Given limited enforcement of restrictions on deceptive online advertising, numerous banner advertisers are willing to resort to trickery to draw attention to their offerings. And plenty of web site publishers stand ready to run deceptive ads. But widespread distribution of deceptive ads also requires another crucial input: intermediary ad networks to broker ad placement and tracking. Key among these intermediaries: Yahoo, whose Right Media marketplace facilitates so much of this traffic that Right Media classifies advertisers' deceptive tactics into dozens of detailed categories. Right Media's detailed ad examination and classification offers a natural opportunity to curtail deceptive online advertising -- but to date, Right Media has chosen instead to let deceptive ads run rampant through its system.

Right Media is the web's largest marketplace for "remnant" (otherwise-unsold) display advertising. Sellers of advertising space would typically rather sell their most desirable space elsewhere, where they can get higher prices. So Right Media is left with relatively less desirable ad inventory. Who wants to buy that space? In part, bottom-feeders whose ads have a tendency to deceive.

In the following sections, I briefly review relevant regulation of online advertising. I then provide examples of false and deceptive advertising running through Right Media: fake user interface ads, false claims of a buddy's supposed romantic interest presented in false widget-type design elements, false statements that "you have won", fake user interface ads using the word "free" in violation of applicable regulation, and miscellaneous other deceptive ads. I present Right Media's classification of deceptive ads, and I consider Right Media's own assessments of the prevalence of each kind of deception. Finally, I examine Right Media's response to this deception, and I suggest more effective alternatives.

Regulation of Online Advertising - Selected Requirements

Though online advertising may seem like a wild west, the FTC offers a set of rules -- embodied in both policy statements and in precedent -- that constrain certain advertising practices. The table below presents requirements relevant to the ads examined in subsequent sections.

Webrewardstream: fake user interface (FUI) banner ads that pretend to convey messages from Windows or from software already on a user's computer

In testing on December 28, 2008, I requested a page on cheatcodes.com. I received the display shown at right

At top-center, within the red box (added by me), Cheatcodes presents a "Message Alert" proclaiming "You have 1 message waiting for you." This message appears with substantially the same color scheme, fonts, layout, and alignment as genuine Windows XP (classic theme) message boxes. This message even shares the distinctive minimize, maximize, and close ("x") buttons common to most Windows applications, and the message's "OK" button matches key details of genuine Windows buttons (including shadow, an inner dotted outline indicating purported cursor focus, and an underlined letter indicating a purported shortcut key). Despite these many similarities to genuine Windows dialog boxes, in fact this message is just an advertisement that resorts to a "fake user interface" (FUI) design to attract a user's attention. In particular, the message is an ad from a company called Webrewardstream of Ontario, Canada.

The Webrewardstream ad was placed onto the Cheatcodes site by Kitara Media (of the Virgin Islands) and Right Media. (packet log)

A sophisticated user might recognize that this "Message Alert" is an ad based on its position on the Cheatcodes web page -- in a place where an ad might be expected to be. But a user unfamiliar with the Cheatcodes site is unlikely to know where that site's ads appear. Moreover, the ad's realistic appearance and tempting word choice are particularly likely to confuse hurried, distracted, or novice users.

Ordinarily, the minimize and "x" buttons serve to dismiss or hide an unwanted window. But on this Webrewardstream ad, pressing minimize or "x" leads to activating the offer, not closing the offer. Thus, the ad's design creates a heightened risk that users will activate the ad in the course of attempting to exit the ad.

The Webrewardstream ad's fake user interface design offers an unearned aura of trustworthiness. Reasonable users might rightly give additional credence to messages that come from programs installed on their computers, for messages from already-installed programs are likely to report actual conditions requiring a user's attention. For example, if software on a user's computer reports "you have 1 message waiting for you," it is highly likely that a user in fact has such a message. By combining the FUI design with a message that seems to be specific to a particular user, the ad creates an appearance of urgency when, in fact, the message is just an ad -- and when the user has actually received no special "message" other than an ordinary commercial offer presented to the general public.

The Webrewardstream ad's substance and format give no indication that the message is an advertisement. In fact, the message and format affirmatively indicate that the message is not an ad: "You have 1 message" is not ordinarily a commercial message, and a Windows message box is not a typical format for presenting a commercial offer. The ad thus violates the FTC's requirement that advertisements must not conceal their commercial nature.

