions seeks preliminary in-junctive relief to keep Tech from continuing to use its “UBER” marks in connection with its services throughout Florida. ECF No. 27. This order partially grants and otherwise denies the motion for a preliminary injunction.
II. PRELIMINARY INJUNCTION STANDARD
A. Basic Standard
“A district court may grant in-junctive relief only if the moving party shows that: (1) it has a substantial likelihood of success on the merits; (2) irreparable injury will be suffered unless the injunction issues; (3) the threatened injury to the movant outweighs whatever damage the proposed injunction may cause the opposing party; and (4) if issued, the injunction would not be adverse to the public interest.” Siegel v. LePore, 234 F.3d 1163, 1176 (11th Cir.2000) (en banc). A “preliminary injunction is an extraordinary and drastic remedy not to be granted unless the movant clearly carries the burden of persuasion as to the four prerequisites.” United States v. Jefferson County, 720 F.2d 1511, 1519 (11th Cir.1983) (quotation and citation omitted).
B. Preliminary Injunctions in Trademark Cases
In a trademark infringement case, there is an extra wrinkle to the preliminary injunction standard. For years, the Eleventh Circuit (along with many other circuits) held that courts should “extend a presumption of irreparable harm once a plaintiff establishes a likelihood of success on the merits of a trademark infringement claim.” N. Am. Med. Corp. v. Axiom Worldwide, Inc., 522 F.3d 1211, 1227 (11th Cir.2008). The appropriateness of this presumption is doubtful, however, after the Supreme Court’s decision in eBay Inc. v. MercExchange, LLC, 547 U.S. 388, 126 S.Ct. 1837, 164 L.Ed.2d 641 (2006). There the Court held that a similar rule in patent cases was wrong, and that “the traditional four-factor framework that governs the award of injunctive relief’ should apply in patent cases. 547 U.S. at 394, 126 S.Ct. 1837.
Although eBay was a patent case and not a trademark case, and although it concerned a permanent rather than a preliminary injunction, courts have applied its reasoning in the context of preliminary injunctions in trademark cases and concluded that the presumption of irreparable harm is no longer good law. See Ferring Pharm., Inc. v. Watson Pharm., Inc., 765 F.3d 205, 216 (3d Cir.2014); see also Herb Reed Enters., LLC v. Fla. Entm’t Mgmt., Inc., 736 F.3d 1239, 1249 (9th Cir.2013). Other courts, while not going so far as to hold that the presumption of irreparable harm is no longer appropriate in trademark cases, have suggested that that is the correct (and inevitable, once the question is squarely presented) result. See, e.g., Salinger v. Colting, 607 F.3d 68, 78 n.7 (2d Cir.2010).
The Eleventh Circuit is in this latter group of courts, having stated in dicta that “a strong case can be made that eBay’s holding necessarily extends to the grant of preliminary injunctions” in a trademark-*1262infringement case. Axiom Worldwide, 522 F.3d at 1228. District courts in the Eleventh Circuit have dealt with Axiom Worldwide and eBay in a variety of ways, but have mostly declined to decide the question. See Sylvan Learning Inc. v. Learning Sols., Inc., 795 F.Supp.2d 1284, 1295-99 (S.D.Ala.2011) (discussing the approaches taken by various courts in the Eleventh Circuit).
This Court is persuaded by eBay that a presumption of irreparable harm is no longer appropriate in a trademark-infringement case once a substantial likelihood of success on the merits is shown. The Supreme Court made clear in eBay that “a major departure from the long tradition of equity practice should not be lightly implied,” and that the Patent Act did not evince an intent on the part of Congress to depart from that tradition. eBay, 547 U.S. at 391, 126 S.Ct. 1837. The Lanham Act, like the Patent Act, provides that courts “shall have power to grant injunctions! ] according to the principles of equity.” 15 U.S.C. § 1116(a) (emphasis added). This does not suggest that irreparable harm should be presumed in a trademark infringement case, but instead indicates that the appropriateness of in-junctive relief must be analyzed using the traditional framework. See Ferring Pharm., 765 F.3d at 215-17.
That said, it is true that in trademark cases, just as in patent cases, “[w]hen it comes to discerning and applying th[e] [usual preliminary injunction] standards, ... ‘a page of history is worth a volume of logic.’ ” eBay, 547 U.S. at 395, 126 S.Ct. 1837 (Roberts, C.J., concurring) (quoting N.Y. Trust Co. v. Eisner, 256 U.S. 345, 349, 41 S.Ct. 506, 65 L.Ed. 963 (1921) (Holmes, J.)). In other words, while it was perhaps inappropriate in the past to presume irreparable harm, that practice was grounded upon the sound principle that the harm associated with trademark infringement is typically irreparable in nature. See, e.g., Processed Plastic Co. v. Warner Commc’ns, Inc., 675 F.2d 852, 858 (7th Cir.1982). So while a court must, in each trademark infringement case, make a finding of irreparable harm before an injunction may issue, that finding will often be made due to the nature of the harm.
C. Discretion to Tailor Relief
“The essence of equity jurisdiction has been the power of the [court] to do equity and to mould each decree to the necessities of the particular case. Flexibility rather than rigidity has distinguished' it.” Weinberger v. Romero-Barcelo, 456 U.S. 305, 312, 102 S.Ct. 1798, 72 L.Ed.2d 91 (1982) (quoting Hecht Co. v. Bowles, 321 U.S. 321, 329, 64 S.Ct. 587, 88 L.Ed. 754 (1944)). This Court has the power, and indeed the obligation, to exercise its authority in a manner that achieves a “ ‘nice adjustment and reconciliation’ between the competing claims” of Promotions and Tech. See id. (quoting Hecht Co., 321 U.S. at 329, 64 S.Ct. 587). This is not an all-or-nothing affair, and even upon a finding of substantial likelihood of success on the merits this Court has significant discretion to tailor relief so as to best serve the interests of the parties and the public.
III. LIKELIHOOD OF SUCCESS ON THE MERITS
Promotions brings a number of claims, but they all require Promotions to establish more or less the same two things: “(1) that [it] had enforceable ... rights in [its] mark or name, and (2) that [Tech] made unauthorized use of it ‘such that consumers were likely to confuse the two.’ ” Custom Mfg. & Eng’g, Inc. v. Midway Servs., Inc., 508 F.3d 641, 647-48 (11th Cir.2007) (quoting Lone Star Steakhouse & Saloon, Inc. v. Longhorn Steaks, Inc., 106 F.3d 355, 358 (11th Cir.1997)); see *1263also PetMed Express, Inc. v. MedPets.com, Inc., 336 F.Supp.2d 1213, 1217-19 (S.D.Fla.2004) (setting out elements of federal trademark infringement, Florida common-law trademark infringement, federal unfair competition, and Florida common-law unfair competition).
A. Effect of Tech’s Federal Registration
1. Effective Date
Tech registered “UBERCAB” in 2010 and “UBER” in 2011. It argues that its registration of the “UBER” mark should “tack back” to the earlier registration date of “UBERCAB” because the two marks “create the same, continuing commercial impression.” ECF No. 63, at 10. Although this is a question of fact, see Hana Financial, Inc. v. Hana Bank, - U.S. -, 135 S.Ct. 907, 912, 190 L.Ed.2d 800 (2015), this Court is confident that Tech would be able to convince a jury that UBER and UBERCAB would be “considered the same mark” by a reasonable consumer, see id. at 910. Accordingly, for purposes of this preliminary injunction proceeding, Tech’s “UBER” mark will be deemed to have been (effectively) registered federally on August 31, 2010.
2. Constructive Use
Registration of a trademark with the United States Patent and Trademark Office constitutes “constructive notice of the registrant’s claim of ownership thereof.” 15 U.S.C. § 1072. Furthermore, “federal registration ... constitutes constructive use of the mark.” Allard Enters., Inc. v. Advanced Programming Res., Inc., 249 F.3d 564, 572 (6th Cir.2001); 15 U.S.C. § 1057(c). In this case, Tech is considered to have used the mark “UBER” everywhere in the United States as of August 31, 2010 because of its registration of that mark.
B. Geographical Scope of Promotions’ Activities
1. Zone of Future Expansion
“[T]he scope of protection accorded [a common-law] mark is coextensive only with the territory throughout which it is known and from which it has drawn its trade.” Tana v. Dantanna’s, 611 F.3d 767, 780 (11th Cir.2010). In addition, a mark may have protection in “a zone of reasonable future expansion” beyond that territory. Tally-Ho, Inc. v. Coast Cmty. Coll. Dist., 889 F.2d 1018, 1027 (11th Cir. 1989). “The extent of the zone is generally considered as of the date of the junior user’s first use.” Id. at 1028.
