*1 Before H AMILTON , B RENNAN , and K IRSCH , Circuit Judges . B RENNAN , Circuit Judge . Pilar Domer submitted an online order to pick up a can of paint at a Menards home improve- ment store. Menards charged Domer a $1.40 fee for the pickup service she selected. Domer commenced this putative class ac- tion, alleging that Menards had not disclosed the pickup ser- vice fee and used the fee to manipulate its prices. Menards *2 moved to compel arbitration of Domer’s claims. The district court granted Menards’s motion, finding that the parties had entered into an arbitration agreement and that Domer’s claims fell within its scope.
Domer appealed, arguing that the arbitration agreement was invalid and unenforceable, and in any event did not cover her claims. We affirm the district court . Menards has shown that its website provided reasonably conspicuous notice of the terms to which Domer would be bound and Domer un- ambiguously manifested her assent to those terms. Addition- ally, each of Domer’s claims arise from or relate to the contract between Domer and Menards. So, Domer’s claims are within the scope of the arbitration agreement.
I
Menard, Inc. is a home improvement retail company that sells goods through brick-and-mortar stores and on its web- site. [1] Customers who purchase products online have three op- tions for receiving their goods. They may go to a local store and locate the item on the shelf, pay a small fee to have a Menards employee locate the item on the shelf and prepare it for pickup, or have the product shipped to their home.
Domer visited the Menards website to purchase a can of paint. Before checking out, she chose option three on the web- site: to have an employee retrieve the correct item from the shelf, prepare it for pickup at the Menards store in Valparaiso, Indiana, and place it at the pickup counter. In exchange for that service, Menards would charge her a $1.40 per item fee.
After Domer selected the pickup option, the website pre- sented her with the final page in the online checkout process :
As shown, the page presented a summary of Domer’s transaction. It included one shaded, bordered box listing her “Billing & Credit Card Information” and an adjoining column with her “Order Summary.” The “Billing & Credit Card Infor- mation” listed her payment information. The box also pre- sented options to receive text updates on the status of her or- der and to use a gift card. The “Order Summary” column listed four line-items: the “Merchandise Subtotal” (the price of the items purchased); the “Processing Fees” (the pickup service fee); the “Sales Tax”; and the “Total” cost. In addition, the “Order Summary” column included a sentence explaining how much money Domer had saved on her purchase, the op- tion to submit her purchase, and a couple of sentences adver- tising the Menards credit card.
The same page included an additional note to purchasers immediately below the “Billing & Credit Card Information” section:
Please note : Gift cards may be used as a tender for any merchandise portion of your order and cannot be applied to purchase of another gift card. Applied gift card tenders will be used first towards your purchase and the remaining bal- ance applied to your entered credit card. By submitting your order you accept our Terms o f Order.
The note began with bold font and used the same size and style font as the surrounding text.
Directly below the note, there were two hyperlinks: “View Return Policy | Terms of Order Information.” The links, like the note, used the same size font as the surrounding text. But the links were green, unlike the black font of the surrounding text. Clicking on the “Terms of Order Information” link opened a text box with the Menards Terms of Order. The first paragraph of the Terms of Order contained an arbitration clause:
READ THIS CONTRACT CAREFULLY . … Purchaser agrees that any and all controversies or claims arising out of or re- lating to this contract, or the breach thereof, shall be settled by binding arbitration adminis- tered by the American Arbitration Association under its applicable Consumer or Commercial Arbitration Rules. Purchaser agrees that all ar- bitrators selected shall be attorneys. This *5 5 provision shall supersede any contrary rule or provision of the forum state. YOUR PURCHASE OF THE PRODUCT ON THIS CONTRACT CONSTITUTES YOUR AGREEMENT TO ALL TERMS AND CONDITIONS STATED ABOVE.
The text box for the Terms of Order used a larger font than the surrounding text.
Domer completed her purchase and picked up her can of paint in the designated store. Domer then commenced this putative class action. [2] Domer’s amended complaint included three claims: (1) violation of the Indiana Deceptive Consumer Sales Act, Ind. Code § 24-5-0.5-4; (2) violation of the Wiscon- sin Deceptive Trade Practices Act, Wis. Stat. § 100.18; and (3) unjust enrichment, see Sands v. Menard , 904 N.W.2d 789, 798 (Wis. 2017) (describing the elements of Wisconsin’s unjust enrichment law). She alleged that Menards had used the pickup service fee to manipulate its prices— artificially deflat- ing them by $1.40 in its advertising materials and then re- couping that difference in the form of an undisclosed fee. A d- ditionally, she argued that the $1.40 processing fee for in-store pickup had not been disclosed, and, had it been disclosed, she would not have purchased the can of paint.
Menards moved to compel arbitration under the Federal Arbitration Act, 9 U.S.C. §§ 3, 4. To Menards, Domer entered into an enforceable arbitration agreement by accepting the Menards Terms of Order when she completed her online pur- chase. As a result, the parties had a valid and binding agree- ment requiring arbitration of any disputes that either arise from or relate to Domer’s online purchase. Menards further contended that Domer’s claims fall within the scope of the ar- bitration agreement and, therefore, must be arbitrated.
The district court ruled for Menards. The court found that the arbitration agreement was enforceable. The totality of the circumstances demonstrated fair and adequate notice of the existence of an agreement and the consequence of proceeding with an online purchase. The court also found that “[a]ll of Domer’s claims are related to her contract of purchase with Menards: Domer agreed to pay Menards money, and Menards provided her with a can of paint in exchange.” So, her claims fell within the scope of the “expansive” arbitration clause to which she had agreed.
II
On appeal, Domer argues that the district court errone- ously granted Menards’s motion to compel arbitration. To her, the arbitration agreement was invalid and unenforceable, and in any event did not cover her claims.
The “[a]rbitrability of a dispute is often a question of law
that does not depend on undisputed facts… .”
Scheurer v.
Fromm Fam. Foods LLC
,
When a factual dispute exists about whether an arbitration agreement was made, parties have options. If the question is brought to a jury, we “must uphold the finding if it is sup- ported by a reasonable basis in the record.” Id. This includes evidence or expert testimony. Additionally, if the judge holds an evidentiary hearing, we review findings of fact “for clear error.” Id. The parties here chose neither of these possibilities. Domer has conceded that here “the facts are undisputed.” [3]
In cases with no factual disputes, the district court decides
the issue as a matter of law. In these instances, like here, our
“review should be de novo.”
