A "free ride" or a $15 discount?

Uber asks existing users to refer friends — promising significant sign up bonuses to both new and existing users for each referral. First, the existing user activates the Free Rides function in the Uber app, revealing the offer and a code that the new user must enter so Uber can track the referral. Quoting from the first screenshot at right (emphasis added):

Share your invite code [placeholder]. Send friends free rides and you’ll get one too, worth $15! Details. INVITE FRIENDS.

A user who taps "Details" sees two additional sentences (quoting from the second screenshot at right):

Share your promo code with friends and they’ll get their first ride free. Once they’ve tried Uber, you’ll automatically get a free ride (up to $15) the next time you use Uber. CLOSE.

Neither screen provides any menu, link, button, or other command offering more details about other requirements or conditions. The text quoted above is the entirety of Uber’s offer.

Uber’s promise is clear — a "first ride free" for the new user, and a "free ride (up to $15)" for the existing user. But Uber’s actual practice is quite different. Most notably, the new user’s "free ride" is also limited to a $15 discount. One might ask whether the "worth $15" in the first screen applies to the friend’s free ride or to the existing user’s free ride or perhaps both. But the Details screen leaves no doubt that this limitation applies only to the existing user. Notice the placement of the "up to $15" parenthetical only in the sentence describing the existing user’s free ride. In contrast, the separate sentence about the new user’s ride promises "their first ride free" with no indication of any maximum value.

These discrepancies create unfortunate surprises for typical Uber customers. Consider the standard workflow. An existing Uber user tells a friend about Uber and, in person, helps the new user install the app and create an account, including entering the existing user’s referral code when prompted. "You’ll get a free ride," the existing user explains, guided by Uber’s simple on-screen offer. The new user then takes an expensive ride; expecting the ride to be free (as promised), the new user might choose Uber for a distance that would otherwise have been a better fit for a train or bus, or the new user might accept a high surge multiplier. Only later does the new user learn that according to Uber, "free" actually meant a $15 discount. The user would have no written evidence of Uber’s "free ride" promise, which was conveyed orally by the existing user. So the new user is unlikely to complain — and my experience (detailed below) indicates that a request to Uber is unlikely to get satisfaction.

I know about these problems because of an experience referring a friend — call her my cousin — in January 2016. I told her she’d get a free ride, but her receipt showed no such benefit. In fact she took her first ride in another country, which prompted me to check for other discrepancies between Uber’s marketing statements and its actual practice.

Piecing together statements from Uber’s support staff and the various disclosures in Uber’s Android, iOS, and mobile web apps, I found five separate restrictions that were not mentioned mentioned anywhere in Uber’s new user offer as presented to existing Android users:

• The new user’s credit only applies to a ride in the country that Uber designates as the new user’s home country or the currency that Uber designates as the new user’s home currency. But Uber’s signup page doesn’t ask about a user’s home country or currency. As best I can tell, Uber automatically sets a user’s home country based on the user’s IP address or location at the time of signup. (Source: Uber staff indicated that "the promo is currency specific" in explaining why my cousin received no discount on her first ride.)
• The existing user’s credit can only be redeemed towards a ride in the country that Uber designates as the existing user’s home country. (Source: Uber’s iOS app, GET FREE RIDES offer, secondary disclosure screen, stating that "discounts apply automatically in your country," emphasis added.)
• The new user’s maximum ride value varies by country. Not only is there a maximum value (contrary to the simple "first ride free" in Uber’s second screen above), but the maximum value is not mentioned to the existing user. (Source: Uber’s iOS app, GET FREE RIDES offer, secondary disclosure screen; and mobile web, new user offer, page footer.)
• All discounts expire three months from issue date. (Source: Uber’s iOS app, GET FREE RIDES offer, secondary disclosure screen.)
• Offer is not valid for UberTaxi. (Source: Uber’s iOS app, GET FREE RIDES offer, secondary disclosure screen; and mobile web, GET FREE RIDES offer, page footer.)