The doctrine of future expansion is clearly applicable when the junior user is the holder of a common-law trademark, but things are perhaps different when the junior user is the holder of a federally-registered trademark and its “use” is constructive use under 15 U.S.C. § 1057(e). The Eleventh Circuit has noted that “federal registration has the practical effect of freezing a prior user’s enforceable trademark rights thereby terminating any right to future expansion beyond the [senior] user’s existing territory.” Dantanna’s, 611 F.3d at 780-81 (emphasis added). This statement seems to indicate that the future expansion doctrine doesn’t apply when the junior user acquires rights through registration rather than through actual use. But the court cited for this principle the Sixth Circuit case Allard Enterprises, which actually held that the future expansion doctrine is applicable in such a case. Allard Enters., Inc., 249 F.3d at 573-75.
Another issue is what type of constructive use can be ascribed to Tech. Its trademark registrations are for “computer software for coordinating transportation services,” ECF No. 41-1, at 2, 10, so perhaps it can only claim constructive use of its marks in connection with its app, and *1264not in connection with the transportation services its app facilitates. In that case, Tech’s first use of its marks in connection with the actual provision of transportation services in Florida wouldn’t occur until it actually used those marks — that is, until people could actually use Tech’s app to get a ride. On the other hand, the constructive use provision states that registration “constitute[s] constructive use of the mark, conferring a right of priority, nationwide in effect, on or in connection with the goods or services specified in the registration.” 15 U.S.C. § 1057(c) (emphasis added). This suggests that Tech should be deemed as having used its marks in Florida as of the date of registration of those marks in the same manner as it is using them now.
It is not necessary to resolve these issues. Even assuming that Promotions is entitled to a “zone of reasonable future expansion” beyond its core trade area in Gainesville, Promotions has not established that it would have expanded in any way prior to the relevant date. Whatever date is used — August 31, 2010, the date of Tech’s earliest registration; June 14, 2011, the date of Tech’s registration of the “UBER” mark; or even mid-2014, when Tech expanded throughout Florida — there is scant evidence that Promotions was at any point on the verge of expanding its , operations throughout Florida. While it is true that Promotions has transported passengers or arranged for the transportation of passengers to many cities in Florida (and some places outside of Florida),5 that does not mean that it is entitled to expand into each and every one of those places and all points in between. Considering (1) Promotions’ actual trade zone, (2) its failure to expand that trade zone, and (3) the relative rarity of its forays beyond that trade zone, it seems highly unlikely that Promotions would succeed in showing that it is entitled to the entire state of Florida — or anything beyond the Gainesville area — as a “zone of reasonable future expansion.” See Tally-Ho, Inc., 889 F.2d at 1027-28. (identifying as key factors in the “zone of reasonable future expansion” inquiry the size of the senior user’s “actual market penetration or reputation” and “the history of the senior user’s past expansion”).
2. Promotions’ Territory
Promotions has mostly served customers in the Gainesville area. Its relatively infrequent forays beyond that area are not enough to make its marks “known” and thus protectable against Tech. This is because “the geographic area in which an unregistered trademark is ‘in use’ is defined as the area.in which the use of [a] ' similar mark would create a likelihood of confusion.” Dorpan, S.L. v. Hotel Meliá, Inc., 728 F.3d 55, 63-64 (1st Cir.2013). It is difficult to believe that Promotions would be able to show that its infrequent trips to Jacksonville, Tampa, etc. were so memorable that the local citizenry was likely to be confused when Tech later moved into town. The one possible exception is Promotions’ bus service to Ocala Poker and Jai Alai in Reddick, Florida. The record is not entirely clear on this point, but it appears that Promotions has been providing service there (perhaps on and off) for many years, and certainly since before 2010. ECF No. 27-5. This might be the only area outside of Alachua County in which Promotions would be able to establish enforceable common-law trademark rights.
*1265As far as Gainesville (and surrounding areas) goes, Promotions is substantially likely to succeed in showing that it has enforceable rights in the “UBER PROMOTIONS” mark (and perhaps “UBER,” too, though whether it does or not wouldn’t change anything). Tech argues, somewhat half-heartedly, that Promotions has not put forth enough evidence of its “market penetration” prior to Tech’s federal registration of its marks, ECF No. 63, at 20-22, but it seems clear that Promotions has been conducting business in the Gainesville area since 2006 and has been using its marks during that entire time, see, e.g., ECF No. 27-5. It certainly seems as if Promotions was sufficiently “known” in the Gainesville area prior to 2010 to have enforceable common-law trademark rights.
C. Likelihood of Confusion in the Gainesville Area
1. Type of Confusion?
Promotions argues that both forward and reverse confusion are occurring and are likely to occur, but admits that reverse confusion seems like a much larger threat. ECF No. 71, at 83-84. That is an understatement. This is, in many ways, a classic reverse confusion case, cf. A&H Sportswear, Inc. v. Victoria’s Secret Stores, Inc., 237 F.3d 198, 228 (3d Cir. 2000), and there is scant evidence that would support a finding that Promotions is substantially likely to prevail on anything other than a reverse confusion theory.6
“[R]everse confusion occurs when ‘the junior user saturates the market with a similar trademark and overwhelms the senior user.’ ” Id. (quoting Fisons Horticulture, Inc. v. Vigoro Indus., Inc., 30 F.3d 466, 475 (3d Cir.1994)). “The harm flowing from reverse confusion is that ‘[t]he public comes to assume the sénior user’s products are really the junior user’s or that the former has become somehow connected to the latter. ... [T]he senior user loses the value of the trademark — its product identity, corporate identity, control over its goodwill and reputation, and ability to move into new markets.’ ” Id. (quoting Ameritech, Inc. v. Am. Info. Tech. Corp., 811 F.2d 960, 964 (6th Cir.1987)). “As in a direct confusion claim, the ultimate question in a reverse confusion claim is whether there is a likelihood of consumer confusion as to the source or sponsorship of a product.” Id. at 229.
2. Factors to Consider
Every circuit uses some %-factor test for likelihood of confusion, where n is an integer that is almost certainly too large for the task at hand.7 The Eleventh Circuit uses a seven-factor test. See Dantanna’s, 611 F.3d at 774-75. None of the factors is dispositive, Carillon Importers Ltd. v. Frank Pesce Group, Inc., 913 F.Supp. 1559, 1564 (S.D.Fla.1996), though some are more important than others, see Dantanna’s, 611 F.3d at 779.
Before jumping in, it is important to note four things. First, “there is nothing magical about these ‘factors’; they operate only as a heuristic device to assist *1266in determining whether confusion exists.” Sullivan v. CBS Corp., 385 F.3d 772, 778 (7th Cir.2004). The multi-factor “test” is not really a test at all, but rather a set of things to consider, or questions to ask, en route to the ultimate question: is there “a likelihood of confusion in the mind[s] of an appreciable number of reasonably prudent buyers[?]” John H. Harland Co. v. Clarke Checks, Inc., 711 F.2d 966, 979 n.22 (11th Cir.1983) (citations and internal quotations omitted).
Second, it is vital to always keep in mind the “actual market conditions and the type of confusion alleged” when conducting the analysis. Malletier v. Burlington Coat Factory Warehouse Corp., 426 F.3d 532, 534 (2d Cir.2005) (Calabresi, J.). This is because “ ‘real world’ confusion is the confusion that the [Lanham] Act seeks to eliminate,” and the existence of such confusion is best assessed when marks are compared “in light of the way in which the marks are actually displayed” or perceived by consumers. Id. at 538-39. Here, reverse confusion is at issue, which means the “market conditions” we should be concerned with are the conditions under which a consumer encounters the senior user’s (Promotions’) mark.
Third, it is also vital to keep in mind “the persons who are (or are not) likely to be confused.” See Clarke Checks, 711 F.2d at 979 n.22. This is a reverse confusion case, which means that the largest class of people who might be confused are people familiar with Tech’s marks who then encounter Promotions’ marks. See Banff, Ltd. v. Federated Dep’t Stores, Inc., 841 F.2d 486, 490 (2d Cir.1988) (tying reverse confusion to “consumers initially aware of [the junior user’s mark who] may believe that [the senior user’s] mark they later encounter originates with [the junior user]”). Certainly confusion could occur with consumers initially aware of Promotions’ mark who then encounter Tech’s mark — this has apparently actually happened, in fact — but this is less likely given the relative commercial strength of the companies and their marks. Promotions primarily serves and advertises to college students and others in the same age range — young people, roughly speaking. See, e.g., ECF No. 27-1, at 8-9 ¶¶ 19-22; ECF No. 27-5; ECF No. 36-1, at 43. Thus, when assessing likelihood of confusion, the, key demographic to keep in mind is young people, particularly college students, in the Gainesville area.8
Fourth, when an alleged infringer’s marks are used in connection with multiple goods or services, it may be that the use of those marks in connection with some goods or services — but not others— creates a likelihood of consumer confusion. See, e.g., H. Lubovsky, Inc. v. Esprit de Corp., 627 F.Supp. 483, 492-93 (S.D.N.Y. 1986) (finding no likelihood of confusion as to the defendant’s use of its marks on sportswear, but a likelihood of confusion as to the use of its marks on shoes). In this case, Tech uses its “UBER” mark in connection with a number of services: UberX, UberXL, UberBLACK, UberEVENTS, and so on. See ECF No. 37, at 7-8 ¶ 9. As explained more fully below in Part III. C.ll, the UberEVENTS service is different from Tech’s other services in that it is *1267more closely related to Promotions’ transportation services. For that reason, it will be treated separately. In what follows, the seven factors are analyzed without taking into account Tech’s UberEVENTS service.