Id.; see Gore v. Alltel Commc’ns,
LLC
,
A. Formation of the Arbitration Agreement
Whether an agreement to arbitrate has been formed is gov-
erned by state-law principles of contract formation.
Gore
, 666
F.3d at 1032. Wisconsin contract law applies here.
Rock Hemp
Corp. v. Dunn
,
The Menards Terms of Order contains an arbitration clause. Domer contends she did not agree to these terms, and therefore did not agree to arbitrate her claims, when she sub- mitted her purchase order on the Menards website.
Under Wisconsin law, the formation of a valid contract re-
quires that “an offer was accepted .”
Wells Fargo Bus. Credit v.
Hindman
, 734 F.3d 657, 667 (7th Cir. 2013) (quoting
Hoeft v.
U.S. Fire Ins. Co.
,
One way to assent to and form a contract online is for a customer to click on an “I Accept” button as part of a “click- wrap” agreement. See id. “Courts around the country have *9 recognized that this type of electronic ‘click’ can suffice to sig- nify the acceptance of a contract.” Id. These agreements differ from “browsewrap” agreements, which provide veiled notice to customers that mere use of the website constitutes agree- ment to various terms and conditions. See Oberstein v. Live Na- tion Ent., Inc. , 60 F.4th 505, 513 (9th Cir. 2023) (comparing clickwrap and browsewrap agreements in the context of online arbitration agreements). Browsewrap agreements typ- ically are not enforced. Id.
But, as here, online agreements often fall somewhere in
between these two forms of agreement. Courts acknowledge
that some companies “rely on simply displaying,” some-
where on a webpage, “a notice of deemed acquiescence and a
link” to the putative terms.
Cullinane v. Uber Techs., Inc.
, 893
F.3d 53, 62 (1st Cir. 2018). When this happens, “purported as-
sent is largely passive.”
Schnabel v. Trilegiant Corp.
, 697 F.3d
110, 120 (2d Cir. 2012). And when assent is passive, a court
will recognize an enforceable contract “only if: (1) the website
provides reasonably conspicuous notice of the terms to which
the consumer will be bound; and (2) the consumer takes some
action, such as clicking a button or checking a box, that unam-
biguously manifests his or her assent to those terms.”
Berman
v. Freedom Fin. Network, LLC
,
Although these questions may involve underlying facts,
they are questions of law. Rather than resolving questions of
fact, we undertake a fact-intensive legal analysis.
See Sgouros
,
Domer did not click a button saying “I Accept” to form the arbitration agreement. So, to prove that the arbitration agree- ment was formed, Menards must satisfy the test for when as- sent is passive. Although we analyze the website checkout page presentation, we do so to decide two legal questions: (1) did the website provide reasonably conspicuous notice of the terms to which the consumer will be bound, and (2) did the consumer take some action that unambiguously mani- fested his or her assent to those terms? See Berman , 30 F.4th at 856.
1.
Reasonably conspicuous notice
We examine notice from the perspective of a reasonable
online shopper—that is, a person who is neither an expert nor
a novice with technology.
See Meyer
,
When deciding whether an online disclosure has afforded fair notice, we consider five elements: (1) the simplicity of the screen; (2) the clarity of the disclosure; (3) the size and color- ing of the disclosure’s font; (4) the spatial placement of the hy- perlink; and (5) the temporal relationship to the user’s action. See, e.g. , id. at 78–79. No single factor is dispositive: the ques- tion is whether the website provided reasonable notice “in light of the whole webpage.” Nicosia v. Amazon.com, Inc. , 834 F.3d 220, 237 (2d Cir. 2016).
Simplicity of the screen.
First, we assess the simplicity of the
information on the screen and the way that information is pre-
sented.
See Meyer v. Uber Techs., Inc.
,
Menards online checkout page includes fields for (1) the user to enter his credit card details, (2) an order summary list- ing costs with a button to submit the order, (3) an option to receive text updates on the status of the order, (4) a note about the use of gift cards and explaining that “By submitting your order you accept our Terms of Order,” (5) hyperlinks to the Terms of Order Information and Return Policy, and (6) a note explaining that the shopper could save money with a Menards credit card. The only non-standard item in this list is the Menards credit card note.
Domer argues the Menards webpage is “sufficiently dis- tracting so as to divert users’ attention from the disclosure and consequences of placing an order.” She compares the check- out page to the interface in Nicosia . 834 F.3d at 240–41. The court in Nicosia found the webpage may not have been clear because of the company’s use of many different fonts (“vari- ous text is displayed in at least four font sizes and six colors (blue, yellow, green, red, orange, and black)”); the presence of numerous extraneous elemen ts (“multiple buttons and pro- motional advertisements”); the spatially dissipated, complex *12 presentation of various categories of transaction-related infor- mation (“customers’ personal address, credit card infor- mation, shipping options, and purchase summary”); and the sheer number of hyperlinks on the page (“between fifteen and twenty- five links”) . Id. at 236–38. But the court “d[id] not hold that there was no objective manifestation of mutual assent here as a matter of law.” Id. at 238. Rather, it concluded “simply that reasonable minds could disagree on the reason- ableness of notice.” Id. (holding that the district court erred in concluding that Nicosia failed to state a claim under Federal Rule of Civil Procedure 12(b)(6)).
In contrast, the Menards page is streamlined, well-spaced, and internally consistent. There is ample white space; nearly all the text and images on the screen are pertinent to the checkout process; and the page is organized into just a few neat boxes and columns ( see “Billing & Credit Card Infor- mation” and “Order Summary”). And the page is not littered with dozens (or even a handful) of hyperlinks. The page has a consistent color and typeface, and only a few items are pre- sented in bold type. So, those items—including the disclosure of the Terms of Order—are likely to catch a user ’s attention. The district court correctly found that the Terms of Order were reasonably conspicuous in this uncluttered presenta- tion.
Clarity of the disclosure.
Where, as here, “terms are not dis-
played” directly to a user, but instead “must be brought up
by using a hyperlink,” a “clear prompt directing the user to
read them” is required.