The table below presents the Uber’s marketing offers in all three platforms, along with the errors I see in each:

	Android	iOS	Mobile Web
Primary disclosure	FREE RIDES. Share your invite code [placeholder]. Send friends free rides and you’ll get one too, worth $15! Details. INVITE FRIENDS.	GET FREE RIDES. They get a free ride and you will too (worth up to $15), after their first ride. Details.	GET FREE RIDES. Sign up now to claim your free gift from [placeholder] ($15 off first ride)*.
Secondary disclosure	Share your promo code with friends and they’ll get their first ride free. Once they’ve tried Uber, you’ll automatically get a free ride (up to $15) the next time you use Uber. CLOSE.	Every time a new Uber user signs up with your invite code, they’ll get their first ride free (value amounts vary by location). Once they take their first ride, you’ll automatically get your next ride free (worth up to $15). Discounts apply automatically in your country and expire 3 months from their issue date. Offer not valid for uberTAXI.	Every time a friend signs up with your invite code, they’ll get their first ride free (value amounts vary by country). Once they use it, you’ll automatically get a free ride to use the next time you Uber (worth up to $15). Offer not valid for UberTaxi.
Errors	New user’s ride is actually limited to $15 (or other amounts in other countries). In contrast, both disclosures indicate that there is no limit to the value of the new user’s ride. Discount only applies to a new user’s first ride in the user’s home country as determined by Uber. Existing user’s discount can only be redeemed in existing user’s home country. Discounts for both the new and existing user expire three months from issue date. Offer is not valid for UberTaxi.	Secondary disclosure plausibly contradicts primary disclosure: Primary disclosure promised "a free ride" for the new user, while secondary disclosure retracts "free ride" and instead offers only a discount. In contrast, FTC rules allow secondary disclosures only to clarify, not to contradict prior statements. Discount only applies to a new user’s first ride in the user’s home country as determined by Uber.	Discount only applies to a new user’s first ride in the user’s home country as determined by Uber. Existing user’s discount can only be redeemed in existing user’s home country. Discounts for both the new and existing user expire three months from issue date.

I first alerted Uber staff to these discrepancies in January 2016. It was a difficult discussion: My inquiries were bounced among four different support representatives, with a different person replying every time and no one addressing the substance of my messages. So I reluctantly gave up.

Six weeks later, a different Uber rep replied out of the blue. He seemed to better understand the problem, and I managed to get two separate replies from him. At my request, he committed to "pass this along to the appropriate team." That said, he did not respond to my repeated suggestion that Uber needed to refund affected consumers.

Eight weeks after my final correspondence with the fifth Uber representative and sixteen weeks after I first alerted Uber to the problem, I see no improvement. Uber’s Android app still makes the same incorrect statements about promotion benefits, verbatim identical to what I observed in January.

Credit on your "next trip" — or later, or not at all

In a variety of circumstances, Uber responds to customer complaints by issuing credits towards a customer’s "next trip." For example, during a recent attempt to ride with Uber in Mexico, I was unable to find or contact the driver. (I was where the on-screen pushpin told me to be, and GPS seemingly told the driver he was where he was supposed to be, but we just couldn’t find each other.) I later received an email receipt showing that I had been charged a cancellation fee. In Uber’s "Help" area, I used Uber’s "I was charged a cancellation fee" feature, and I was immediately advised (emphasis added):

We’ve credited your Uber account. Thanks for letting us know what happened. A credit has been added to your Uber account. This credit will apply to your next trip.

Imagine my surprise when, upon returning to the US a few hours later, I took another Uber ride but received no such credit.

It seems Uber’s notion of credits is in fact country-specific or currency-specific. My problem resulted from difficulty finding a driver in Mexico, where I don’t live and rarely travel. Far from applying the credit on my "next trip," it seems Uber’s systems will carry the credit forward to my next journey in Mexico. (See Uber’s Payment screen, "Credit Balances" section, showing the amount of the cancellation fee as a credit in the currency associated with the country where that fee was charged.) But this is much less useful to users traveling internationally. For example, Uber might impose a time limit on the credit — analogous to Uber’s undisclosed three month limit for use of a "free ride" credit (as revealed in the preceding section). And some users may never return to (or never again take Uber rides in) certain countries or currencies. The plain language of "your next trip" of course purports to protect users against all these contingencies; "your next trip" means a trip denominated in any currency, perhaps soon but perhaps indefinitely in the future. Uber’s actual practice is plainly less favorable to consumers.