3. Strength of the Mark
The first factor to be considered is the strength of Promotions’ marks. Dantan-na’s, 611 F.3d at 774-75. However, because this is a reverse-confusion case, it is the conceptual and not the commercial strength of Promotions’ marks that matters. See A&H Sportswear, 237 F.3d at 231-32. The parties disagree on this point, with Promotions claiming that “UBER” and “ÜBER” are arbitrary or fanciful in nature, and thus “strong” marks, ECF No. 27, at 19, and Tech claiming that they are descriptive (at least when combined with “PROMOTIONS”) and thus weaker, ECF No. 63, at 26-27. Furthermore, Tech points out that many companies have registered marks with “UBER” in them, thus reducing the conceptual strength of Promotions’ marks. ECF No. 63, at 27.
It seems clear that Promotions’ use of “UBER” by itself is at least suggestive, if not arbitrary. Although Promotions’ founders may have chosen the name of their company because it means “the best,” id. at 26, their motivation or intent behind picking the name is irrelevant; what matters is what impression the mark leaves on consumers. The difference between a descriptive mark and a suggestive mark is that the former “identifies a characteristic or quality of an article or service” while the latter “suggests ... a characteristic of the goods or services and requires an effort of the imagination by the consumer in order to be understood as descriptive.” Vision Ctr. v. Opticks, Inc., 596 F.2d 111, 115 (5th Cir.1979).9 Put another way, if competitors “would be likely to need the term[ ] used in the trademark in describing their products,” the mark is likely descriptive. See id. (quoting Union Carbide Corp. v. Ever-Ready, Inc., 531 F.2d 366, 379 (7th Cir.1976)). Clearly “UBER” by itself is not descriptive under this test.
“UBER PROMOTIONS” or “UBER PROMOTIONS” is a closer call, not least because of the “self-laudatory mark” principle. This principle is simply the idea that, generally speaking, a term such as “best” or “super” that “extol[s] some feature or attribute of ... goods or services” is descriptive in nature and likely to be too weak to be trademarked. 2 McCarthy on Trademarks and Unfair Competition § 11:17, Westlaw (4th ed., database updated Dec. 2015) [hereinafter McCarthy]. The German term “fiber” means “above” or “over” or “beyond,” and can also be combined with other German terms to form compound nouns such as “Ubermensch” (“Superman”) and adverbs such as “fiberaus” (“extremely” or “exceedingly”).10 The English term “fiber” or “uber” — which of course is just an Anglicized version of “fiber”- — -means “being a superlative example of its kind or class.”11 Tech argues that “ÜBER PROMOTIONS” is “conceptually weak” because, like “Super Promotions” or “Best Promotions,” it is self-laudatory and descriptive. ECF No. 63, at 26-27.
*1268This argument is not very convincing. A term such as “uber” “may slide along the continuum between suggestiveness and descriptiveness depending on usage, context, and other factors that affect the relevant public’s perception of the term.” In re Nett Designs, Inc., 236 F.3d 1339, 1341 (Fed.Cir.2001) (discussing the term “ultimate”). “Uber,” while not an extremely uncommon term, is less frequently encountered — and thus less likely to be perceived by the relevant public as merely self-laudatory — than a term like “super” or even “ultimate.” Simply put, Promotions would likely be able to convince a factfin-der that “ÜBER PROMOTIONS” is more suggestive than descriptive, and deserving of trademark protection even without any secondary meaning. That said, “UBER” and especially “UBER PROMOTIONS” are not especially strong marks conceptually, and so this factor tips in favor of Tech — that is, slightly against a finding of likelihood of confusion.
A Similarity of the Marks
The second factor is the similarity of Promotions’ and Tech’s marks. Dantanna’s, 611 F.3d at 775. In assessing similarity, it is important to view the marks as an ordinary consumer would, and not in an artificial context. See Sun-Maid Raisin Growers of Cal. v. Sunaid Food Prods., Inc., 356 F.2d 467, 469 (5th Cir.1966); see also A&H Sportswear, 237 F.3d at 216 (noting that when comparing marks, “an effort must be made to move into the mind of the roving consumer”). This means, among other things, remembering that memory is imperfect, and that a consumer encountering one mark after exposure to another mark may be confused even when the marks differ greatly in some important respects. See, e.g., Northam Warren Corp. v. Univ. Cosmetic Co., 18 F.2d 774, 775 (7th Cir.1927) (“Whether there is an infringement of a trade-mark does not depend upon the use of identical words, nor on the question as to whether they are so similar that a person looking at one would be deceived into the belief that it was the other; but it is sufficient if one adopts a trade-name or a trade-mark so like another in form, spelling, or sound that one, with a not very definite or clear recollection as to the real trade-mark, is likely to become confused or misled.”).
Promotions can only hope to prevail on a reverse-confusion theory, so our typical consumer is someone who is familiar with Tech and who then encounters Promotions. But there are many classes of consumers “familiar with” Tech — there are those who have used Tech’s services by accessing them through Tech’s app; those who have perhaps been along for an Uber ride with someone else who but who have not themselves downloaded or used Tech’s app; and those who have heard of Tech and have a general idea of what services it provides, but who have never used Tech’s app or been along for a ride with a Tech driver partner. And there are many ways in which a consumer might “encounter” Promotions — through its website; by seeing a flyer; through its Facebook page; or by finding it through a Google search. In assessing the similarity of the marks, each type of familiarity and each type of encounter must be kept in mind.
Tech is probably correct that for Tech users/customers, the visual differences between Tech’s app and Promotions’ logo, together with the fact that Promotions usually includes “Promotions” in its name, suggest that confusion is relatively unlikely. EOF No. 63, at 28-29. But the visual differences are much less significant for those consumers less familiar with Tech. The examples of Tech’s advertising and promotional materials provided in the record suggest that the most important feature of Tech’s mark is the word “Uber” *1269and not any particular design element.12 See, e.g., ECF No. 27-18, at 2; ECF No. 27-20, at 4; ECF No. 27-21, at 6-9. Moreover, many consumers may have just heard of Tech or read a news story referencing it, in which case the visual differences are quite unimportant. Similarly, the use of “Promotions” by Promotions is not very important in assessing similarity in the eyes of consumers relatively less familiar with Tech. Those consumers will simply remember “Uber” and will recognize that relatively uncommon word in Promotions’ advertisements. Cf. A&H Sportswear, 237 F.3d at 216 (“the general rule that marks should be viewed in their entirety does not undermine the common-sense precept that the more forceful and distinctive aspects of a mark should be given more weight [in assessing mark similarity], and the other aspects less weight”). In addition, Tech' often uses “Uber Promotions” in connection with various promotional campaigns, which could easily create confusion. ECF No. 27-10.
In sum, Tech’s and Promotions’ marks are fairly similar, at least to a certain portion of the public. Tech may be quite correct that “anyone who uses [Tech’s] services and comes across [Promotions’] advertising will know the difference,” ECF No. 63, at 29, but there are many who have come across Tech but have not used its services. For that group, the marks may be quite similar. Cf. Stork Restaurant v. Sahati, 166 F.2d 348, 358-59 (9th Cir.1948) (noting that while a “worldly-wise passerby would not be ... deceived” into thinking that “an unpretentious night club displaying the sign ‘Stork Club’ in or near San Francisco ... was in any way affiliated with the celebrated New York establishment,” “[t]he law ... protects not only the intelligent, the experienced, and the astute, ... [but] also the ignorant, the inexperienced, and the gullible”). This factor tilts slightly in favor of Promotions.
5. Similarity of Services
The third factor is the “similarity between the goods and services offered under the two marks.” Dantanna’s, 611 F.3d at 775. Because the ultimate “likelihood of confusion” question is whether consumers may come to believe that Promotions’ services are in some way affiliated with Tech’s services, see Axiom Worldwide, 522 F.3d at 1222 n.9, it is not necessary that the services offered by the two companies be precisely the same, or even that they be competitive with one another; rather, confusion may result where the services are so related that a single entity is likely to be behind both. See E. Remy Martin & Co., S.A. v. Shaw-Ross Int’l Imports, Inc., 756 F.2d 1525, 1530 (11th Cir.1985).
