Looks like Google is again taking steps to blur the visual distinction between paid and organic search results. A reader at Google Blogoscoped posted an e-mail recently received from Google about his/her AdSense account:

We’re writing to let you know about a coming change to the appearance of your Google ads. Your ads currently display the default Google color palette, Seaside (formerly known as Mother Earth). In the near future, we plan to update the default palette to Open Air, a new palette containing the same set of colors, but without the blue border. We’ve found that many publishers prefer the cleaner look of this palette and have also seen that a blended color palette performs better for them – attracting user interest while still maintaining the distinction between ads and content with the ‘Ads by Google’ label. …If you don’t choose to update your ad code, your ads will begin displaying the Open Air palette within the next 10 days.

While Google has previously given website owners the ability to customize the format of their AdSense ads, with this announcement, Google is taking steps on their own to blur the separation between paid and unpaid content. As one commenter noted, “The fine line between good and evil is dwindling.”

[image from Philipp Lenssen]

Jim was kind enough to ask me for a contribution on the ethical implications of paid search. My brief contribution, “The Value Implications of the Practice of Paid Search,” warns of the potential value implications of the practice of paid inclusion and paid placement of sites within search engine results, focusing on three interrelated values of moral and ethical import: freedom from bias, privacy and trust. It closes with a call to action for the search engine and paid search community to engage in value-sensitive design to ensure such values are protected in the conception and design of these important and powerful knowledge tools, rather than being merely retrofitted after completion.

John Battelle reports on a recent new feature within Google search results where “related queries” get inserted within your original search results. For example, when you search for “on demand,” you also get search results for “Comcast on demand.” (see image).

Battelle received the following statement from Google’s PR explaining the feature:

Google is testing an automated technique for detecting when an alternate query might help users find what they are looking for more quickly. For these searches, which are both commercial and non-commercial in nature, Google displays one or more alternate queries together with a preview of their top results.

What do they mean by “both commercial and non-commercial” alternate queries? Is this yet another case of the blurring of advertising with search results?

While Battelle maintains that “certainly Google would never do paid inclusion,” that’s certainly what it sounds like. Did Comcast pay to have search results using their company’s name appear whenever someone searches for “on demand”? Comcast certainly could use Google’s AdWords program to have “sponsored ads” appear whenever someone searches for “on demand.” But if these “alternative queries” are indeed paid inclusions, Google should, at the very least, disclose what is happening.

I’ve only come across one other example of when such alternative results were inserted into a page of results: a search for “us” includes alternate results for “us magazine,” which provides links to US Magazine, FHM and Newsweek. What’s odd about both the “on demand” and “us” search examples of this phenomenon is that the inserted results are decidedly commercial – they provide links to commercial interests. Do other examples exist that merely provide links to non-commercial information? Might a search for “volcano” provide alternate results for “hawaiian volcano”?

It seems a simple question: are these inserted queries organic search results – simply the 2nd most common search query related to the original query – or are they motivated by Google’s advertising relationships? I’m not looking for a conspiracy – only a fuller explanation.

More blurring of search engine results and advertising:

A paper presented at the Sixth ACM Conference on E-commerce by Penn State School of Information Sciences and Technology professor Jim Jansen indicates that consumers exhibit a substantial measure of bias against sponsored links. The report details a study in which 56 participants, ages 18 to 29, gauged the results from 330 e-commerce result queries entered into a leading search engine, and the researchers concocted a fake search engine to display one page of the results from the queries. Sponsored links and “organic” results returned automatically by the search engine’s algorithmic functions were flipped on half the pages, and the researchers discovered that participants first referenced the results identified as organic on over 80 percent of the searches. From their release:

“Prior research had noted a bias against sponsored links, but the question remaining was whether sponsored links were as good as organic links,” Jansen said. “What our study shows is that even when the returned results are exactly the same, people still view what they thought of as the organic results as better.”

The quality of the sponsored links isn’t the issue; it’s the placement of the results, he added.