Here too, predictable consumer responses increase the harm from Uber’s approach. If a consumer was charged improperly and felt Uber’s response was out of line, the consumer might pursue a credit card chargeback. But when Uber tells the consumer "We’ve credited your Uber account" and "This credit will apply to your next trip," there’s every reason to think the problem is completely fixed. Then the consumer may forget about the problem; certainly the consumer is less likely to diligently check future Uber receipts for a credit that was slated to be automatic and guaranteed. In addition, consumers are vulnerable to the passing of time: If a consumer rides with Uber only occasionally, the permissible time for a chargeback may have elapsed by the time the consumer’s next ride.

Update (May 25, 2016): Four weeks after I reported this discrepancy to Uber, I received a reply from an Uber representative. He confirmed that I did not receive the promised credit for the general reason I described above — a credit provided in one currency, while my next trips was in a different currency. I reminded him that Uber’s statements to users say nothing of any such restriction. I also pointed out that Uber is capable of converting currencies, and I encouraged him to assure that other users, similarly situated, are appropriately refunded. So far as I know, Uber has not done so.

Others’ reports

Checking with friends and colleagues, and receiving further reports from strangers, I’ve learned about a fair number of other Uber billing errata. For example, one user confidently reported that when a driver cancels a ride — perhaps seeing a surge in another app, getting lost, learning that the passenger’s destination is inconvenient, or just changing his or her mind — Uber still charges the passenger a cancellation fee. I haven’t been able to verify this, as I don’t have an easy way to cause a driver to cancel. But in the Uber help tool, the "I was charged a cancellation fee" menu offers as one of its preset reasons for complaint "My driver cancelled" — confirming that Uber’s systems charge cancellation fees to passengers when drivers cancel. Of course Uber’s systems can distinguish who pressed the cancel button, plus Uber could ask a driver the reason for cancellation. I see no proper reason for Uber ever to charge a passenger a cancelation fee if it is the driver who elected to cancel.

Users with experience with this problem, or other Uber contracting or billing errata, should feel free to contact me. I’ll add their experiences here or in a follow-up.

Update (June 4): Readers alerted me to UberPool drivers repeatedly charging for two passengers purportedly riding, when only one passenger was actually present, increasing the charge to passengers.

Update (June 4): In federal litigation against Uber, blind passenger Tiffany Jolliff reports that not only did multiple Uber drivers refuse to transport her and her service dog, but Uber charged her a cancellation fee each time a driver refused to transport her.

Lessons from Uber’s billing errors

I see four distinct takeaways from these billing errors. First, Uber’s engineering, testing, and legal teams need to sharpen their focus on billing, promotions, and credits. The coding of a promotional offer is inextricably linked to the marketing text that describes the offer. Similarly, the coding of a customer service benefit must match the text that explains the benefit to users. Both should be checked by attorneys who specialize in advertising law and consumer protection. Instead, in the problems I described here, Uber’s billing logic seems to be entirely separate from the text presented to consumers. It is particularly striking to see Uber’s three separate textual descriptions of the new user promotion — all three of them incorrect, yet in three different ways. Even a basic attorney review would have flagged the discrepancies and identified the need to inquire further. An advanced attorney review, fully attuned to FTC disclosure rules, might also question what appears in the primary disclosure versus the secondary disclosure. Attorneys might reasonably caution Uber against repeatedly and prominently promising "FREE RIDES" when the company’s actual offer is a discount.

Second, Uber’s overcharging is both large and long-lasting. I reported the new user promotion problems in January 2016, although they probably began considerably earlier. (Perhaps Uber will respond to this article by determining, and telling the public, when the problems began.) In response to this article, I expect that Uber will fix these specific problems promptly. But given Uber’s massive operations — many thousands of new users per month — the aggregate harm is plausibly well into the millions of dollars.