Despite the fact that both Tech’s and Promotions’ services involve passenger transportation, there is no doubt that the services differ in many important respects. Tech’s services all involve driver partners using their own vehicles to transport customers. Promotions, in contrast, arranges *1270for or provides transportation via more traditional passenger transportation vehicles such as charter buses, party buses, and limos. Tech’s focus is on the transportation of individuals or small groups, whereas Promotions focuses more on group transportation. Farzad Dep. 247-48.
These differences are not so great that they weigh heavily against confusion, particularly in light of Tech’s expanding roster of transportation options. Around August 2015, for instance, Tech introduced UberXL to the Gainesville market. ECF No. 37-1, at 144. This option allows customers to request larger cars such as SUV's or minivans. Id. UberPOOL, a service that allows customers traveling in similar directions to share a car and save money, is not yet available in Gainesville, but it very well could be soon. ECF No. 37, at 18 ¶ 27. This growing family of Tech services and options makes it more likely that a consumer might mistake Pro-, motions’ services as being in some way affiliated with Tech.
Those more familiar with Tech will likely be saved from confusion by the fact that Promotions’ services lack the most salient (or perhaps infamous) feature of Tech’s services: the use of driver partners, many of whom are moonlighting as drivers, ECF No. 63, at 159, using their own vehicles as taxis. On the other hand, those who only know that Tech has something to do with transportation or that it provides something like a taxi service could very well think that Promotions and Tech are affiliated based on Promotions’ services. Ultimately, this factor doesn’t weigh heavily in favor of either party, though it tilts slightly in favor of Tech.
6. Similarity of Sales Methods
The fourth factor is the “similarity of the actual sales methods used by the holders of the marks, such as their sales outlets and customer base.” Dantanna’s, 611 F.3d at 775. “This factor takes into consideration where, how, and to whom the parties’ products [or services] are sold.” Frehling Enters., Inc. v. Int’l Select Grp., Inc., 192 F.3d 1330, 1339 (11th Cir.1999).
This factor weighs heavily in favor of Tech. Tech’s services are accessed through an app on a smartphone or other mobile device, whereas one must call, e-mail, text, or send a Facebook message to Promotions in order to set up a ride. Farzad Dep. at 183. Anyone who has used Tech’s services has necessarily downloaded and used its app, and it seems relatively unlikely that such users would think that Promotions was affiliated with Tech. And even people who have only heard of Tech may very well know that its services have something to do with an app, and may therefore be less likely to be confused when confronted with Promotions’ services, which are accessed via quite different mechanisms.
7. Similarity of Advertising Methods
The fifth factor is the “similarity of advertising methods” used by Tech and Promotions. Dantanna’s, 611 F.3d at 775. Both Tech and Promotions make extensive use of social media to advertise. See, e.g., ECF Nos. 27-10, 36-1; Gore Dep. 268-71. In fact, there is evidence that some of Tech’s “affiliates” — people who advertise Tech’s services via social media in exchange for a commission — have placed advertisements for Tech’s services on Promotions’ Facebook page. ECF No. 36-1. While Tech is correct that the look of Promotions’ advertisements is quite different from the look of Tech’s advertisements, ECF No. 63, at 33, the fact remains that both companies’ extensive use of social media — and especially Tech’s affiliates’ use of Promotions’ own Facebook page to post advertisements — raises the likelihood that consumers will come into contact with both marks, which in turn raises the likeli*1271hood of confusion. See Frehling Enters., 192 F.3d at 1340/
Furthermore, Tech has apparently used Google AdWords to ensure that certain searches conducted using Google will lead to an ad for Tech being displayed above the search results.13 ECF No. 27-1, at 13-14 ¶ 33. Tech is of course allowed to advertise, but advertising in this way will increase the likelihood of consumer confusion. Overall, this factor weighs slightly in favor of Promotions.
8. Intent to Infringe
The sixth factor is the “intent of the alleged infringer to misappropriate the proprietor’s good will.” Dantanna’s, 611 F.3d at 775. This factor needs to be tweaked in a reverse-confusion case, since the concern is not that an alleged infringer will try to use a plaintiffs good name to gain business, but rather that it will “push[ ] its [smaller] rival out of the market,” thus “usurping] [the rival’s] business identity.” A&H Sportswear, 237 F.3d at 232. Furthermore, only intent to “usurp identity” through confusion is relevant; there is nothing unlawful about intending to compete. Id.
It is difficult to believe that Tech acted with the intent to push Promotions out of the Gainesville market by using a confusingly similar mark. Elephants don’t look out for gerbils when they plow through the bush. Tech adopted its mark before it ever knew of Promotions and expanded nationally thereafter. Surely it knew that it would collide with Promotions once it entered Gainesville, but that doesn’t mean it intended to confuse consumers and thereby hurt Promotions. This factor is not particularly relevant to the analysis in this case.
9. Actual Confusion
The final factor is “the existence and extent of actual confusion in the consuming public.” Dantanna’s, 611 F.3d at 775. Evidence of actual confusion “is the most persuasive evidence in assessing likelihood of confusion.” Id. at 779. Clearly understanding this, the parties vigorously dispute how to interpret the numerous phone calls received by Promotions from customers looking for Tech. Promotions claims that these phone calls constitute “extraordinary” evidence of actual confusion, ECF No. 27, at 26, while Tech argues that these calls are the result of “inattention and indifference,” not confusion. ECF No. 63, at 34-37.
Each party overstates its case. Undoubtedly some of the people who called Promotions looking for Tech were indeed “confused” within the meaning of the Lan-ham Act — that is, they understood they were calling Promotions and thought Promotions was affiliated with Tech. But Tech doesn’t have an easy-to-find phone number,14 so it is likely that many of the callers thought they were calling Tech. This is supported by the large number of callers inquiring about becoming driver partners, see ECF No. 27-15, at 2-8 — a key feature of Tech’s business model completely missing from Promotions’ business model. This theory is further bolstered by the fact that if one searches “Uber Gainesville phone number” or “Uber Gainesville phone” using Google, it is Promotions’ phone number that comes up. Given what’s in the record, it’s likely that a large fraction of the callers didn’t even grasp that they were calling something called Uber Promotions, but instead thought they were *1272calling Uber Technologies. This is akin to being given a wrong number by a telephone operator.15
If this is confusion at all, it is not the type of confusion that’s important in a trademark-infringement case.16 “Not all conceivable forms of confusion are relevant to trademark infringement. Confusion is relevant when it exists in the minds of persons in a position to influence the purchasing decision or persons whose confusion presents a significant risk to the sales, goodwill, or reputation of the trademark owner.” Hotel Meliá, 728 F.3d at 64-65 (quotations and citations omitted). A person who does not even grasp that he is talking to Promotions on the phone, or that Promotions even exists separate and apart from Tech, is obviously not going to form a bad opinion of Promotions based on his “confusion,” nor is Promotions going to lose any business. And that’s assuming that the caller can even be termed “confused” — as Tech points out, courts in similar factual situations have suggested that such misdirected communications can be attributed to inattentiveness or carelessness. See Therma-Scan, Inc. v. Thermoscan, Inc., 295 F.3d 623, 636 (6th Cir.2002).
The e-mails received by Promotions from customers looking for Tech are slightly more helpful to Promotions’ case, but some of them again suggest carelessness as much as, if not more than, confusion. The e-mail from a customer in Vietnam is clearly intended for Tech — the e-mailer followed up with an “oops” e-mail almost immediately. ECF No. 27-16, at 9-10. The Los Angeles customer appears to be someone so easily confused that even trademark law cannot protect her. Id. at 6-8.
The two instances of confusion on the part of former customers of Promotions are more helpful to Promotions’ case. These customers clearly understood what Promotions is, understood what Tech is, and thought they were affiliated. Farzad Dep. 127-30; Friedman Dep. 140. Similarly, the evidence in the record of a few Tech customers — and even some driver partners and others affiliated with Tech — thinking that Tech and Promotions were somehow related is evidence of actual confusion within the meaning of the Lanham Act.17 *1273See ECF No. 27-22, at 5; Gore Dep. Ex. 4, at 3, 5-6,12-13.
In sum, there is definitely evidence of actual confusion in the record, though not as much as Promotions claims or as little as Tech claims. The evidence of customers calling and e-mailing Promotions looking for Tech is in many ways the least important subset of this evidence. The types of requests being made by the callers, the fact that Tech doesn’t have an easy-to-find phone number, the fact that someone conducting a Google search for a local Gaines-ville number for Tech would get Promotions’ number, the relative commercial strength of the two companies — all of these facts strongly suggest that most of the callers were not confused within the meaning of the Lanham Act, but were either misdirected, careless, or both. As for the e-mails, there are only five of them presented in the record, at least two of which could easily be chalked up to carelessness. Furthermore, many of the callers and e-mailers who were confused were likely quickly put straight, and “[s]hort-lived confusion or confusion of individuals casually acquainted with a business is worthy of little weight.” Safeway Stores, Inc. v. Safeway Discount Drugs, Inc., 675 F.2d 1160, 1167 (11th Cir.1982). That said, it strains credulity to think that at least some of the many dozens of callers were not actually confused, and certainly two of the e-mailers were. The evidence of confusion on the part of former Promotions customers and current Tech customers and driver partners is worth more weight than the phone call and e-mail evidence, see id. but there are only a few such instances. Still, even a few instances of actual confusion on the part of customers is “worthy of some consideration.” Id. Putting all of this together, the actual confusion factor weighs in favor of Promotions, but it by no means weighs so heavily that a finding of a likelihood of confusion is inevitable.