Decisions of the BBB's National Advertising Division confirm the impropriety of the fake user interface tactics in this Webrewardstream ad. In 2001, the NAD examined Bonzi Buddy ads which, like this ad, appeared to be genuine Windows messages. The NAD described the practice then at issue: "a 468x60 ... banner advertisement designed to look like a Microsoft Windows error message that includes [a] blue active title bar [and] the familiar three boxes in the right corner, which purportedly allow the user to reduce, restore or close the window [with] a rectangular 'OK' box ... to the right of the text." The NAD cited the BBB's Code of Online Business Practices, requiring that "online advertisers should not disguise advertising as technical or desktop functionality when doing so would mislead customers into clicking on the advertisement thinking that they were actually performing a technical function." The NAD concluded that the banner then at issue appeared to be "something other than paid commercial advertising," and the NAD therefore recommended that Bonzi cease using banner ads that appear to be messages from a computer's operating system. NAD Case #3797 - Bonzi.com Software - July 12, 2001. The same reasoning applies equally to this ad.

If a user clicks any portion of the Webrewardstream ad, the user is taken through Kitara Media to Right Media and onwards to Webrewardstream. The resulting landing page claims a user can receive a "free Nintendo Wii" -- mentioning only in small-font text, "below the fold" (hence not immediately visible) on a standard 1024x768 screen, that receiving the promised product requires "meet[ing] eligibility requirements." Even then, Webrewardstream's landing page nowhere mentions the number of requirements or the number of sponsor offers a user must complete. Nor does the landing page list specific offer requirements or even provide examples of representative offers. These omissions fall far short of the FTC's requirements for lead generation landing pages.

I see ample basis for Right Media to reject this Webrewardstream ad: The ad deceptively appears to come from software already on a user's computer, and it is disguised to appear to be a genuine Windows message rather than an ordinary ad. The ad's minimize and "x" buttons do not function in the usual way. The ad falsely claims "you have 1 message waiting for you" when in fact the user has not received a bona fide message. The ad's landing page fails to disclose material terms, and its disclosures (such as they are) appear where few users are likely to notice. Finally, the ad and landing page fall far short of the FTC's requirements for lead generation services.

Tatto Media: Statements that are literally false; deception compounded through use of fake user interface ads

In testing on December 30, 2008, I requested a page on PerfSpot. I received the display shown at right:

Highlighted within a red box (added by me), the top of the page presents an advertisement created by Tatto Media of Seattle and Boston. The ad appears to be a part of the PerfSpot page: It matches the format, style, and color scheme of the PerfSpot site, and it even lacks a border to mark its boundary from PerfSpot. The ad's deception is compounded by its substance: a blue "Crush Center' banner, buttons labeled "(2) New Messages", "Sent Messages" , and "New Friend Requests", and a bubble reading "One of your buddies has a crush on you." Because both substance and format blur the boundary between web site and advertisement, the ad violates the FTC's requirement that advertisements must not conceal their commercial nature.

Despite the ad's claims , a user receiving this ad does not actually have two new messages, nor does such a have a buddy with a crush, any "new friend requests" or any "sent messages." Rather, these claims and buttons are simply a ruse designed to attract a user's attention.

The Tatto ad was placed onto PerfSpot's site through Right Media, which tracked and facilitated the placement. (packet log)

Clicking Tatto's ad takes a user through Right Media to Tatto Media, which attempts to tell a user a $9.99/month horoscope subscription service. In the course of Tatto's sales pitch, the deception of initial deception is compounded through its subsequent landing pages:

• Tatto conveys a false sense of urgency. Tatto claims that "your love is waiting" and shows a 20-second countdown timer. Tatto thereby encourages users to act quickly, without taking the time to read detailed disclosures of costs.
• Tatto falsely indicates that its message was prompted by a specific identifiable user. Tatto's reference to "your love" falsely suggests that a specific (albeit unnamed) person has expressed interest in the specific user receiving the ad, when that is not the case.
• Tatto's fee disclosures are hard to read and easy to overlook. Only some of Tatto's subsequent landing pages mention Tatto's fees in on-screen text. (Landing pages with disclosures: steps 1 and 4. Without disclosures: steps 2 and 3.) Furthermore, even when Tatto's fees are disclosed, they appear at the far edge of the landing page -- unlikely to catch a user's eye, especially in light of the sense of rush that Tatto's countdown timer causes. Moreover, the disclosures are strikingly hard to read against a patterned multicolor background.
• Tatto charges users through a mechanism that has a tendency to deceive. Tatto charges users through their cell phones -- never requiring that a user provide ordinary payment information (e.g. a credit card). Reasonable users may not realize that Tatto has the ability to place charges on their credit cards; users may think services are free of charge since no credit card number is requested.