Third, my experience calls into question whether Uber’s customer service staff are up to the task of investigating or resolving these problems. Writing in to customer service is fruitless; even with screenshots proving the discrepancy in the new user promotion, Uber was slow to promise a refund to match the marketing commitment. (It took five separate messages over eight weeks!) In fact, even after promising the refund in a message of March 16, 2016, that refund never actually occurred. Similarly, I promptly alerted Uber to the "next ride" credit not provided — but ten days later, I have received neither the promised credit nor any reply. Others have reported the shortfalls of Uber’s customer service staff including ineffective responses, a focus on response speed rather than correctness, and insufficient training. My experience suggests that these problems are genuine and ongoing. Users with the most complicated problems — Uber system flaws, discrepancies between records, billing errors — appear to be particularly unlikely to achieve resolution.

Finally, users lack meaningful recourse in responding to Uber’s overcharges. In each of the problems I found, Uber is overcharging a modest amount to each of many thousands of customers. Ordinarily, this would be a natural context for class action litigation, which would allow a single judge and a single set of lawyers to figure out what happened and how to set things right. But Uber’s Terms and Conditions purport to disallow users to sue Uber at all, instead requiring arbitration. Furthermore, Uber claims to disallow group arbitration, instead requiring that each consumer bring a separate claim. That’s inefficient and uneconomical. Uber’s arbitration clause thus provides a de facto free pass against litigation and legal remedies. Of course many companies use arbitration to similarly immunize themselves against consumer claims. But Uber’s controversial activities, including the overbilling described here among many other disputes, give consumers extra reason to seek judicial oversight.

Next steps

Just last week, Uber formed a paid advisory board of ex-regulators, most with competition and consumer protection expertise. These experts should exercise independent judgment in looking into the full breadth of Uber’s problems. I doubt overbilling was previously on their agenda, but my experience suggests it should be. To investigate, they might review all customer service threads with five or more messages, plus look for all messages attaching screenshots or mentioning "overcharge" or "promised" or "contract." Tthey shouldn’t rely merely on Uber staff summaries of customer experience; with an advisory board of superstars, the group should bring independent judgment to assessing and improving the company’s approach.

Meanwhile, Uber’s response should include a full refund to each and every user who was overcharged. For example, when Uber promised a "free ride" to an Android user who in turn referred a friend, Uber should provide the friend with a refund to achieve exactly that — not just the $15 discount that may be what Uber intended but isn’t what the offer actually said. Since Uber may be disinclined to offer these refunds voluntarily, I’m alerting consumer protection and transportation regulators in appropriate jurisdictions to help push to get users what they were promised.

We examine online affiliate marketing programs in which merchants oversee thousands of affiliates they have never met. Some merchants hire outside specialists to set and enforce policies for affiliates, while other merchants ask their ordinary marketing staff to perform these functions. For clear violations of applicable rules, we find that outside specialists are most effective at excluding the responsible affiliates, which we interpret as a benefit of specialization. However, in-house staff are more successful at identifying and excluding affiliates whose practices are viewed as “borderline” (albeit still contrary to merchants’ interests), foregoing the efficiencies of specialization in favor of the better incentives of a company’s staff. We consider the implications for marketing of online affiliate programs and for online marketing more generally.

We examine how unfavorable social comparisons differentially spur employees of varying hierarchical levels to engage in deception. Drawing on literatures in social psychology and workplace self-esteem, we theorize that negative comparisons with peers could cause either junior or senior employees to seek to improve reported relative performance measures via deception. In a first study, we use deceptive self-downloads on SSRN, the leading working paper repository in the social sciences, to show that employees higher in a hierarchy are more likely to engage in deception, particularly when the employee has enjoyed a high level of past success. In a second study, we confirm this finding in two scenario-based experiments. Our results suggest that longer-tenured and more successful employees face a greater loss of self-esteem from negative social comparisons and are more likely to engage in deception in response to reported performance that is lower than that of peers.