10. Likelihood of Confusion?
The “intent to infringe” factor is unimportant, the “strength of the mark” factor does not really favor a finding of confusion, and the “similarity of marks” and the “similarity of advertising methods” factors together favor a finding of confusion. The remaining three factors — actual confusion, similarity of services, and similarity of sales methods — deserve more discussion.
The one factor that weighs most heavily against Promotions is the dissimilarity in the sales methods of the two companies. Tech’s services are accessed through a smartphone app.18 One cannot call for a ride with one of Tech’s driver partners, nor can one go to a website and reserve a ride. This is in stark contrast to the manner in which Promotions’ services are accessed. This difference is critically important for the following reason: anyone who has used Tech’s services necessarily understands that those services are accessed via a smartphone app, and is unlikely to think that Tech has all of a sudden started to make its services accessible via other means. The analysis of the “similarity of services” factor (see Part III.C.5 above) also suggests that people who have used Tech’s services are unlikely to be confused when they encounter Promotions’ marks.
What this means is that the group of people who might be likely to be confused are people who have heard of, but not used, Tech’s services. But the people who *1274matter most in this group are those who will actually encounter Promotions’ mark, and that subgroup is likely largely comprised of college-aged (and slightly older) people. This subgroup is relatively tech-sawy19 and therefore likely to be sensitive to the differences between the sales methods of Tech and Promotions, even if all they know about Tech is that it provides transportation services through an app. Moreover, if a person in this subgroup has heard of Tech, she might be aware that Tech is available in many locations, not just Gainesville, and the local focus of Promotions’ advertising will not tend to create confusion.
Were there no evidence of actual confusion, the foregoing analysis based on the other factors would counsel against finding that Promotions had shown a substantial likelihood of success on the issue of likelihood of confusion, though it would be a close call. But there is evidence of actual confusion, and the Eleventh Circuit has repeatedly emphasized the importance of such evidence. See, e.g., Caliber Auto. Liquidators, Inc. v. Premier Chrysler, Jeep, Dodge, LLC, 605 F.3d 931, 936 (11th Cir. 2010) (“evidence of actual confusion is the most weighty consideration” in the. likelihood of confusion analysis). Great weight is given to this factor for two reasons. First, actual confusion is direct evidence of precisely the occurrence that the “likelihood of confusion” inquiry is trying to assess the probability of. Second, giving great weight to evidence of actual confusion helps ensure that a judge doesn’t rely (perhaps subconsciously) too heavily on his oum subjective view of the likelihood of confusion. Cf. Jada Toys, Inc. v. Mattel, Inc., 518 F.3d 628, 633 (9th Cir.2008) (holding that a finding of no likelihood of confusion based solely on dissimilarity of marks was improper because it created “the potential for a judge to elevate his or her own subjective impressions of the relative dissimilarity of the marks over evidence of, for example, actual confusion”). In analyzing the other factors, this Court has of course tried to put itself into the shoes of an average consumer, and has tried to avoid giving too much weight to its own subjective impressions, but such efforts are perhaps doomed. See Societe Anonyme de la Grande Distillerie E. Cusenier Fils Aine & Cie. v. Julius Wile Sons & Co., 161 F.Supp. 545, 547 (S.D.N.Y.1958) (“[A] finding on infringement is by necessity a subjective determination by the trial judge based on his visceral reactions as to the likelihood of confusion the allegedly infringing mark will create in the minds of the public.”). Giving serious weight to evidence of actual confusion helps counteract the empathetic deficiencies that necessarily distort the analysis of the remaining factors.
It is true that there is not a huge amount of evidence of actual confusion in the record — perhaps five or so instances of people who truly thought that Promotions and Tech were related despite having a more-than-casual relationship with the two entities, plus some fraction of the callers. *1275But the Eleventh Circuit’s case law makes clear that even a few instances of consumer confusion can be sufficient to establish actual confusion. See Safeway Stores, 675 F.2d at 1167 (two instances of customer confusion enough so that “there [wa]s some indication, though slight, of actual confusion”); see also AmBrit, Inc. v. Kraft, Inc., 812 F.2d 1531, 1544 (11th Cir.1986) (noting that “it takes very little evidence to establish the existence of the actual confusion factor”). This is because “[ajctual confusion by a few customers is ... evidence of [a] likelihood of confusion by many customers.” Freedom Sav. & Loan Ass’n v. Way, 757 F.2d 1176, 1185 (11th Cir.1985) (citation omitted).
It’s a close call, but Promotions has established a substantial likelihood of success on the merits of the “likelihood of confusion” element of its claims. Because it has also established that it’s substantially likely to succeed in showing that it has enforceable trademark rights in the Gaines-ville area in the “UBER PROMOTIONS” mark, Promotions is substantially likely to succeed in showing that Tech has infringed on its mark by using its “UBER” marks in the Gainesville area.
11. UberEVENTS
Everything that has been said so far has not taken into account Tech’s new “Uber-EVENTS” service. That service, if allowed to grow in the Gainesville area pending trial, would substantially increase the risk of consumer confusion for three independent reasons. First, the more “Uber_” services there are, the more likely it is that a consumer might think that Promotions is just another branch on the Uber Technologies tree. Second, UberE-VENTS is accessed through a webpage, thus eliminating one of the key differences between Tech and Promotions. Third, the service allows users to coordinate transportation at a set time in the future for a large number of people, just like Promotions. Granted, the method by which people are transported is quite different from Promotions — one imagines that most driver partners’ vehicles are less lively than a party bus — but the nature of the service is far closer to the services offered by Promotions than any of Tech’s other offerings.
To put this in terms of the factors, the “similarity of sales methods” factor is less tilted towards Tech and the “similarity of services” factor tips in favor of Promotions when assessing Tech’s use of its UBER marks in connection with its Uber-EVENTS service. While the question of whether Promotions is substantially likely to succeed on the merits of its claim vis-a-vis UberX, etc. is a close call, these minor adjustments to the factors push Promotions’ claim vis-a-vis UberEVENTS well past the edge. Promotions has easily met its burden of showing that it is substantially likely to succeed on the merits of its claim that Tech’s use of its UBER marks in connection with the UberE-VENTS service creates a likelihood of consumer confusion.
D. Laches and Estoppel
Tech asserted laches and estoppel as affirmative defenses in its original answer, ECF No. 15, and again in its answer to the amended complaint, ECF No. 70, but it didn’t press those defenses at all in its opposition to the preliminary injunction motion, ECF No. 63, or at the hearing. These defenses would certainly fail in relation to Tech’s UberEVENTS service, which didn’t even start until after this lawsuit was brought. As for the remainder of Tech’s offerings, Tech’s failure to press any argument is fatal to it at this stage. When a party moving for preliminary in-junctive relief has shown ,a substantial likelihood of success on the merits of a claim, the burden shifts to the nonmoving party *1276to show that it is substantially likely to succeed on the merits of its affirmative defense. See Gonzales v. O Centro Espírita Beneficente Uniã, do Vegetal, 546 U.S. 418, 428-30, 126 S.Ct. 1211, 163 L.Ed.2d 1017 (2006) (holding in a case concerning RFRA that “the burdens at the preliminary injunction stage track the burdens at trial”); LSSi Data Corp. v. Comcast Phone, LLC, 696 F.3d 1114, 1123 n.11 (11th Cir.2012) (applying Gonzales in a case involving the Communications Act of 1934); Perfect 10, Inc. v. Amazon.com, Inc., 508 F.3d 1146, 1158 (9th Cir.2007) (applying Gonzales in a copyright case). Tech has clearly not met that burden, because it has not pressed these defenses at all in opposition to Promotions’ motion for a preliminary injunction.
IV. IRREPARABLE HARM
“[Pjlaintiffs seeking preliminary [injunc-tive] relief [must] demonstrate that irreparable injury is likely in the absence of an injunction.” Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 22, 129 S.Ct. 365, 172 L.Ed.2d 249 (2008) (emphasis in original). Moreover, it is not all future injuries that matter, but only those that would occur “before a decision on the merits can be rendered.” Id. (quoting 11A Charles A. Wright et al., Federal Practice and Procedure § 2948.1 (2d ed. 1995)).