Sgouros
,
The Menards disclosure provides such a prompt for the
consumer. It explicitly tells the user: “
Please note
: … By sub-
mitting your order you accept our Terms of Order.” Courts
have compelled arbitration based on similar prompts.
See, e.g.
,
Meyer
,
Domer relies on
Sgouros
to argue that the Menards “
Please
note
” disclaimer was not conspicuous enough. But in
Sgouros
,
the bold text beneath the scroll box informed the purchaser
that clicking on the box constituted his authorization for the
company to obtain his personal information—it said “nothing
about contractual terms.”
Sgouros
,
On the Menards website, the prompt and the disclosure are separated by two sentences about gift cards, but they con- stitute only one line of text. This does not render the prompt inconspicuous or misleading. As the district court correctly observed, the bold font stating “ Please note ” “encourage[s] a reasonable user to read through to the end of the notice.” That Domer did not read the gift card sentences is not relevant. W hat matters here is that a reasonable internet user would have seen the bold font stating “ Please note ” against the clean, white background and would have been on notice to read the two lines of text following the prompt. See Meyer , 868 F.3d at 79 (“While it may be the case that many users will not bother reading the additional terms, that is the choice the user makes; the user is still on inquiry notice.”). This factor also weighs in favor of the finding that the Terms of Order were reasonably conspicuous.
Design of the hyperlinks and disclosure. The hyperlinks to the Terms of Order are offset from the white background in a bright, green color, which, as the district court observed, con- trasts with the black text of the disclosure immediately above it. See Meyer , 868 F.3d at 78 (disclosure was conspicuous in part because “the dark print contrasts with the bright white background”).
It does not matter that the text was green rather than the
“typical[] blue.”
Berman
,
The disclosure begins with the bolded “ Please note ”, drawing the customer’ s attention. It is placed alone below the “Billing & Credit Card Information,” where a customer is likely to see it. That Menards chose to place the hyperlinks below instead of within the disclosure does not increase the likelihood that a customer would overlook them. Rather, hav- ing the words “Terms of Order” listed twice, once in the dis- closure and once in the hyperlink, makes it more likely the customer sees them.
The district court correctly found that the font of the dis- closure and hyperlinks are readable and distinct from the sur- rounding text. So, this element weighs in favor of affirming the district court’s finding of fair notice.
Spatial placement of the disclosure.
Text advising users of
terms should be spatially coupled with the act deemed to
manifest assent to those terms.
See Starke
,
Domer correctly contends that the page lists the disclosure
near the bottom with language discussing gift card use. But
the disclosure and hyperlinks are placed directly below the
“Billing & Credit Card Information” box, which a reasonable
shopper undoubtedly reviews, and they are part of the natu-
ral visual flow of reviewing the page.
See Schnabel
, 697 F.3d at
127 (“[T]he presentation of these terms at a place and time that
the consumer will associate with the initial purchase or en-
rollment, or the use of, the goods or services from which the
recipient benefits at least indicates to the consumer that he or
she is taking such goods or employing such services subject
to additional terms and conditions that may one day affect
him or her.”). The user can see the disclosure and click the
hyperlink while viewing the rest of the checkout information
and does not need to scroll far, if at all, to read them.
See
*17
Meyer
,
Again, the cases Domer relies on are distinguishable. In
Wilson
, the company’s notice was in small, unbolded text, and
it was placed underneath large, colorful buttons .
See Wilson v.
Redbox Automated Retail, LLC
,
So, the notice and disclosure are not, as Domer suggests, “hidden” away. And this element weighs in favor of affirming the district court’s finding of a reasonably conspicuous notice.
Temporal relationship.
The disclosure is temporally con-
nected to the required act of the user—in other words, the
user encounters the disclosure on the same page where the
order will be placed.
See Meyer
,
To summarize on this first prong of whether an agreement was made when assent is passive, the Menards checkout page is not a visually bewildering screen. As the district court found, the website provided reasonably conspicuous notice of the terms to which the consumer, Domer, will be bound.
2.
Assent to the Terms of Order
On the second prong, Domer argues that because the
checkout webpage did not reasonably communicate the exist-
ence of the Terms of Order, she did not unambiguously as-
sent.
See Cullinane
,
Domer is incorrect. “There is nothing automatically offen- sive about such agreements, as long as the layout and lan- guage of the site give the user reasonable notice that a click will manifest assent to an agreement.” Sgouros , 817 F.3d at 1033–34. As explained above, the Menards notice was reason- ably conspicuous, so Domer manifested her assent to accept the Terms of Order by going through with the purchase. Un- like with pure clickwrap agreements, clicking “SUBMIT ORDER” “does not specifically manifest assent t o the addi- tional [arbitration] terms, for the purchaser is not specifically asked whether she agrees or to say ‘I agree.’” Nicosia , 834 F.3d at 236–37.
Nevertheless, Domer unambiguously gave her assent to the arbitration provision. Although Domer’s assent “was not express, we are convinced that it was unambiguous in light of the objectively reasonable notice of the terms, as discussed in detail above.” Meyer , 868 F.3d at 79; see Register.com, Inc. v. Verio, Inc. , 356 F.3d 393, 403 (2d Cir. 2004) (“[R]egardless whether [a user] did or did not say, ‘I agree’ … [the user’s] choice was either to accept the offer of contract, taking the in- formation subject to t he terms of the offer, or, if the terms were not acceptable, to decline to take the benefits.”).
As described above:
There is ample evidence that a reasonable user would be on inquiry notice of the terms, and the spatial and temporal coupling of the terms with the registration button indicated to the *20 consumer that he or she is … employing such services subject to additional terms and condi- tions that may one day affect him or her. … A reasonable user would know that by clicking the registration button, he was agreeing to the terms and conditions accessible via the hyper- link, whether he clicked on the hyperlink or not. Meyer , 868 F.3d at 79– 80 (citation and quotations omitted). Reasonable consumers understand there will be terms and conditions associated with using a website. That clicking the register button had two functions —purchase of an order and assent to the Terms of Service—does not render Domer’s as- sent ambiguous. See id. at 80. So, clicking the “SUBMIT ORDER” button demonstrated Domer’s unambiguous mani- festation of assent to the arbitration agreement.