Deceptive Messages When Users Try to Exit
• Tatto places its key disclosures where users are unlikely to notice them, in a format that makes them hard to read. Tatto's detailed fee description appears off-screen ("below the fold") on a 1024x768 PC. A user must scroll to see the full disclosure. Even there, the disclosures are particularly hard to read due to their color and font size, and the patterned multicolored background.
• Tatto makes further false claims to try to retain consumers who seek to leave. If a user attempts to leave Tatto's landing page, Tatto pops open a subsequent message: "CRUSH ALERT: One of your friends from Franklin has a crush on you. Enter your cell number to find out who." (See the first inset image at right.) This statement is also false: Tatto does not know who (if anyone) might have a crush on the user seeing Tatto's ad.
• Tatto imposes a further hindrance to exit through nonsensical dialog box instructions. If a user attempts to leave Tatto's landing page, Tatto attempts to dissuade the user through an exceptionally confusing JavaScript popup that asks: "Are you sure you want to navigate away from this page? PRESS 'CANCEL' BELOW TO CONTINUE. Press OK to continue, or cancel to stay on the current page." (See the second inset image at right.) Notice the nonsensical self-contradictory text: "Press cancel to continue. ... Press OK to continue." On this basis, a reasonable user can only guess how to easily leave Tatto's site.
• Tatto's actual offer-- a cell phone horoscope service -- is unrelated to Tatto's initial pitch (touting friend requests and crushes).

I preserved these practices in videos showing Tatto's deceptive ad placement and landing pages, and Tatto's efforts to discourage and confuse users attempting to leave Tatto's site.

Tatto's deceptive ads are particularly puzzling in light of recent Washington Attorney General litigation about these same advertising practices. The Washington AG carefully documented the same practices I presented and described above -- including Tatto falsely claiming that a someone has a "crush" on a user seeing an ad, falsely claiming that the person with the crush is attempting to contact the user, creating a false sense of urgency, and impeding exit. In a November 2008 assurance of discontinuance, Tatto agreed to cease these practices. Yet I saw these ads more than a month later. These ads therefore appear to violate Tatto's agreement with the Washington Attorney General.

I see ample basis for Right Media to reject this Tatto ad: The ad makes claims that are not grounded in actual fact. The ad is disguised to look like part of a web page rather than admitting that it is separate advertising content. The ultimate service importantly differs from anything mentioned in the initial ad. Moreover, the landing page also makes false claims, rushes a user to make a decision without appropriate time to consider the consequences, and presents disclosures where they are hard to read and unlikely to be noticed. These practices continue despite a state consumer protection agreement exactly requiring the discontinuance of such practices. However, despite these significant and multiple causes for concern, Right Media nonetheless continues to run the ad.

Webrewardstream: False statement that "you have won"

In testing on January 7, 2009, I requested Cartoonnetwerck.com, a typosquatting site confusingly similar to Cartoonnetwork.com, a popular kids site. Cartoonnetwork's links to Google indicate that Cartoonnetwerck was registered by a Google partner with code client=ca-dp-highlands19_3ph_xml -- probably Highlands International Investment, reportedly one-and-the-same as Domain Doorman, Capitol Domains, Belgium Domains, and others; the subject of typosquatting litigation by Dell and others.

Cartoonnetwerck opened a popunder, shown at right after activation. (The popunder would also become visible after a user closed the covering browser window.) The popunder showed a Webrewardstream ad loaded from Cpxinteractive (of Westbury, NY) and Right Media. (packet log)

The Webrewardstream popunder says "Congratulations!! You have won todays contest in Franklin!!" (s.i.c.). (Franklin is a Massachusetts town often associated with IP addresses in Boston, based on the IP Whois information publicly associated with Boston IP addresses.) However, in fact I had not "won" anything, and as best I can tell, there was no "contest." Rather, this message was simply an ad attempting to lure me to substantially the same Webrewardstream offer shown above.