Online advertising might seem to be the most measurable form of marketing ever invented. Comprehensive records can track who clicked what ad–and often who saw what ad–to compare those clicks with users’ subsequent purchases. Ever-cheaper IT makes this tracking cost-effective and routine. In addition, a web of interlocking ad networks trades inventory and offers to show the right ad to the right person at the right time. It could be a marketer’s dream. However, these benefits are at most partially realized. The same institutions and practices that facilitate efficient ad placement can also facilitate fraud. The networks that should be serving advertisers have decidedly mixed incentives, such as cost savings from cutting corners, constrained in part by long-run reputation concerns, but only if advertisers ultimately figure out when they’re getting a bad deal. Legal, administrative, and logistical factors make it difficult to sue even the worst offenders. And sometimes an advertiser’s own staff members prefer to look the other way. The result is an advertising system in which a certain amount of waste and fraud has become the norm, despite the system’s fundamental capability to offer unprecedented accountability.

Google’s Advertising Labels in 2014

How does online advertising become less effective than advertisers expect and less effective than measurements indicate? The current research explores problems that result, in part, from malfeasance by outside perpetrators who overstate their efforts to increase their measured performance. In parallel, similar vulnerabilities result from mistaken analysis of cause and effect–errors that have become more fundamental as advertisers target their advertisements with greater precision. In the paper that follows, the author attempts to identify the circumstances that make advertisers most vulnerable, notes adjusted contract structures that offer some protections, and explores the origins of the problems in participants’ incentives and in legal rules.

In an online experiment, we measure users’ interactions with search engines, both in standard configurations and in modified versions with clearer labels identifying search engine advertisements. In particular, for a random subset of users, we change “Sponsored links” or “Ads” labels to instead read “Paid Advertisements.” Relative to users receiving the “Sponsored link” or “Ad” labels, users receiving the “Paid Advertisement” label click 25% and 27% fewer advertisements, respectively. Users seeing “Paid Advertisement” labels also correctly report that they click fewer advertisements, controlling for the number of advertisements they actually click. Results are most pronounced for commercial searches and for vulnerable users with low education and little online experience.

In today’s posting, Paul Kominers and I examine leading online apartment search services and evaluate the disclosures associated with their paid listings. We find paid placement and paid inclusion listings at each site, but disclosures range from limited to nonexistent. Where disclosures exist, they are largely hidden behind multiple intermediate pages, effectively invisible to most users. We propose specific ways these sites could improve their disclosures, and we flag their duties under existing law.

Advertising Disclosures in Online Apartment Search

While Google has certainly profited from selling advertisements to rogue pharmacies, that’s just one of many areas where Google sells unlawful advertisements. Here are six other areas where I’ve also seen widespread unlawful AdWords advertisements:

• Advertisements charging for something that’s actually free. I’ve documented scores of AdWords advertisements that attempt to trick users into paying for software that’s widely available for free — charging for RealPlayer, Skype, WinZip, and more.
• Advertisements promising “free” service but actually imposing a charge. I have also flagged dozens of advertisements promising “100% complimentary” “free” “no obligation” service that actually comes with a monthly charge, typically $9.99/month or more. Promising “free” ringtones, these services rarely ask users for their credit card numbers. Instead, they post charges straight onto users’ mobile phone bills — combining carrier-direct billing with deceptive advertising claims in order to strengthen the illusion of “free” service.
• Copyright infringement – advertisements touting tools for infringing audio and video downloads. For example, in 2007 media companies uncovered Google selling advertisements to various download sites, typically folks charging for Bittorrent clients. These programs helped users download movies without permission from the corresponding rights-holders, which is a double-whammy to copyright holders: Not only did labels, studios, artists, and filmmakers get no share of users’ payments, but users’ payments flowed to those making tools to facilitate infringement.
• Copyright infringement – advertisements touting counterfeit software. For example, Rosetta Stone in six months notified Google of more than 200 instances in which AdWords advertisers offered counterfeit Rosetta Stone software.
• Advertisements for programs that bundle spyware/adware. At the peak of the spyware and adware mess a few years ago, distributors of unsavory software used AdWords to distribute their wares. For example, a user searching for “screensavers” would receive a mix of advertisements — some promoting software that worked as advertised; others bundling screensavers with advertising and/or tracking software, with or without disclosure.
• Mortgage modification offers . Consumers seeking mortgage modifications often receive AdWords advertisements making deceptive claims. A recent Consumer Watchdog study found AdWords advertisers falsely claiming to be affiliated with the US government, requiring consumers to buy credit reports before receiving advice or help (yielding immediate referral fees to the corresponding sites), and even presenting fake certification logos. One prominent AdWords advertiser had previously faced FTC litigation for telemarketing fraud, while another faced FTC litigation for falsely presenting itself as affiliated with the US government. Other advertisers suffer unsatisfactory BBB ratings, and some advertisers falsely claim to have 501(c)(3) non-profit status.