* * *
Menards has shown that (1) its website provided reasona- bly conspicuous notice of the terms to which Domer would be bound, and (2) Domer’s click of the button “SUBMIT ORDER” unambiguously manifested her assent to those terms. See Berman , 30 F.4th at 856. So, the arbitration agree- ment was formed when Domer submitted her order.
B. Scope of the Arbitration Agreement
Once the parties “have a contract that provides for arbitra-
tion of some issues between them, any doubt concerning the
scope of the arbitration clause is resolved in favor of arbitra-
tion as a matter of federal law.”
Gore
,
After concluding that Menards and Domer had formed a valid arbitration contract, the district court found that Domer’s claims all arise from or relate to her contract of *21 purchase with Menards. So, Domer’s claims fall within the scope of the arbitration agreement.
Domer claims this court has no binding precedent for cases like this one where “the conduct underlying the claims occurred prior to the formation of the purchase contract.” And she asks this court to “clarify” its existing case law and adopt a bright-line rule that claims like hers are categorically beyond the scope of even the most broadly worded arbitra- tion clauses.
Domer is correct that this court has not expressly ruled
that claims arising from action taken prior to the formation of
an arbitration agreement are within the scope of the agree-
ment. But such a ruling is not required. We have explained
that claims may be arbitrated when they fall “within the scope
of the agreement” to arbitrate.
Rock Hemp Corp.
The Menards Terms of Order represents a valid, binding agreement to arbitrate and is framed in broad terms that state, in part:
Purchaser agrees that any and all controversies or claims arising out of or relating to this con- tract, or the breach thereof, shall be settled by binding arbitration … .
Each of Domer’s claims “arises out of” or “relates to” her con-
tract with Menards. We see this in her theory of liability and
in the injuries she claims. As the district court explained,
“[t]he theory behind her consumer protection and unjust en-
richment claims is that Menards misled her about the price
that she would pay for that purchase.” A party “may not
avoid a contractual arbitration clause merely by ‘casting its
complaint in tort.’”
Sweet Dreams Unlimited, Inc. v. Dial-A-Mat-
tress Int’l, Ltd.
,
Additionally, the only harms Domer alleges are rooted in, and related to, the arbitration contract. [4] Domer disagrees, ar- guing that her claims are not related to the contract. Accord- ing to Domer, she “rel[ies] solely on the allegation that Menards induced and enticed customers to patronize its web- site through false representations regarding the price of items listed on its website.” This “occurred prior to and inde- pendently from the terms of the contract containing the arbi- tration provision.”
Domer believes this temporal argument differentiates her
case from cases like
Sweet Dreams
and
Kiefer
.
Sweet Dreams,
1
F.3d at 640 (plaintiff alleged the defendant had “fraudulently
induced it to continue making expenditures … after the
Agreement had expired”);
Kiefer
,
Domer also argues that none of her claims require her to “reference [] the terms of the purchase contract” and are therefore not “encompassed by the scope of the arbitration provision.” But under I ND . C ODE § 24-5-0.5-4:
A person relying upon an uncured or incurable deceptive act may bring an action for the dam- ages actually suffered as a consumer as a result of
the deceptive act or five hundred dollars ($500), whichever is greater.
I ND . C ODE § 24-5-0.5-4 (emphasis added);
see Castagna v. New-
mar Corp.
,
Similarly, the elements of a private action under W IS . S TAT .
§ 100.18 are: (1) the defendant advertised the product; (2) the
advertising was misleading; and (3) the plaintiff
suffered pecu-
niary loss
because of the misleading advertising.
See
W IS . S TAT .
§ 100.18;
Tietsworth v. Harley-Davidson, Inc.
,
Further, Wisconsin’s unjust enrichment law requires proof of:
(1) a
benefit conferred on the defendant
by the
plaintiff; (2) appreciation or knowledge by the
defendant of the benefit; and (3) acceptance or
retention of the benefit by the defendant under
circumstances making it inequitable to do so.
Sands
,
As shown in the emphasized language above, each of
these claims require an injury to Domer: (1) “damage actually
suffered,” (2) “suffered pecuniary loss,” or (3) “benefit con-
ferred on [Menards].” I ND . C ODE § 24-5-0.5-4; W IS . S TAT .
§ 100.18;
Sands
,
Even if the conduct had occurred before formation of the
arbitration agreement, the Supreme Court and Seventh Cir-
cuit have compelled arbitration in those instances.
See, e.g.
,
Prima Paint Corp. v. Flood & Conklin Mfg. Co.
,
Domer cites to authority that is non-binding and distin-
guishable. Her alleged injuries would not have existed with-
out the one contract that included the arbitration agreement.
Cf. Matalka v. Home Point Fin. Corp.
,
In the alternative, Domer suggests this court adopt a standard she believes the Fifth and Sixth Circuits use to deter- mine whether an arbitration provision encompasses a claim. According to Domer, those circuits ask whether a claim can be maintained without reference to the agreement at issue. So, she argues her claim is “likely outside the scope of the *26 arbitration agreement.” But Domer’s analysis of our fellow circuits’ standards is faulty.
The Fifth and Sixth Circuits, like this court, do not apply a
test distinct from the framework used to assess whether a
claim falls within the scope of an arbitration agreement. Once
the validity of the agreement has been confirmed, d oubts con-
cerning the scope of arbitrability are resolved in favor of arbi-
tration.
See Safer v. Nelson Fin. Grp., Inc.
,
The general rules of contract formation, the elements of Domer’s claims, and precedent assure us that the district court correctly found that Domer’s claims are within the scope of the arbitration agreement.
III
We compliment the district court for its good work in this case. Domer and Menards formed an arbitration agreement which encompasses Domer’s claims. For these reasons, we A FFIRM the district court’s decision.
H AMILTON , Circuit Judge , concurring in the judgment. Start from the intuition that there is something odd about judges looking at online user interfaces with magnifying glasses and debating font color and size and the placement of hyperlinks. This case illustrates why courts should try to rethink how we approach the issues of “click-wrap” and “browse-wrap” con- tracts of adhesion in online commerce. One modest step in that project would be to treat whether a “browse-wrap” user interface gives sufficient notice of hidden but hyperlinked terms and conditions as a question of fact rather than a ques- tion of law. In an appropriate case, the issue should be sub- mitted to a jury, including in cases subject to the Federal Ar- bitration Act.