This Webrewardstream ad's substance and format suggest that it is an ordinary web site message rather than an advertisement. For one, the ad falsely claims that "you have won" -- a statement unlikely to appear in an ordinary advertisement, but conceivable for a web site (which might offer a bona fide contest or promotion). Moreover, the message's bare design, simple appearance, and overall context and placement further suggest that it is a bona fide message, not a commercial solicitation. I therefore conclude that this ad falls short of the FTC's requirement that advertisements must not conceal their commercial nature.

If a user clicks the ad, the user is taken through Cpxinteractive to Right Media and onwards to Webrewardstream.

When I clicked through to the ad's landing page, the landing page initially opened in a window too small for me to see or read any of the page's details or disclosures. instead, the landing page showed only the "enter your email" text box and submit button. As a result, the landing page encourages users to register without ever seeing the disclosures necessary to make an informed decision. Meanwhile, even after maximizing the landing page, disclosures still appeared only in vague language below the fold. These disclosures feature the same deficiencies described in the preceding Webrewardstream section.

I see ample basis for Right Media to reject this Webrewardstream ad: The ad falsely claims 'you have won," when the ad is actually an ordinary commercial offer shown to the general public. The ad's landing page opens in an unreasonably small window, failing to disclose material terms, with its disclosures (such as they are) where few users are likely to notice. The ad and landing page fall far short of FTC requirements in the similar ValueClick case (discussed above).

Consumergain fake user interface popups with improper use of the word "free"

In testing on December 24, I browsed the Cinevivo web site, a parked domain hosted on Information.com. I received the popunder shown at right (after activation) -- a freestanding 720x300 fake user interface window. The ad loads from Cpxinteractive (of Westbury, NY) and Right Media, ultimately selecting an ad for Consumergain (of San Francisco). (packet log)

The Consumergain ad shares numerous design characteristics with genuine Windows message boxes. Notice the ad's color scheme, matching the distinctive Windows XP theme of white, grey, and dark blue. The ad also mimics XP standards in its blue left-edge stripe with a white-on-black icon at top-center. Furthermore, the ad copies standard windows through its wizard-style buttons in the bottom right corner, its bottom top-right heading, its smaller type below, and its use of standard Windows sans serif fonts throughout. For comparison, see the genuine Windows XP "add printer wizard" -- featuring these same design characteristics.

The Consumergain ad falsely claims to report "system status" -- when in fact the message is simply an ad, not a report of the status of a user's computer. Moreover, the ad claims "your urgent attention is required!" although the ad does not actually require any special attention. These false claims reinforce the deception caused by the ad's fake user interface design -- causing a reasonable user to mistakenly conclude that this message comes from Windows, from a computer's "system," or from other software already on a user's computer.

The Consumergain ad falsely claims that a user has "been chosen" when in fact the ad is shown to the general public, and the user has not been chosen or offered anything out of the ordinary.

The Consumergain ad repeatedly claims its offer is "free." Although the word "free" is accompanied by an asterisk, the asterisk and footnote do not cure the ad's improper use of the word free. For one, the footnote appears only in grey-on-grey type -- unusually hard to read. Moreover, the footnote fails to disclose the specific requirements associated with receiving the free gift card -- exactly contrary to the FTC's guide concerning use of the word "free", which requires that all such conditions be "set forth clearly and conspicuously at the outset of the offer."

If a user clicks the ad, the user is taken from Information.com to Cpxinteractive to Right Media and onwards to Consumergain.

Consumergain's ad and landing page fall short of the FTC's requirements. Consumergain's initial advertisement nowhere mentions that a purchase is required in order to receive the promised "free gift card" -- but FTC requirements clearly indicate that the initial advertisement must disclose that purchase is required. Consumergain's landing page does include this disclosure, but Consumergain presents the disclosure in a typeface that is likely to be overlooked. In particular, Consumergain uses light blue text against a medium blue background, dramatically smaller and less noticeable than surrounding text (such as a glowing red all-caps restatement of the supposed gift card), whereas the FTC requires use of the same color, font, and size as the claim of "free." Finally, in explaining its lead generation requirements, Consumergain mentions the number of offers a consumer must complete only in small type at the far bottom of the page. A user must scroll to see this text, and a user must click a separate hyperlink to learn key details (such as which specific offers are available). Hyperlinks and small type at the bottom of the page violate the FTC's requirement of "conspicuous" disclosure of consumers' specific obligations.