Google’s Revenue from Deceptive Advertisements

Google does not report its revenues for specific sectors, so it is generally difficult to know how much money Google receives from particular categories of unlawful advertisements or from particular unlawful practices. That said, in some instances such information nonetheless becomes available. For example, the Wall Street Journal reported that Google charged $809,000 to one company advertising tools for unauthorized audio and video downloads. In 2006, I estimated that Google charged more than $2 million per year for advertisements distributing spyware and adware shown when users search for the single keyword “screensavers.” Scaling up to other keywords pushing spyware and adware, I suggested Google collects $25+ million per year for advertisements distributing spyware and adware.

Importantly, when AdWords advertisements deliver users into unlawful sites, the majority of the profits flow to Google. Consider a keyword for which several advertisers present similar unlawful offers. The advertisers bid against each other in Google’s auction-style advertising sales process — quickly bidding the price to a level where none of them can justify higher payments. If the advertisers are similar, they end up bidding away most of their profits. Indeed, most of these advertisers have low marginal costs, so their profit approaches their revenue, and Google even collects the majority of their revenue. In 2006 I ran an auction simulation to consider bidder behavior with 10 bidders and 20% standard deviation in per-click valuations; I found that in this situation, advertisers on average paid 71% of their revenue to Google. Drawing on litigation documents, the WSJ reports similar values: EasyDownloadCenter and TheDownloadCenter collected $1.1 million from users, but paid $809,000 (74%) to Google for AdWords advertising. Consumer Watchdog’s revenue estimates, drawn from Google’s own traffic estimation tools, reveal that an advertiser would need to pay more than $6 million per year to capture all the clicks from searches for “credit repair” and “bad credit.”

The Scope of Google’s Involvement — and Resulting Liability

Multiple sources have revealed Google’s far-reaching involvement in facilitating and supporting deceptive advertisements. For example, Google staff supplied EasyDownloadCenter and TheDownloadCenter with keywords to reach users, including “bootleg movie download,” “pirated,” and “download harry potter movie.” Similarly, plaintiffs in Goddard v. Google alleged that not only did Google show deceptive advertisements for “free ringtones” and similar searches, but Google’s own systems affirmatively suggested that advertisers target the phrase “free ringtones” (deceptive, since the advertisers’ service weren’t actually free) when advertisers requested only the word “ringtones.” Google’s involvement also extends to financing. For example, the WSJ reports that Google extended credit to EasyDownloadCenter and TheDownloadCenter — letting them expand their advertising effort without needing to pay Google in advance (as most advertisers must). In short, Google knew about these deceptive advertisements, profited from them, and provided assistance to the corresponding advertisers in the selection of keywords, in the provision of credit, and otherwise.

One might naturally expect that Google is liable when its actions cause harm to consumers — especially when Google knows what is occurring and profits from it. But the Communications Decency Act potentially offers Google a remarkable protection: CDA § 230 instructs that a provider of an interactive computer service may not be treated as the publisher of content others provide through that service. Even if a printed publication would face liability for printing the same advertisements Google shows, CDA § 230 is often interpreted to provide that Google may distribute such advertisements online with impunity. Indeed, that’s exactly the conclusion reached in Goddard v. Google, finding that even if Google’s keyword tools suggests “free ringtone” to advertisers and even if Google is aware of fraudulent mobile subscription services, Google is not liable to affected consumers.