The issue here is one we all encounter with virtually every online purchase. Somewhere in the steps to complete the pur- chase, the online merchant includes on a screen a link that al- lows the customer to access a long statement of terms and conditions. They’re written in impenetrable legalese and usu- ally include an arbitration clause, often even purporting to waive a right to collective or class arbitration. Does anybody read them? Empirical data indicate that only two or three cus- tomers in a thousand spend more than one second looking at those terms. [1]
I address in Part I how far the realities of online commerce have drifted from the law of contracts. Part II addresses the false assumptions underlying the reasonable notice require- ment. Part III addresses the realities of user-interface design and consumer behavior. Part IV summarizes how courts have been deciding these issues about the terms of contracts and argues that we should be treating them as issues of fact. Part V briefly addresses jury trials under the Federal Arbitration Act.
I. “Mutual Assent” in the Era of Digital Consumer Contracts of
Adhesion
The common law of contracts emerged from the paradigm
of agreements between two repeat-player merchants negoti-
ating all the terms to which both assent. Courts ask if there
was a meeting of the minds, shown through evidence of ob-
jective actions and communications. E.g.,
Household Utilities,
Inc. v. Andrews Co.
, 71 Wis. 2d 17, 29, 236 N.W.2d 663, 669
(1976) (explaining that a “‘meeting of the minds’ between par-
ties … is required to establish the existence of a contract”);
State v. Bembenek
,
Mass retail commerce required courts to adapt that para- digm for a world of high-volume contracts of adhesion. The Supreme Court of New Jersey wrote two generations ago:
The traditional contract is the result of free bar-
gaining of parties who are brought together by
the play of the market, and who meet each other
on a footing of approximate economic equality.
*30
In such a society there is no danger that freedom
of contract will be a threat to the social order as
a whole. But in present-day commercial life the
standardized mass contract has appeared. It is
used primarily by enterprises with strong bar-
gaining power and position. “The weaker party,
in need of the goods or services, is frequently
not in a position to shop around for better terms,
either because the author of the standard con-
tract has a monopoly (natural or artificial) or be-
cause all competitors use the same clauses. His
contractual intention is but a subjection more or
less voluntary to terms dictated by the stronger
party, terms whose consequences are often un-
derstood in a vague way, if at all.”
Henningsen v. Bloomfield Motors, Inc.
,
Now, in online transactions, merchants use contracts of adhesion with terms available only through hyperlinks that the consumer might or might not even notice, let alone read and understand. These online practices have taken us even further from the paradigm of contract law. Digital terms and conditions are “ubiquitous, and their terms are voluminous, onerous, and complex. Because they are practically costless to reproduce and update, businesses use them more frequently and alter them often.” Nancy S. Kim, Adhesive Terms and Rea- sonable Notice , 53 Seton Hall L. Rev. 85, 91 (2022). This dy- namic “creates the potential for unfair surprise, particularly when those terms are at odds with longstanding consumer *31 expectations and with contract law’s default rules.” Mark A. Lemley, The Benefit of the Bargain , 2023 Wisc. L. Rev. 237, 266 (2023).
Digital terms of service are contracts of adhesion. They may be described as “click-wrap” or “browse-wrap” con- tracts (or hybrids), in which a retailer states only that there are terms and conditions that govern a transaction and that by completing the purchase, you accept those terms. See Kim, 53 Seton Hall L. Rev. at 96 (“Online, where there is no option to sign, courts have determined that a ‘manifestation of consent’ could be something other than a signature; it could be a click on the icon that expresses acceptance.”). The contract pre- sented in this case is a “browse-wrap.” The relevant infor- mation is not contained in the notice itself. Instead, the “notice provides no information other than that legal terms are availa- ble.” Id. at 104. The consumer then has the burden of seeking out “terms behind links,” and those links “do not necessarily signal whether the hidden information is important or triv- ial.” Id.
Under these conditions, what happens to the meeting of the minds and manifestations of mutual assent that are the foundations of contract law? Judge Brennan’s opinion for the court accurately reflects the state of the emerging case law, but those cases deserve rethinking.
Under the emerging case law, where the merchant uses a
design that prevents customers from completing transactions
without affirmatively “clicking” that they agree to the linked
terms, courts generally enforce those terms. The so-called
“click-wrap” design requires customers to decide whether
they want to spend the time to review those terms. If they de-
cide not to—usually a perfectly reasonable decision—courts
*32
treat that choice like signing a contract without reading it.
They enforce the available but unread terms of such contracts.
See
Sgouros v. TransUnion Corp.
,
The problem here is different, going under the rubric of a “browse-wrap” user interface. See, e.g., Oberstein , 60 F.4th at 513. Menards, like many other merchants, does not require customers to “click” that they accept the terms. Instead, Menards puts a hyperlink somewhere in the screens a cus- tomer will encounter along the path to making a purchase. Maybe the customer will notice, maybe not. When the cus- tomer completes the transaction, has she, in the language of contract law, manifested her assent to those terms? And how should courts decide?
The majority answers with what it calls “a fact-intensive
legal analysis,” ante at 9, citing
Sgouros
,
To see the problem with this approach, take a look at the
two Uber payment screenshots below, taken from cellphones.
Each had a hyperlink for “Terms of Service and Privacy Pol-
icy,” and each hyperlink would lead a user to terms that in-
cluded a mandatory arbitration clause. But each used a
“browse- wrap” technique, not requiring affirmative assent
with a click. One circuit found that one screen provided users
with reasonable notice that they were agreeing to a manda-
tory arbitration clause by signing up for the service. A differ-
ent circuit found that the other did not. Compare
Cullinane v.
Uber Technologies, Inc.
,
These conflicting results for functionally identical inter-
faces are not surprising, given what courts have been doing
for almost a decade. We have been speculating about which
design features will put the hypothetical “reasonable person”
on “reasonable notice” that a digital purchase will bind him
to a long document of legal terms and conditions. See, e.g.,
Sgouros
,
I believe this emerging multifactor test takes courts far from the law, far from judges’ competence, and far from the practical realities of online commerce and user-interface de- sign and consumer behavior.