I see ample reason for Right Media to reject this Consumergain ad. The ad is disguised to appear to be a Windows message, and its false "system status" and "attention is required" titles reinforce the deception caused by fake user interface design. The ad repeatedly uses the word "free" in violation of FTC requirements for use of the word "free" generally and for lead generation specifically. Furthermore, the ad's landing page falls short of the disclosure substance and format the FTC demanded of other lead generation services.

Additional deceptive ads

I have also observed the following additional deceptive ads (among numerous others). Each ad was distributed through Right Media in December 2008 to January 2009. I see ample reason for Right Media to reject each of these ads, for the reasons listed in the right column.

	This ad is shown broadly -- not just to the "10,000th visitor." The viewer is not actually the "#1 winner."
	The ad references only one specific entity that might plausibly be the ad's sponsor -- the US government. But the banner does not actually promote the US government. Rather, the advertiser offers a paid service not affiliated with the US government.
	Contrary to the ad's title, the user has not actually "won" anything. The banner is shown to the general public.
	Contrary to the ad's text, the user is not actually the #857,336 visitor, and the user has not actually "won" anything. Rather, this banner is shown to the general public.
	The ad seeks to distribute software that is entirely new to a user's PC -- not any kind of "update." The software is not affiliated with Microsoft, despite the ad's use of the distinctive Windows logo. The Vista-style minimize and "x" buttons activate the ad, rather than dismissing it.
	The ad presents a mock (albeit animated) progress bar. Despite the ad's appearance, no task is actually in progress, and the user's computer is not actually "receiving [a] message." The Vista-style minimize and "x" buttons activate the ad, rather than dismissing it.

It is easy to find scores of similar ads. In the subsequent sections, I attempt to quantify the breadth of Right Media's deceptive ads based on information Right Media itself collected.

Right Media's taxonomy of deceptive ads -- and Right Media's assessment of their prevalence

As early as 2006, Right Media recognized the problem of deceptive ads in its marketplace. Publishers repeatedly complained of ads inappropriate for their sites -- ads showing unwelcome sexually-explicit content and ads installing spyware, to be sure, but also ads engaging in the forms of deception shown above. In response, Right Media developed Media Guard -- a classification system whereby Right Media staff examined each ad and tagged it with into any of 160 categories. In October 2006, Right Media even posted its Media Guard categories for public inspection. In that initial posting, Right Media listed seven kinds of physical deception and five kinds of deceptive language:

By September 2008, Right Media's analysis of deception had been reduced to five kinds of physical deception and four kinds of deceptive language shown in the table below. For each kind of deception, Right Media offered a clarifying statement and an example banner ad, as presented in the second and third columns of the table.

Right Media offers publishers the ability to allow or deny each of these deceptive categories of advertising. For example, a publisher can check a box to exclude all ads with, e.g., false error messages. As a publisher considers these custom exclusions, Right Media shows the publisher the effect of each exclusion -- how many ads would remain available if a publisher elected to ban, e.g., ads false error messages. To check the effects of excluding each category of ads, in September 2008 I logged into a Right Media Direct Media Exchange, accessing Right Media's assessment of all ads in its Remix Media network. By default, Right Media proposed to exclude only the 0.43% of Remix Media ads with graphic adult material, other prohibited content, or ActiveX, leaving 99.57% of ads available. I opted to exclude the various deception categories in sequence, one at a time, and I noted the percent of ads that would remain after each such exclusion. (See the next to last column in the table below.) By subtracting the remaining percentages from 99.57%, I computed the proportion of September 2008 ads that, in Right Media's judgment, exhibited each listed characteristics. See values in the final column in the table below. Click the arrow in the first column to see these values as presented on Right Media's site, along with Right Media's descriptions and example ads.