The broad application of CDA § 230 immunity has attracted ample criticism. For example, a 2000 DOJ study concluded that “substantive regulation … should, as a rule, apply in the same way to conduct in the cyberworld as it does to conduct in the physical world.” Yet CDA § 230 unapologetically invites Google to show all manner of unlawful advertisements that would create liability if distributed by traditional publishers. And if Google can turn a blind eye to advertisers using its ad platform to defraud or otherwise harm users, advertisers will do so with impunity. For Google to escape liability is all the more puzzling when Google reaps most of the profits from advertisers’ schemes.

CDA § 230 includes several important exceptions. For example, the CDA does not immunize violations of criminal law — and rogue pharmacies implicate various criminal laws, giving rise to the liability for which Google now expects to pay $500 million. But this exception may be broader than critics yet realize. For example, large-scale copyright infringement and distribution of counterfeit goods may also create criminal liability for the underlying advertisers, hence excluding Google from the CDA safe-harbor for the corresponding advertisements.

Ultimately, I stand by my 2006 conclusion: “Google ought to do more to make ads safe.” Since then, Google’s revenue and profit have more than doubled, giving Google that much greater resources to evaluate advertisers. But I wouldn’t say Google’s users are twice as safe — quite the contrary, deceptive and unlawful advertisements remain all too widespread. Kudos to the Department of Justice for holding Google accountable for unlawful pharmacy advertisements — but there’s ample more work to be done in light of Google’s other unlawful advertisements.

When a user enters a search term that matches a company’s trademark, Google often shows results for the company’s competitors. To take a specific example: Searches for language software seller "Rosetta Stone" often yield links to competing sites — sometimes, sites that sell counterfeit software. Rosetta Stone think that’s rotten, and, as I’ve previously written, I agree: It’s a pure power-play, effectively compelling advertisers to pay Google if they want to reach users already trying to reach their sites; otherwise, Google will link to competitors instead. Furthermore, Google is reaping where others have sown: After an advertiser builds a brand (often by advertising in other media), Google lets competitors skim off that traffic — reducing the advertiser’s incentive to invest in the first place. So Google’s approach to trademarks definitely harms advertisers and trademark-holders. But it’s also confusing to consumers. How do we know? Because Google’s own documents admit as much.

Today Public Citizen posted an unredacted version of Rosetta Stone’s appellate brief in its ongoing litigation with Google. Google had sought to keep confidential the documents that ground district court and appellate adjudication of the dispute, but now some of the documents are available — giving an inside look at Google’s policies and objectives for trademark-triggered ads. Some highlights:

• Through early 2004, Google let trademark holders request that ads be disabled if they used a trademark in keyword or ad text. But in early 2004, Google determined that it could achieve a "significant potential revenue impact" from selling trademarks as keywords. (ref)
• In connection with Google’s 2004 policy change letting advertisers buy trademarks as keywords, Google conducted experiments to assess user confusion from trademarks appearing in search advertisements. Google concluded that showing a trademark anywhere in the text of an advertisement resulted in a "high" degree of consumer confusion. Google’s study concluded: "Overall very high rate of trademark confusion (30-40% on average per user) … 94% of users were confused at least once during the study." (ref)
• Notwithstanding Google’s 2004 study, Google in 2009 changed its trademark policy to permit the user of trademarks in advertisement text. Google estimated that this policy change would result in at least $100 million of additional annual revenue, and potentially more than a billion dollars of additional annual revenue. Google implemented this change without any further studies or experiments as to consumer confusion. (ref)
• Google possesses more than 100,000 pages of complaints from trademark holders, including at least 9,862 complaints from at least 5,024 trademark owners from 2004 to 2009. (ref)

Kudos to Public Citizen for obtaining these documents. That said, I believe Google should never have sought to limit distribution of these documents in the first place. In other litigation, I’ve found that Google’s standard practice is to attempt to seal all documents, even where applicable court rules require that documents be provided to the general public. That’s troubling, and that needs to change.