II. The Flawed Assumptions Underlying the Reasonable Notice Re-
quirement The multifactor test the majority applies today replaces the familiar requirement of actual mutual assent with one of “conspicuous notice” or “reasonable notice.” Nonetheless, it imports other assumptions from traditional contract law that do not reflect reasonable consumer behavior in online trans- actions. The result is that every day, many times a day, rea- sonable consumers give away key legal rights without realiz- ing they have done so.
Implicit in this “conspicuous notice” or “reasonable no- tice” standard is a “duty to read.” See Michael L. Rustad & Thomas H. Koenig, Wolves of the World Wide Web: Reforming Social Networks’ Contracting Practices , 48 Wake Forest L. Rev. *36 1431, 1453 (2014) (explaining courts have expanded the longstanding contract principle of a “duty to read” to digital consumer contracts of adhesion). Courts have generally held that if there is a sufficiently conspicuous notice that binding terms and conditions exist, we assume (or our reasoning pre- tends) that the reasonable consumer has read them (even if doing so would be more costly than the transaction itself) and has consented to them by completing the transaction.
A “duty to read” makes sense in a negotiated transaction
between sophisticated parties. But empirical studies show
this fundamental assumption does not hold true when it
comes to digital contracts of adhesion, particularly where
user interfaces are designed to discourage consumers from
reading them. As noted above in note 1, virtually no consum-
ers read the fine print. See Bakos, 43 J. Legal Stud. at 3 (0.2
percent of consumers looked at digital terms of software li-
cense agreement for more than one second); Marotta -
Wurgler,
I suspect that, as a policy matter, we do not actually want consumers to take the time to read every digital contract they make. For example, in the time it takes to read Microsoft’s terms and conditions for its ubiquitous Office suite of prod- ucts, you could instead read Shakespeare’s Macbeth . [3] One es- timate found it would take average Americans 250 hours each *37 to read all the digital contracts to which they agree in a given year. [4] Another study found it would take the equivalent of 76 eight-hour workdays to read all the digital privacy policies (usually separate documents incorporated by reference into digital terms of service) to which the average internet user has agreed in a given year. [5] If all U.S. consumers read all the dig- ital privacy policies they accept, one study estimated, it would cause a total productivity loss of $781 billion annually. See Aleecia M. McDonald & Lorrie Faith Cranor, The Cost of Read- ing Privacy Policies , 4 I/S J.L. & Policy Info. Soc’y 543, 563–65 (2008). Adjusted for inflation in 2024, that number would now be above $1 trillion, with a “t.”
Even assuming the “reasonable consumer” is willing to devote so much time to reading these contracts, that does not tell us whether consumers understand the terms to which they are binding themselves. In an empirical test of 500 hybrid- wrap digital contracts on popular websites and applications, 99.6% were written at a reading level above that of the average adult. See Uri Benoliel & Shmuel I. Becher, The Duty to Read the Unreadable , Bos. Col. L. Rev. 2255, 2275, 2278–79 (2019). Mi- crosoft’s “privacy statement” runs 190 pages in print, with many hyperlinks embedded in it. One study applied a stand- ard measure for reading difficulty and found that the privacy *38 policies on popular sites such as CNN, Zoom, Disney, Airbnb, and major league baseball were more difficult to read and un- derstand than Immanuel Kant’s notoriously dense Critique of Pure Reason . [6]
For an attorney or judge, spending an entire day parsing such complex legal language might be daunting but not im- possible. For the average consumer who just wants to place an online order to pick up that last tool or can of paint to finish a do-it-yourself project, the reading assignment is beyond un- reasonable.
The emerging legal standard is also based on an often- false assumption of choice and competition—that if a con- sumer is dissatisfied with a contract term, she can simply de- cline to complete a transaction and take her business else- where. See Jacques Crémer et al., Fairness & Contestability in the Digital Markets Act , 40 Yale J. Reg. 973, 986 (2023) (“In to- tally competitive markets, with totally rational consumers with no cognitive limitations, … consumers would read all the terms of the contracts proposed by the different suppliers and would be able to carry out a well- founded cost benefit analysis among the different offers. In practice, no consumer can do so.”). For a modest consumer transaction like plaintiff Domer’s, this assumption seems very distant from real con- sumer behavior.
This assumption also ignores the reality of market concen- tration— not just in specific industries, but with regard to the *39 terms typically included in these digital contracts. See id . at 980 (digital gatekeepers “have gained the ability to easily set commercial conditions and terms in a unilateral and detri- mental manner for their … end users.”). [7] See also Lemley, su- pra, at 262 (“Take it or leave it may also not be a practical op- tion if the industry has coalesced around similar contract terms.”); and Thomas D. Haley, Illusory Privacy , 98 Ind. L. J. 75, 79– 80 (2022) (finding in survey of 222 platform terms from major websites and smartphone apps that most key terms were shared, such as allowing the platform to change the terms of the agreement unilaterally without consumer con- sent). [8]
III. Real Consumers and the Science of User-Interface Design
Given what we know about how a reasonable consumer interacts with digital contracts of adhesion, the threshold question of fair notice takes on heightened importance. The literature reviewed in Part II casts doubt on whether it is ever reasonable to enforce digital contract terms against ordinary *40 consumers. Enforcing those terms is doubly unreasonable if we cannot even be sure that consumers know that they have agreed to such terms.
Reasonable notice is produced through the interaction of real consumer behavior and user-interface design choices. Plenty of data shows that (a) almost no consumers even glance at the fine print on websites, let alone click through links to review the terms and conditions, and (b) companies design their interfaces carefully to influence consumer behav- ior to maximize profits. Neither point is surprising. But they show that our “reasonable notice” cases on whether these in- terface designs result in binding con tract terms bear little re- lation to actual consumer behavior or to the traditional con- tract principle of mutual assent.
Because user-interface design is technical and empirically testable, it should be possible for judges or juries to evaluate evidence as to whether or not real-life consumers are on fair notice that they have agreed to a long list of legal terms every time they complete a purchase online or download a new ap- plication on their devices. At heart, the way a reasonable con- sumer responds to a particular user interface is a factual issue. A plaintiff willing to present the necessary evidence may be able to show genuine disputes of material fact that can then be tried before a jury.