Characteristic	Description	Example	% of Right Media Remix ads
			remaining after excluding this characteristic	with this characteristic
Mimics Windows/Mac/Unix dialog boxes	Creatives that intend to convince the user that they are seeing a genuine dialog box when they are not.		94.62%	4.95%
False expand/minimize/close buttons	Creative resembles or mimics Windows expand/minimize/close buttons to entice users into clicking through.		96.04%	3.53%
False web elements (i.e. pull-down menus, buttons, links)	Creative contains pull down menus that are only images but do not function.		92.85%	6.72%
False error messages	Creatives that intend to convince the user that they are seeing a genuine error message when they are not.		97.60%	1.97%
False downloading messages	Creatives that intend to convince the user that they are seeing a genuine downloading message when they are not.		99.34%	0.23%
"You have been chosen to win"	Any creative that claims that a viewer has been specially selected in some way for a prize or gift.		96.94%	2.63%
"Free" with NO disclosure language	An offer for an item that is free, or that uses the word free or other similar language, implying the reader doesn't need to pay for the item. an example of a free offer might be, "Get a free Ipod". There is no disclosure language or footnote about any qualifications to the free offer that might require participation on the part of the consumer.		81.79%	17.78%
"Free" with disclosure language	An offer for an item that is free, or that uses the word free or other similar language, implying the reader doesn't need to pay for the item. an example of an offer with disclosure might be, "Get a free cell phone with a one year subscription to Mobile Carrier X." The offer language does contain an asterisk, with corresponding footnote, or other legible disclosure language or footnote.		90.33%	9.24%
Deceptive or questionably realistic offer	(I mistakenly failed to preserve Right Media' s description for this characteristic.)		99.57%	0%

In principle, an ad might suffer from several of these characteristics. For example, an ad might present a false error message and also falsely claim a user has been chosen to win a prize. If ads present from multiple deceptive characteristics, it would be inaccurate simply to add the numbers in the final column in order to attempt to determine how many ads feature one or more of the listed characteristics. That said, if is impossible for an ad to satisfy both "free" categories. (A single ad cannot both have disclosure language and lack such disclosures.) Combining those rows, those two characteristics jointly apply to 27% of Right Media's Remix Media ads. Leaving a conservative allowance for overlap among the other characteristics, I conclude that these characteristics collectively apply to at least 35% of Right Media's Remix Media ads.

Right Media's response to publisher concern

Right Media encouraged publishers to accept ads tagged with negative categories. For example, in a 2006 message to Direct Media Exchange publishers, a Right Media staff person commented that ads in the "'free' and similar representations with NO disclosure language" category "can get a lot of clicks and conversions and should probably be allowed" -- completely ignoring the ads' tendency to deceive and their apparent violation of applicable FTC requirements. As to ads that are difficult to close or exit, the Right Media staff person encouraged publishers to ignore the problem since publishers get paid no matter users' experience: "It is something that may affect the user after they click the ad, but won't create undesirable behavior on your site." See these quotes in context (Right Media Direct Media Exchange account required) or in a screenshot.

Right Media was similarly unconcerned when publishers flagged deceptive ads. For example, in November 2007, one publisher was alarmed by "someone has a crush" ads similar to the Tatto offer flagged above. The publisher complained: "this is a downright plain deceptive ad made to ripoff people ... any website owner with a conscience would not put an ad on their site like this." In response, Right Media staff simply told the publisher how to remove the ad from the publisher's site -- but Right Media did nothing to protect the many users at other publishers' sites. See the discussion in context (account required) or in a screenshot.

Right Media's "community" discussion (registration required) preserves numerous other publisher complaints of dubious ads, and various further Right Media staff attempts to downplay the problem.

Right Media's response to initial scrutiny and public concern

In October 2008, I mentioned Right Media's deceptive ads to Bob McMillan of IDG News Service. Bob subsequently covered the problem of deceptive ads running through Right Media.

I had hoped Yahoo would respond to this news coverage and public concern by excluding all ads that fall short of public expectations or, in any event, do not provide what governing consumer protection law requires. After all, just last spring Yahoo then-CEO Jerry Yang said Yahoo sought "to be the best corporate citizen it can be." Dropping deceptive ads seems a natural step in that direction -- indeed, an unusually easy step, thanks to Yahoo's preexisting systems to categorize deceptive ads.

Instead, it seems Yahoo responded to McMillan's coverage by attempting to hide the problem. For one, Media Guard no longer classifies ads into types of "deceptiveness." Instead, Right Media renamed the "deceptiveness" category of characteristics -- now calling those categories "ambiguous or unclear attributes." But Right Media's first classification system was more clear and more straightforward: The listed characteristics are exactly designed to deceive, and Right Media only makes the problem harder to understand by whitewashing its characterization of the ads' effects.

Furthermore, Right Media's Media Guard now only provides two categories of physical deception ("Mimics system messages / alerts / functionality" and "False web elements (i.e. pull-down menus, buttons, links)") and three categories of deceptive language ("'You have been chosen to win'", "'Free' with NO disclosure language", and "'Free' with disclosure language"). See a screenshot of Media Guard as it now stands. Right Media's removal of the other four "deceptiveness" categories stands in sharp contrast to Right Media's unambiguous promise to offer "an unprecedented level of detail and accuracy" in its categorizations. Far from "unprecedented ... detail," Right Media's classifications are now less detailed than they were last fall.