If preliminary injunctive relief is denied in this case, what will likely happen? First of all, Tech will likely continue to grow in the Gainesville area, just as it has grown steadily in the past year and a half. ECF No. 44-1, at 2; ECF No. 37, at 29 ¶ 51; Gore Dep. 247-48. More and more people will become familiar with Tech, either by downloading its app and using its services, encountering its advertisements, or hearing about it through word-of-mouth. Some of these people (and some people already familiar with Tech) will become confused when they encounter Promotions’ marks, and may mistakenly associate Promotions with Tech.
But so what? Where’s the harml The harm usually cited as flowing from confusion in a reverse-confusion case is the loss to the senior user of the “value of the trademark — its product identity, corporate identity, [and] control over its goodwill and reputation.” A&H Sportswear, 237 F.3d at 228. This harm is usually deemed irreparable because loss of reputation, goodwill, etc. is not easy to quantify and can’t readily be remedied with money damages or a post-trial injunction. See Kraft Foods Grp. Brands LLC v. Cracker Barrel Old Country Store, Inc., 735 F.3d 735, 741 (7th Cir.2013) (Posner, J.) (“[I]rreparable harm is especially likely in a trademark case because of the difficulty of quantifying the likely effect on a brand of a nontrivial period of consumer confusion.”). It seems clear that if Promotions is harmed by confusion between now and the time of trial, that harm will be difficult, if not impossible, to quantify, and will thus be irreparable.
Still, the harm must be “likely” to occur for injunctive relief to be appropriate. Three considerations suggest that it is likely. First is the already-mentioned anticipated growth of Tech in the Gainesville market. The bigger Tech gets, the more likely confusion becomes, and the less control Promotions has over what people think of it. Tech’s growth has been rapid, and much harm could be done to Promotions’ reputation in the nine-plus months until trial.
Second is Tech’s UberEVENTS service, which is, for reasons discussed above in Part III.C.ll, likely to exacerbate consumer confusion. Because it is more directly competitive with Promotions’ services, UberEVENTS may even cause Promotions to lose business due to forward or initial interest confusion.
*1277The third consideration is the massive amount of negative press attention Tech receives. Promotions has put into evidence many examples of negative press, see ECF No. 27-19, but this Court, whether through the formal mechanism of judicial notice or the less-formal mechanism of gathering background facts,20 could point to many more. For instance, consider the top Google News search results (as of 5:00 p.m. on February 11) for “uber florida”:
[[Image here]]
With all due respect to Tech, Promotions has every reason not to want potential customers and other members of the public to associate it with a company that has inspired protests in cities around the world.21 Absent any injunctive relief, confusion is likely to damage Promotions’ reputation and goodwill and cause irreparable harm in the time between now and a resolution on the merits.
V. BALANCE OF HARDSHIPS
Promotions bears the burden of showing “that the threatened injury to [it] outweighs whatever damage the injunction may cause to” [Tech]. Davidoff & Cie, S.A. v. PLD Int’l Corp., 268 F.3d 1297, 1300 (11th Cir.2001). In conducting this measurement, it is important to remember that the harm to Tech will depend on the particular form of injunctive relief considered. Put a different (but completely equivalent) way, “the scope of an injunction should be determined by balancing harm to the plaintiff, other means of avoiding such harm, and the relative inconvenience to the defendant.” Foxtrap, Inc. v. Foxtrap, Inc., 671 F.2d 636, 640 (D.C.Cir.1982) (per cu-riam). As discussed above in Part II.C, this Court has discretion to tailor relief so as best to serve the interests of the parties and the public.
*1278A. Promotions’ Requested Relief
Promotions asks this Court “to enjoin [Tech]’s use of the mark UBER, and any similar variation, in connection with its software app, its transportation services, event services and related brand names, in and to the State of Florida and to Florida consumers of [Promotions’] Services or [TechJ’s Services.” ECF No. 27, at 39. As discussed above in Part III.B.2, Promotions is only substantially likely to succeed in showing that it has enforceable trademark rights against Tech in the Gainesville area, so this Court will certainly not enjoin Tech from using its marks anywhere else in Florida. The most Promotions can hope for is an injunction that would bar Tech from using its marks in connection with its services in Alachua County.
The parties disagree as to the practical effect that such an injunction would have. Promotions- claims that Tech need not stop operating in the Gainesville market altogether — it could instead make a new app with a new name and operate as a sort of parallel universe version of itself in the Gainesville area. ECF No. 27, at 37; ECF No. 71, at 140-45. (Actually, Promotions proposed that Tech do this throughout Florida, consistent with its request for an injunction covering the entire state. Id.) Tech points out that this would be infeasible, if not impossible, to do from a technical standpoint, and would in any event not be economically wise. ECF No. 37, at 24-26 ¶¶ 39-^3.
Promotions’ proposal does indeed seem impracticable for two reasons. First, there are the technological and economic impediments identified by Tech. Second, forcing Tech to change its name only in Gainesville would force it to give up the value of its trademark — Tech would basically be starting over, and would have to mount a massive campaign to introduce itself to the Gainesville public. Promotions’ proposal seems to ignore the value of name recognition and brand identity, which are the very things that trademark law seeks to protect.
In truth, the practical effect of enjoining Tech’s use of its marks in the Gainesville area would probably be that Tech would simply stop operating there. See ECF No. 37, at 28 ¶ 49. This seems like a drastic remedy. The harm to Tech would be great — not in terms of lost profits, perhaps, but in terms of setting back Tech’s efforts to build a customer base in the area. Tech has been operating in Gaines-ville for almost a year and a half and has made significant efforts to grow its business there; forcing it to leave town for nine months would do incalculable harm to that progress.
While it is true that “the injury a defendant might suffer if an injunction were imposed may be discounted by the fact that the defendant brought that injury upon itself,” Novartis Consumer Health, Inc. v. Johnson & Johnson-Merck Consumer Pharmaceuticals Co., 290 F.3d 578, 596 (3d Cir.2002), the amount of such discounting should depend on the willfulness of the infringement. Cf. Opticians Ass’n of Am. v. Indep. Opticians of Am., 920 F.2d 187, 197 (3d Cir.1990) (taking into account the fact that the defendant had “openly, intentionally, and illegally appropriated the [plaintiffs] marks, despite being warned not to” in conducting the balancing analysis). Here, there is no real indication that Tech acted with an intent to cause consumer confusion in order to benefit itself.
Additionally, the fact that some of the harm to Promotions could have been avoided had Promotions filed suit earlier must be taken into account. Promotions waited over a year after Tech moved into the Gainesville market to file suit. While its delay may not be so unjustified that *1279laches would be a defense, the delay still should be factored into the balance of hardships analysis. Cf. Tom Doherty Assocs., Inc. v. Saban Entm’t, Inc., 60 F.3d 27, 39 (2d Cir.1995) (“A district court should generally consider delay in assessing irreparable harm.”).22
Putting all of this together, it seems clear that Promotions has failed to meet its burden of showing that the injury it might suffer absent preliminary injunctive relief outweighs the injury to Tech that would result from enjoining Tech’s use of its marks in the Gainesville area pending trial.
B. Arming Squirrels With Bazookas
An injunction along the lines of the one requested by Promotions would be inappropriate for another reason. An injunction may be, and in many cases is, valued differently by different parties. See, e.g., BEM I, LLC v. Anthropologie, Inc., 301 F.3d 548, 553 (7th Cir.2002) (Posner, J.). For instance, if Promotions were a Manhattan-or Chicago-based company, complying with a preliminary injunction preventing Tech from using its marks in Promotions’ area for nine months might well cost Tech many, many times more money than Promotions could ever hope to make in profits. Such an injunction would effectively be a windfall to Promotions, as it could sell the right to enforce the injunction to Tech in exchange for a sum far exceeding its anticipated profits but still far below the cost to Tech of obeying the injunction. The award of a permanent, rather than a preliminary, injunction of that nature — entered after a finding of infringement — might be appropriate in such a case since the enjoined activity would violate trademark law. Cf. In re Brand Name Prescription Drugs Antitrust Litig., 123 F.3d 599, 610 (7th Cir.1997) (Posner, J.) (injunction preventing a defendant from engaging in unlawful activity doesn’t deprive the defendant of a legally-protected interest, and therefore value of injunction to defendant shouldn’t count in determining whether diversity jurisdiction threshold is met). But a preliminary injunction is more problematic because the plaintiff is receiving something of great value to the defendant (the injunction) based on a mere likelihood of success, which raises the distinct possibility that the entry of the preliminary injunction will wrongfully give the plaintiff a windfall.