A. The Realities of the “Reasonable Consumer” As we explained recently:
R easonable consumer behavior is not a matter of pure economic theory. Rather, reasonable consumer behaviors are “ matters of fact, subject to proof that can be tested at trial, even if as *41 judges we might be tempted to debate and spec- ulate further about them.” In establishing rea- sonable consumer behavior, what matters is “how consumers actually behave—how they perceive advertising and how they make deci- sions.”
Kahn v. Walmart Inc.
,
“‘[W]e have often stressed that consumers are likely to ex-
hibit a low degree of care when purchasing low-priced, eve-
ryday items.’ This low degree of care does not make consum-
ers unreasonable—it makes them human, and even economi-
cally rational when search costs and transaction costs are in-
cluded in the utility calculus.”
Kahn
,
Predictable tendencies in consumer behavior mean that retail settings can be engineered to in- fluence consumers in ways they (meaning we) do not fully anticipate or appreciate. “‘[M]arket outcomes frequently will be heavily influenced, if not determined, by the ability of one actor to control the format of information, the presenta- tion of choices, and, in general, the setting *42 42
within which market transactions occur,’ allow-
ing some to ‘exploit those tendencies for gain.’”
Kahn
,
These predictable tendencies in consumer behavior mean that a reasonable consumer may be more likely to overlook the existence of terms than a judge engaged in a careful exam- ination of a single user interface. Extrapolating from our own experiences as busy consumers risks undervaluing the uncon- scious processes and conditions that shape consumer deci- sion-making.
B. User-Interface Design and Consumer Behavior Do these realities matter? The businesses seeking to en- force digital contracts certainly believe they do. Just as brick- and-mortar supermarkets stock higher-margin items at eye- level to nudge shoppers into spending more, see Kahn , F.4th at 596 n.3, online retailers strategically place and format certain information (while obscuring other information) to in- fluence consumer behavior.
User-interface design experts focus carefully on how the details of a digital interface influence users’ interactions with it. See Jamie Luguri & Lior Jacob Strahilevitz, Shining a Light on Dark Patterns , 13 J. Legal Analysis 43, 48 (2021) (“e-com- merce[] firms run thousands of consumers through identical interfaces at a reasonable cost and see how small software tweaks might alter user behavior”); Colin Gray et al., The Dark (Patterns) Side of UX Design , In Proceedings of the 2018 CHI *43 Conference on Human Factors in Computing Systems (CHI ’18), Association for Computing Machinery, Paper 534, 2 (2018) (user-interface design is “inherently a persuasive act”).
Research can measure how various features change the likelihood of consumers taking certain actions when using a website or application. “Choice architects have long under- stood that contrasting visual prominence can be used to nudge choosers effectively into a choice the architect prefers.” Luguri & Strahilevitz, 13 J. Legal Analysis at 55. For example, one study found that the most effective design strategies in- clude “hidden information” and “smaller print in a less visu- ally prominent location.” Id. at 47. In a controlled study, such a design choice doubled the number of consumers willing to complete a purchase. Id. at 73–75. The authors of the study concluded that such design choices “might cause consumers to misunderstand material terms of a contractual offer they have just accepted.” Id. at 94; see also Gray et al., supra, at 7 (explaining that the “primary motivator behind hidden infor- mation is the disguising of relevant information as irrele- vant”). These authors could have been describing Menards’ design and placement of its hyperlink to the terms and condi- tions at issue here.
If businesses like Menards invest in designing their user
interfaces “to influence the behavior of real consumers,”
courts should not allow them to enforce hidden contract
terms with arguments that “assume consumer behavior in
idealized markets with ‘
perfect
information,
perfect
competi-
tion, and
no
transactions costs.’”
Kahn
,
These empirical studies on consumer behavior in response to a particular user interface are exactly the type of evidence an expert could present to a jury to show that a consumer did or did not have notice of the terms and conditions found on a digital website. The evidence might show, for example, that the interface has been designed to distract or mislead the con- sumer, or at least to manipulate the consumer into proceeding without bothering to look at the terms of the contract of adhe- sion.
IV. The Multifactor Test — Law or Fact?
As noted, the majority enforces the arbitration clause in Menard’s terms and conditions by applying “a fact-intensive legal analysis” that considers (1) the simplicity or clutter of the screen; (2) the clarity of the disclosure; (3) the size and col- oring of the disclosure’s font, in the context of the rest of the screen; (4) the “spatial placement” of the hyperlink; and (5) the temporal relationship between the hyperlink and the steps the customer must take to complete the transaction. Ante at 10.
I have been tempted to disagree with the majority’s appli- cation of this multifactor test, which requires the decision- maker to study high-resolution screenshots of various user in- terfaces. On further consideration, though, I concluded that such a debate would not be useful. First, plaintiff Domer chose to litigate this case without trying to raise material fac- tual disputes about fair notice. That’s why I concur in the judgment here. Second, and more fundamental, the prospect of debates among judges about such details of a user interface as font size and color and the exact placement and labels for hyperlinks to incomprehensible fine print shows , in my view, that we need to think about these problems in a different way, *45 one that is more consistent with older approaches to contract formation.
As Justice Kagan said recently, “we’re a court. We really
don’t know about these things. You know, these are not like
the nine greatest experts on the internet.” Transcript of Oral
Argument at 45–46,
Gonzalez v. Google
,
Rather than debating among ourselves our impressions about font size and color, the placement of hyperlinks, and the choice between click-wrap and browse-wrap agreements, we should start treating these issues about user-interface de- sign as questions of fact. We should invite sufficiently moti- vated parties to test alternative designs and to conduct dis- covery into merchants’ actual design choices. If judges will not do so, then legislators and/or consumer-protection agen- cies should consider stepping in with m ore specific disclosure requirements and regulatory safe harbors.
I recognize that there may be room to debate whether the contract terms in online commerce result from a race to the bottom or a race to the top. Proponents of the browse -wrap interface design might well assert that busy consumers prefer not to be bothered with detailed contract terms they don’t care about, and that competitors’ common terms and conditions reflect an evolution toward an efficient solution to common issues. Those assertions may be true, but they are assertions of fact, subject to empirical p roof and rebuttal.