McMillan quotes Yahoo spokeswoman Kristen Wareham defending Right Media's practices: "The Exchange doesn't make a judgment on that type of ad category. It's up to the publisher to select the type of ad that works on their page." Perhaps it's Yahoo's prerogative to design its systems that way. But as publishers, advertisers, policy-makers, and users see the unsavory ads Right Media systematically offers, Right Media's reputation will drop further -- ultimately tainting Yahoo's name too.

Next steps

Despite Right Media's removal of numerous key categories of deceptive ads, deceptive ads remain widespread in Right Media's marketplace. Recall the examples above, all observed during December 2008 to January 2009.

Consumers need not accept ads specifically designed to deceive. Nor does Yahoo need to facilitate the placement of these ads through its advertising systems. We ought to demand better -- particularly from a company that styles itself a leader in ethical practices.

All these examples fall within the broad task of assessing ad quality. But at the same time that Right Media matches publishers with advertisers, Right Media equally matches advertisers with publishers. In so doing, Right Media faces the significant challenge of publisher quality. For years, I've been finding Right Media placements in spyware-delivered pop-ups (example), in ads that spyware programs insert into other publishers' sites (without permission of, or payment to, those publishers) (example), and in banner farms that consist of nothing but ads (example). I found interesting new examples this fall -- including programs and sites loading literally dozens of Right Media ads at once, without showing users even a single ad. I'll present these practices in a separate article later this spring.

Meanwhile, it seems that the public cannot look to Yahoo for effective supervision of the Right Media marketplace. So I will attempt to find other channels to protect unsuspecting consumers. Later this week, I expect to file complaints with the BBB's National Advertising Division about a variety of Right Media advertisers (some of those listed here, among others). I expect a favorable response from the BBB in light of the BBB's criticism of prior deceptive fake user interface ads from Bonzi.com, Casale Media's Spyware Stormer, and Bullseye Media's ConsumerSavingCenter. That said, the NAD's procedures will prohibit me from discussing these complaints during the review process (rule E.(iii).(a)), so I will be unable to comment on NAD outcomes until those matters are complete. In the interim, I will do my best to bring these deceptive ads -- and Right Media's role in distributing them -- to the attention of appropriate consumer protection authorities nationwide.

Right Media's response to this piece

Right Media provided PC Magazine's Larry Seltzer with the following statement:

“Right Media is deeply committed to providing a high-quality experience for advertisers, agencies, networks and publishers. We have rigorous platform standards and guidelines for our members that we expect them to follow, which include preventing the use of our system in a misleading, deceptive or illegal manner.  For example,  since Exchange members classify their own content, we expect all advertisers to accurately categorize their creatives and associated landing pages and we expect that publishers will select the types of ads that are appropriate for their websites. Our Media Guard tool for creatives also provides additional safeguards for the Exchange.   If we learn that an ad is in conflict with laws or regulations, our expectations or our guidelines, we take action to remove it from the Exchange as quickly as possible.” 

To my eye, Right Media's response ignores the key issues. Right Media claims to prohibit using its system in a misleading, deceptive or illegal manner -- but Right Media's own ad categories reveal that Right Media knows deception is widespread. If Right Media wants to put teeth behind its prohibitions on deceptive and illegal uses of its system, it should kick out the many ads that violate these prohibitions.

Furthermore, it's hard to put much faith in the final sentence of Right Media's statement. Consider, e.g., the "'Free' with NO disclosure language" category that, in Right Media's own analysis, comprises 17%+ of Right Media's Remix Media ads. The FTC specifically requires that when receiving a free item is contingent on purchasing additional terms, all such requirements "should be set forth clearly and conspicuously at the outset of the offer." That is, the FTC requires appropriate disclosures on ads offering something for "free" -- whereas 17%+ of Right Media's ads promote "free" products but lack exactly that disclousre. Far from removing deceptive ads "as quickly as possible," Right Media let this single category of deceptive ads grow to 17%+ of its Remix ad inventory, and Right Media staff even encouraged publishers to allow ads featuring this deception, Despite Right Media's claims, Right Media's actions are the very opposite of removing unlawful ads as quickly as possible.

Posted: January 14, 2009. Sign up for notification of major updates and related work.