The blurring of search engine results and advertising is a major concern given how search engines have become the “center of gravity” for the distribution of information and the acquisition of knowledge. ClickZ reports on a Consumer Reports WebWatch study which claims that Search engines continue to “inadequately inform consumers about the influence that advertising dollars have on placement and rankings of search results”:

A follow-up to a 2004 study examining the ways search engines identify and explain paid search results, the latest study found little improvement over last year, and concluded that none of the 15 search engines examined had satisfactorily disclosed their practices.

“Follow-up research confirms these insufficient efforts to inform consumers about the financial forces at work every time they hit the search button,” said Jorgen Wouters, a WebWatch consultant and author of the report.

Despite many changes among the sites in the study, WebWatch found that nearly half the sites stayed the same, a third got worse, and only three of the 15 sites improved. Key areas considered were prominence of disclosure headings, along with clarity and accessibility of disclosure statements.

The study found that many of the search engines moved or changed the color of these elements from a bright color like red to gray. Several also removed the direct hyperlink from the “sponsored links” headings to the disclosure page, making it more difficult for users to find that information.

Of the top search engines, AOL Search, Google, and Yahoo! Search Marketing were given good marks for disclosure, but showed minimal change over the year before. Ask Jeeves and Yahoo! Search were downgraded for making headings less visible, and removing hyperlinks to disclosure statements. MSN Search was the only major search engine to show improvement, largely because MSN discontinued its paid inclusion and content promotion programs, giving them less to disclose.

Wouters insisted that WebWatch is not on a mission to abolish paid placement or paid inclusion, but rather wants engines to fully disclose that certain results appear because of a paid relationship and not necessarily because of the site’s relevance to the user’s search query.

[via Andy Beal]

One of the things Google brags about with its Adsense and Adwords programs is the high relevancy of the ads placed alongside its search results. Their logic is that the more relevant the ads are to the search, the less intrusive they are. Similarly, they’ve banned the use of banner ads, and try to make the advertising match the aesthetics of entire search engine results page.

These claimed benefits can also do harm. The difference between search engine results and paid advertising is blurring. The more relevant and similar in appearance the ads are, the harder it will become for a user to distinguish them from their “real” search results. This Pew Report, for example, reveals that only 18% of users searchers say they can tell which results are paid for and which are not. Even worse, 62% were unaware that someone has paid for some of the results they see when they carry out a search.

And the situation might grow more dire. First, this blogger discovered a way to eliminate the borders around Google’s Adsense ads – the main feature distinguishing the ads from the content. After removing the borders from the paid advertising, and making them look identical to the rest of the content, he bragged about a 70% increase in clickthrough rates. He is unmoved at the fact that people were misled into believing the Adsense advertisements were regular non-biased links from his site.

And now, Google will soon officially facilitate this kind of deceptive action. Google is testing new ad formats that allow site owners to have greater control over how many ads are show and what the overall size and format of the ad block is. This is a bigger problem than click-fraud; this relates to the tenuous role search engines play in delivering knowledge. The blurring of search results and ads challenges the common view of search engines as non-biased and egalitarian gatekeepers of information.

Andy Beal at Search Engine Lowdown notes this report of the Pew Internet and American Life Project’s study which suggests that only 18% of search engine users can tell the difference between paid and natural search engine results. Beal seems to applaud this:

So what does that tell us? Well, to me it suggests that the search engines have everything finely tuned and doing a great job. If the average user cannot tell the difference, it likely means that they are finding the most relevant results in both paid and natural listings.

Of course, this could also mean that our search results are being polluted by paid advertising placements without the user’s awareness, reducing the validity and reliability of those results. Contrary to Beal’s enthusiasm, users aren’t finding the most relevant results if a link was artificially ranked; they’re finding the results with the best ability to pay for placement. The fact that users are not aware of the placement of advertising within the tools they use to find information is an indication that our search tools need to become more transparant as to any biases that exists within their results.