Advocates of treating these problems as questions of law might also argue that treating them as questions of fact would invite expensive and inconclusive battles of experts. Online merchants would be left guessing how to design their user- interfaces to secure enforceable terms. To some extent, those risks are a tradeoff chosen by online merchants who wish to bind consumers to digital contracts of adhesion through browse-wrap agreements. Merchants could avoid much of the uncertainty of litigation by using click-wrap agreements, which are usually enforced.
Moreover, those risks do not change the inherently factual nature of the questions. Nor do those risks justify having judges apply their own experience as consumers to decide factual questions that, in the real world of online commerce, are the subject of genuinely expert study and design. If a plaintiff is willing to invest in the needed expertise, courts and juries should listen . Another alternative, more likely to offer clearer rules, would be for legislatures or consumer-protec- tion agencies to establish more specific guideline s for enforc- ing terms and conditions for consumer transactions. Legisla- tures and agencies are also better positioned to craft rules that address the false assumptions underlying the reasonable no- tice requirement.
In any event, the foundation of contract law is mutual as- sent. How can there be a meeting of the minds if the average person has (quite reasonably) never seen, let alone read and understood, the contract terms? Or if consumers are function- ally deprived of an opportunity to opt out of the binding terms? It would be virtually impossible to navigate our mod- ern world without binding yourself to click-wrap and browse-wrap contracts, yet empirical data show that only a *47 vanishingly small segment of consumers can give anything approaching meaningful consent to their terms. Courts should account for those realities in deciding these disputes over contract terms.
V. Jury Trials on Contract Formation under the FAA
Finally, this reasoning points toward trying before juries
whether to enforce hyperlinked arbitration clauses in online
contracts of adhesion. Under the Federal Arbitration Act,
courts must look to state contract law to determine whether
such an agreement exists and is valid. See
Tinder v. Pinkerton
Security
,
“The burden is on the party opposing arbitration to “iden-
tify a triable issue of fact concerning the existence of the agree-
ment in order to obtain a trial on the merits of the contract.”
Tinder
, 305 F.3d at 735. Whether there is a meeting of the
minds to form a contract is generally a question of fact under
Wisconsin contract law. See
National Steel Serv. Ctr., Inc. v.
Wollin Silos & Equip., Inc.
,
Ultimately, however, b ecause plaintiff Domer did not try to raise any genuine issues of material fact here, I concur in the judgment affirming dismissal of her case .
Notes
[1] Menards is owned by the defendant Menard, Inc. For clarity this opinion uses “Menards” throughout, as did the district court.
[2] The district court had jurisdiction under the Class Action Fairness Act of 2005 (codified at 28 U.S.C. § 1332(d)). Domer is a citizen of Indiana. Menard, Inc. is incorporated and maintains its principal business offices in Wisconsin. Menards also regularly conducts business and operates store locations in Indiana. There are more than one hundred members in the putative class and the amount in controversy is greater than $5,000,000. We have jurisdiction over the appeal of the district court’s final judgment pursuant to 28 U.S.C. § 1291.
[3] Oral Argument, 5:30–5:40.
[4] See Domer’s Amended Class Action Complaint, at ¶¶ 48, 52 (citing $1.40 processing fee as an expenditure she would have avoided but-for the alleged price deception); id. at ¶ 52 (same for the purchase itself); id. at ¶ 63 (describing the claimed pecuniary losses); id. at ¶¶ 60, 63 (alleging that the unjust enrichment conferred on Menards is the $1.40 fee).
[1] See Yannis Bakos et al., Does Anyone Read the Fine Print? Consumer Attention to Standard-Form Contracts , 43 J. Legal Stud. 1, 3 (2014) (finding that 0.2 percent of consumers looked at digital terms of software license agreement for more than one second); Florencia Marotta-Wurgler, Will In- creased Disclosure Help? Evaluating the Recommendations of the ALI’s “Princi- ples of the Law of Software Contracts” , 78 U. Chi. L. Rev. 165, 179–81 (2011) (enhanced disclosure of terms via a “click-wrap” contract increased con- sumer reading by only 0.36% as compared to a “browse-wrap” contract).
[2] Spoiler alert: The image on the left is from
Cullinane
,
[3] Nicholas LePan, Visualizing the Length of the Fine Print, for 14 Popular Apps , Visual Capitalist (Apr. 18, 2020), https://www.visualcapital- ist.com/terms-of-service-visualizing-the-length-of-internet-agreements/ [https://perma.cc/VS83-32UM].
[4] David Berreby, Click to agree with what? No one reads terms of service, studies confirm , The Guardian (Mar. 3, 2017), https://www.theguard- ian.com/technology/2017/mar/03/terms-of-service-online-contracts-fine-print [https://perma.cc/PE2L-2LBH].
[5] Alexis C. Madrigal, Reading the Privacy Policies You Encounter in a Year Would Take 76 Work Days , The Atlantic (Mar. 1, 2012), https://www.theat- lantic.com/technology/archive/2012/03/reading-the-privacy-policies-you- encounter-in-a-year-would-take-76-work-days/253851/ [https://perma.cc/RY25-4E85].
[6] See Kevin Litman-Navarro, Opinion, We Read 150 Privacy Policies. They Were an Incomprehensible Disaster. , N.Y. Times (June 12, 2019), https://www.nytimes.com/interactive/2019/06/12/opinion/facebook- google-privacy-policies.html [https://perma.cc/29Y9-CZTZ].
[7] The Crémer article quotes in turn Council Regulation 2022/1925, 2022 O.J. (L. 265) 1, Recital (13) (“Regulation of the European Parliament and of the Council on Contestable and Fair Markets in the Digital Sector (Digital Markets Act)”).
[8] Consumers are increasingly likely to have contracted away their right to bring a consumer antitrust action against the very entity that sub- jected them to such binding terms. Crémer et al., 40 Yale J. Reg. at 979 (internal quotation omitted); see also Mark A. Lemley & Christopher R. Leslie, Antitrust Arbitration and Merger Approval , 110 Nw. Univ. L. Rev. 1, 46–47 (2015) (“Because firms with strong bargaining power can impose onerous terms on their business partners, a monopolist could make all sales contingent on the customer’s agreement to waive his or her right to sue for an antitrust violation in federal court, or to bring any form of class- wide arbitration.” (footnote omitted)).