Certainly this is a concern in any preliminary injunction proceeding — that’s .why movants must post injunction bonds, see Nokia Corp. v. InterDigital, Inc., 645 F.3d 553, 560 (2d Cir.2011) — but it’s particularly problematic in reverse confusion cases, which usually involve a small, regional plaintiff taking on a large national behemoth. One court characterized the problem in the following terms:
“A broad, far-reaching injunction barring defendant!, a large company,] from selling shoes in any manner connected to its Esprit mark would cost the defendant millions. So far as the evidence shows, on the other hand, plaintiffs business would be virtually unaffected whether an injunction is granted or denied. There is no convincing showing that plaintiff would lose a single sale if it dropped Esprit and instead used Rober-tina, or another mark. At present, the principal value of the Esprit mark to plaintiff lies in the price plaintiff can extract from defendant for its sale. A far reaching injunction would so damage defendant that it would permit plaintiff to demand a very high price.
*1280It is surely not the proper function of an equitable decree to give plaintiffs mark an extortion value far beyond the value it possess in plaintiffs business. From .this point of view, it may be questioned whether plaintiff should receive any injunction at all.”
H. Lubovsky, 627 F.Supp. at 496-97.
To slightly tweak a metaphor offered by this Court during the hearing, a preliminary injunction should not serve as a bazooka in the hands of a squirrel, used to extract from a more fearsome animal a bounty which the squirrel would never be able to gather by his own labors — at least not when the larger animal is mostly without sin.
C. Narrowly-Tailored Relief
While an injunction along the lines requested by Promotions would harm Tech more than the absence of any relief would harm Promotions, a more narrowly-tailored injunction — one that would do less harm to Tech while still preventing some harm to Promotions — might be in order. In particular, an injunction aimed at (1) preventing people looking for Tech’s phone number from coming across Promotions’ phone number and (2) preventing UberE-VENTS from causing further confusion would be appropriate.
Many, many people have called Promotions looking for Tech. Some of these people are undoubtedly confused, with that confusion resulting from the fact that Promotions comes up when one searches “Uber Gainesville phone” or “Uber Gaines-ville phone number” on Google. Requiring Tech to set up a local (352 area code) number to handle calls from Tech customers and driver partners in the area would help reduce confusion and would not be unduly burdensome to Tech. Relief of this nature is often awarded to reduce confusion in trademark cases. See generally McCarthy § 30:7.
As for UberEVENTS: as discussed above in Part III.C.ll, if Tech is allowed to continue to use its “UBER” marks in connection with UberEVENTS in the Gainesville area, there is an increased risk of consumer confusion — including forward confusion and loss of sales — and harm to Promotions’ ability to control its name in the event transportation field. UberE-VENTS is a new service, and “[ujsage in th[e] [Gainesville] community is minimal.” ECF No. 71, at 73. With that in mind, it seems clear that an injunction preventing Tech from using its marks in connection with UberEVENTS pending a resolution of the merits of this case would do very little damage to Tech while preventing a non-negligible amount of harm to Promotions. It also seems that compliance with such an injunction would not be unduly burdensome from a technological standpoint — Tech would just have to ensure that any customer inputting an Alachua County address into the form on the UberE-VENTS page would be unable to proceed.
Tech’s lack of an easy-to-find phone number and its recently-launched UberE-VENTS service are two facets of its business that are likely to cause confusion among customers but are also not so central to Tech’s operation that injunctive relief would- do much harm to Tech. For that reason, an injunction (1) requiring Tech to maintain a local phone number, and to take steps to make the number easily findable; and (2) preventing Tech from using its marks in connection with its UberE-VENTS service in the Gainesville area pending trial would do less harm to Tech than the lack of any injunctive relief would do to Promotions.
VI. PUBLIC INTEREST
A preliminary injunction may not issue unless a movant shows that the “injunction is in the public interest.” Winter, *1281555 U.S. at 20, 129 S.Ct. 365. In the trademark infringement context, this means taking into account a number of things: the public’s interest in not being confused, Kos Pharmaceuticals, Inc. v. Andrx Corp., 369 F.3d 700, 730 (3d Cir.2004); the public’s interest “in encouraging, not stifling, competition,” Calvin Klein Cosmetics Corp. v. Lenox Laboratories, Inc., 815 F.2d 500, 505 (8th Cir.1987); and “the broader economic implications of [the] grant of the preliminary injunction,” id.
A. Promotions’ Requested Relief
In this case, enjoining Tech from using its marks in connection with its services in. Gainesville would not serve the public interest. While such an injunction would certainly reduce consumer confusion, its negative effects would be far more serious. As discussed above in Part Y.A, Tech would likely (at least temporarily) pull out of the Gainesville market, which would negatively impact Tech’s driver partners. ECF No. 37, at 26-27 ¶¶ 44-45. Tech would also have to cease participating in its “Freedom in Motion” program, in which it “facilitates on-demand, subsidized transportation services for residents of senior citizen communities,” and in the “Safe Rides” program, “which gives university students discounted rates on their rides when returning home late at night.” Id. at 28-29 ¶ 50.
The public’s interest in having on-demand transportation would also be dis-served by granting the relief Promotions requests. Tech provides a lower-cost alternative to traditional taxi services that has apparently grown in popularity with Gainesville riders as of late. See id. at 29 ¶ 51. Yanking that away — particularly when Promotions is not. directly competitive with Tech and could not fill the void left by Tech’s departure23 — would deprive citizens of a service they’ve come to rely on for transportation.
B. Narrowly-Tailored Relief
On the other hand, the more limited relief discussed above in Part V.C would serve the public interest. The limited injunction would help reduce confusion, but it would not force Tech to abandon any of its programs. While enjoining any further growth of UberEVENTS might impact Tech’s driver partners, the effect is likely to be very minor given the fact that Uber-EVENTS is relatively new and has not yet taken off in Gainesville.
VII. SECURITY
“[A] court may issue a preliminary injunction ... only if the movant gives security in an amount that the court considers proper to pay the costs and damages sustained by any party found to have been wrongfully enjoined or restrained.” Fed. R. Civ. P. 65(c). Tech asked for a security bond of $64 million, but that was based upon a broad injunction covering all of Florida. ECF No. 63, at 44. Given the very limited scope of the injunction, a bond in that amount would be excessive.
It’s very hard to estimate how much money Tech might lose by not operating its UberEVENTS service for nine months,24 and it’s also not clear how much it might cost Tech to set up a phone number and to ensure that that number comes up when people conduct the appro*1282priate searches on Google, Bing, etc. Thinking in terms of orders of magnitude, it seems that one thousand (103) dollars would be too low and one hundred thousand dollars (105) would be too high. Ten thousand (104) dollars may or may not be close, but it’s certainly in the right ballpark.
VIII. CONCLUSION
Promotions is substantially likely to succeed in showing that Tech has infringed on its common-law trademark rights in the Gainesville area. It is also likely to suffer irreparable harm absent any injunctive relief pending a trial on the merits. However, enjoining Tech from using its marks in connection with any of its services in the Gainesville area would do more harm to Tech than Promotions would suffer in the absence of injunctive relief, and would also not serve the public interest. A more limited injunction aimed at preventing people looking for Tech’s phone number from coming across Promotions’ phone number and at preventing Tech from using its marks in connection with its UberEVENTS service would do less harm to Tech than Promotions would suffer in the absence of any relief and would also serve the public interest.
Accordingly,
IT IS ORDERED:
1. Plaintiffs motion for a preliminary injunction, EOF No. 27, is GRANTED IN PART and DENIED IN PART.
2. Defendant (including its officers, agents, servants, employees, attorneys, and other persons in active concert or participation with it) is preliminarily enjoined from using the UBER mark (or any variant thereof) in connection with the UberEVENTS service in the Gainesville market (Alachua County) until further notice.
a. Defendant (including its officers, etc.) shall not advertise the Uber-EVENTS service in Alachua County or cause it to be advertised in Alachua County until fur- . ther notice. A posting promoting UberEVENTS placed on the Fa-cebook wall or page of an entity with its principal place of business in Alachua County is an advertisement. A posting promoting Uber-EVENTS placed on the Facebook wall or page of a real person whose usual place of abode or residence is in Alachua County is an advertisement.
b. Defendant must ensure that if a person attempts to “book” an event in Alachua County through the UberEVENTS webpage, that booking is not allowed to be completed.
3.Defendant must set up a 352-area-code phone number to handle phone calls.
a. Defendant must ensure that this phone number is listed in all available directories.
b. Defendant must ensure that a search conducted with the Google, • Yahoo, or Bing search engine using the keywords “Uber Gaines-ville phone” or “Uber Gainesville phone number” returns a result containing Defendant’s 352-area-code number along with words clearly indicating that the result is associated with Defendant. Such words may include “driver partner,” “app,” or “ride.” Defendant must ensure that this result, while prominently displayed on the search results page, does not replace the result for Plaintiffs phone number that is currently returned when a search is con*1283ducted using these keywords.25 Compliance may entail using the search engines’ paid advertisement features such as Google Ad-Words:
c. Defendant must ensure that a search conducted with the Google, Yahoo, or Bing search engines using the keywords “Uber promotions Gainesville phone” or “Uber promotions Gainesville phone number” does not return a result with Defendant’s 352-area-code number.
4. On or before Friday, February 19, 2016, Defendant shall notify this Court if it has an objection to the security bond amount being set at ten thousand dollars ($10,000.00). If Defendant has no objection, Plaintiff shall post security with this Court in the amount of ten thousand dollars ($10,000.00) on or before Monday, February 22, 2016. If Defendant has an objection, this Court will consider such objection and enter a later order setting the amount of the security bond and a time by which Plaintiff must post such a bond.
SO ORDERED on February 16, 2016.
APPENDIX
*1284[[Image here]]
. Both the "UBER” and "UBERCAB” marks "consist[ ] of standard characters without claim to any particular font, style, size, or color.” ECF No. 41-1, at 2, 10.
. For a graphical depiction of Tech’s expansion, see Ellen Huet, Uber's Global Expansion in Five Seconds, Forbes.com (Dec. U, 2014), http ://www.forbes. com/sites/ellenhuet/2 014/ 12/11/ubers-global-expansion/.
. The UberBLACK service — which is not currently available in Gainesville, see ECF No. 63, at 156 — allows customers to get rides in higher-end vehicles from driver partners who are professional drivers. Id. at 158-59.
. Throughout this opinion, this Court will generally use the terms "trademark” and "mark” "in a broad and generic sense to denote the entire field of trademarks, service marks, trade names, and trade dress.” See Platinum Home Mortg. Corp. v. Platinum Financial Grp., Inc., 149 F.3d 722, 726 n.l (7lh Cir.1998). The differences between these types of protectable marks are not material to this dispute.
. Promotions identifies a number of instances in which it has transported people from Gainesville, and a handful of occasions on which it has provided transportation for peo-pie whose point of origin was outside the Gainesville area. See ECF No. 27-1, at 2-8 ¶¶ 4-19; ECF No. 63, 120-32.
. "[Allegations of forward confusion and reverse confusion do not form distinct claims— they are alternative theories that can be used separately or together in a trademark infringement claim under the Lanham Act.” Trouble v. Wet Seal, Inc., 179 F.Supp.2d 291, 297 (S.D.N.Y.2001).
. See Barton Beebe, An Empirical Study of the Multifactor Tests for Trademark Infringement, 94 Cal. L. Rev. 1581, 1582-83 (2006) ("Each circuit has developed its own formulation of the test. In the Eighth and Tenth Circuits, the test consists of six factors. Other circuits, such as the Seventh, typically use seven factors, while still others, such as the Second and the Ninth, typically use eight. The Third Circuit uses ten and the Federal Circuit thirteen ...”).
. This is not to say that other demographics don’t matter at all. Of course some older people may come across Promotions’ marks, and even if they aren't likely to use Promotions’ services, their confusion may ultimately do harm to Promotions. See, e.g., Beacon Mut. Ins. Co. v. OneBeacon Ins. Grp., 376 F.3d 8, 16-17 (1st Cir.2004) (noting that "the likelihood of confusion inquiry is not limited to actual or potential purchasers, but also includes others whose confusion threatens the trademark owner’s commercial interest in its mark,” and that "[r]elevant confusion among non-purchasers may well extend beyond the confusion of those persons positioned to influence directly the decisions of purchasers”).
. Decisions of the Fifth Circuit handed down prior to October 1, 1981 are binding within the Eleventh Circuit. Bonner v. City of Prichard, 661 F.2d 1206, 1207 (11th Cir.1981) (en banc).
. These definitions were found using Tech-nische Universitat Chemnitz's online translation service. See BEOLINGUS-Your Online Dictionary, http ://dict.tu-chemnitz. de/dings. cgi?lang=en;service=deen (last visited Feb. 2, 2016).
.See Uber, Merriam-Webster.com, http:// www.merriam-webster.com/dictionary/uber (last visited Feb. 2, 2016).
. In the time between the hearing and this Order, Tech "rebranded” with a new logo and graphic scheme. See, e.g., Brian Mastroianni, Uber re-brands itself with a new logo, CBS News (Feb. 2, 2016), www.cbsnews.com/ news/uber-rebrands-itself-with-new-app-logo/. The new "look,” while still conservative compared to Promotions' graphic presentation, is slightly more colorful. This new look in and of itself doesn’t affect the "similarity of marks” analysis in any significant way given that the people most likely to be confused have probably only heard of Tech. The fact of the change — and the fact that it was so well-publicized — might tend to increase reverse confusion by creating doubt among those familiar with Tech about what its "look” really is, but such an effect would not change this Court's conclusion that the “similarity of marks” factor weighs slightly in favor of Promotions.
. This Court confirmed Promotions' allega- ' tions to that effect by conducting a Google search using the search terms "gainesville uber.” The results page features an ad for Tech above the search results.
. Tech does have an emergency phone number that can be found using its app. Gore Dep. 159-60.
. During Michael Farzad’s deposition, he received two phone calls from "confused'1 customers. After he mentioned this to Tech’s counsel, the following exchange took place:
Q. With respect to the calls that you received today, were either of those individuals confused at the point in which they were using your services?
A. No. They’re calling me because they think I'm Uber Technologies and they want to reserve services and ask questions.
Q. So those individuals were looking for Uber Technologies?
A. Right, and they left messages on my phone and they — they’re thinking that they're speaking to Uber Technologies.
ECF No. 63, at 136-37. This doesn’t prove anything, of course, but it tends to support the conclusion that many of the callers to Promotions don’t even realize they're talking to Promotions.
. At least not in this case. If Tech were the senior user and Promotions the junior user, the fact that consumers were being diverted to Promotions — even if they initially thought they were dealing with Tech and were in fact trying to reach Tech — would be highly relevant on an "initial interest confusion” theory in a suit brought by Tech. See Axiom Worldwide, 522 F.3d at 1221-22. But such confusion is not very important to this case, because such callers are not confused as to source, just misdirected. There is not, to this Court's knowledge, a “reverse initial interest confusion” theory of trademark infringement, nor would one make sense; the plaintiff proceeding under such a theory would be hard-pressed to articulate what harm it was suffering.
. Some of this confusion might be "forward” or "direct” confusion — people encountering Tech and thinking it’s related to Promotions.' Evidence of such confusion is relevant even in a case relying on a reverse-confusion theory. See A&H Sportswear, 237 *1273F.3d at 233 ("CT]he manifestation of consumer confusion as ‘direct’ or ‘reverse’ may merely be a function of the context in which the consumer first encountered the mark. Isolated instances of ‘direct’ confusion may occur in a reverse confusion case, and vice-versa.”).
. With the exception of the UberEVENTS service, which will be discussed later.
. According to a recent study, 86% of Americans aged 18-29 own a smartphone, compared to only 68% of all adults in the United States who own a smartphone. Pew Research Center, Technology Device Ownership: 2015 7 (2015), http://www.pewinternet.org/files/ 2015/10/PI-2015-10-29_device-ownership_ FINAL.pdf. A slightly older study conducted in the United Kingdom suggests that younger people are generally more tech-sawy than older people. See Techie teens are shaping how we communicate, Ofcom.org.uk (Aug. 7, 2014), http://media.ofcom.org.uk/news/2014/ cmr-uk-2014/. These studies do not conclusively establish that the young people most likely to come across Promotions' marks are precisely the people least likely to be confused, but they do tend to support that commonsense view.
. By "background facts” this Court means "facts ... designed to increase the reader’s understanding of a case by placing the adjudicative facts in an illuminating context.” Richard A. Posner, Reflections on Judging 137 (2013).
. See Marc Santora & John Surico, Uber Drivers in New York City Protest Fare Cuts, N.Y. Times (Feb. 1, 2016), http://www. nytimes.com/2016/02/02/nyregion/uber-drivers-in-new-york-city-protest-fare-cuts. html?_r=0; see also Alissa J. Rubin & Mark Scott, Clashes Erupt Across France as Taxi Drivers Protest Uber, N.Y. Times (June 25, 2015), http://www.nytimes.com/2015/06/26/ business/international/uber-protests-france. html.
. Since delay bears on the question of whether irreparable harm is likely to occur, it also bears on the question of the extent of irreparable harm that is likely to occur, which then goes into the balance of hardships analysis.
. Promotions suggested at the hearing that riders in downtown Gainesville can use their buses if they see one roaming around at night and "flag it down.” ECF No. 71, at 80, 83. Even if this is true, 'this is no substitute for Tech’s services, which allow users to request rides originating anywhere in the Gainesville area.
. In their papers and at the hearing, both parties focused on the probable effects of a statewide injunction prohibiting Tech from using its marks in connection with any of its services. The parties did not present much, if any, evidence as to the likely effects of more limited relief.
. Appendix A contains a screenshot of a Google search for "uber gainesville phone number” showing where Tech must ensure that its 352-area-code number shows up.