Expect to see the issue of how well, or poorly, search engine disclose paid content to come up this month. Reason? Consumer WebWatch will be issuing an "anthropological" report later this month showing confusion about disclosure of paid listings based on a study of 17 web surfers.
You can read a fascinating account of the findings in this transcript from a search engine panel held during April's National Summit on Web Credibility (a good catch spotted by the SiteLines web searching blog and later reviewed by Pandia). I'll revisit the report when it is formally issued. In the meantime, here are some quick comments based on what's already been revealed.
None of the 17 people studied were familiar with paid placement before the survey began. In fact, it was a requirement that they know nothing about it. Instead, they were invited to learn about paid placement after doing some search activities during an "enlightenment" phrase, in which they were apparently pointed at some likely pages at each search engine that provided disclosure information.
Some participants found this disclosure information unclear, buried or written more for advertisers than to inform searchers. I'm not surprised by this, yet it's also important to note that the US Federal Trade Commission guidelines for search engines don't expect consumers to have to follow disclosure links at all to determine what's a paid placement link. Instead, the FTC guidelines recommend that the links be labeled clearly in some way right on the search results page itself.
Given this, I wonder how the testing would have gone if people were asked if they could tell what were paid placement links on a results page just by looking at the page, rather than hunting for disclosure information. It may be that their explorations gave them more information than they needed, creating confusion.
Of course, it is reasonable to expect some people to explore search engine help pages to understand how a search engine operates, which includes how it makes use of paid content. And though the sample is extremely limited, anyone who has ever read some of the search engine help pages will find themselves readily nodding with the bad reactions some in the survey had.
Not all help files are bad, however. In my recent search engine paid disclosure survey, I found it easy to locate good explanations from both MSN Search and AOL Search. Finding Yahoo's disclosure page was much harder, but when I eventually got to the What are the different sections on the Search results page? information, I found a visual guide with good disclosure references. At least, good references from my point of view.
Saying "Sponsored" Isn't Clear Enough?
What about the labels of paid placement right on the search results page? It was surprising to discover that some in the survey found the word "sponsored," which is now nearly universally used, to be unclear. Moreover, in the panel Q&A, it was astounding to hear someone suggest that search engines are using the term "sponsored" as a "veil" to disguise that some links are sold.
The FTC itself suggested that "sponsored" was the type of disclosure language it liked rather than more euphemistic terms such as "featured." The fact that so many search engines now use "sponsored" in deserves praise for doing what the guidelines were recommending.
Indeed, it should be noted that the search engines have "cleaned up their acts" big time since the FTC recommendations first came out. I just resurveyed the situation, which you can find summarized on the Buying Your Way In: Search Engine Advertising Chart at Search Engine Watch -- now with new color coding!
Disclosure Has Improved
Last year, just after the recommendations were released, 25 percent of the search engines surveyed failed to meet paid placement requirements (3 out of 12). The figure's now dropped to only 9 percent -- or just 1 out of 11 in the latest survey. As for that one, Netscape Search, it failed because some new "sidebar" ads it is testing, which are well removed from editorial results, aren't properly labeled. It's an easy fix to make.
The change is more dramatic when it comes to paid inclusion. Last year, NO search engine with a paid inclusion program provided proper disclosure. This year, half of them do. If the survey had been done two months ago, the score would have been 75 percent compliance -- but good disclosure that both Yahoo and AllTheWeb had done was lost during recent design changes.
So things have gotten better. But is better good enough? If people aren't understanding "Sponsored" to mean "Ads," as the new research suggests, then perhaps the language will need to change again.
Paid Inclusion Not Surveyed
The report is also interesting in that it didn't try to understand the impact of paid inclusion. It was considered too confusing to survey and was also seen as not tainting results, in that it doesn't influence rankings.
Certainly paid inclusion is a difficult concept to explain. However, this type of content probably deserves far more attention than paid placement listings, which are generally separated from editorial result and well disclosed.
In contrast, paid inclusion interacts directly with editorial results, and we have only the search engines to take at face value when they say it doesn't impact rankings. After seeing ranking boosts explained away for technical reasons, I've become more convinced that paid inclusion content may ultimately have to be labeled or segregated just like paid placement content. The paid inclusion section of my Buying Your Way In chart explains this more and references some key articles on the topic.
Can't We All Just Get Along?
If the latest research prompts news stories that slam paid listings, expect the search engine industry to fight back with research of its own. The industry is already discussing this, looking to freshen up some past research from last year showing that users are aware of paid listings and that the majority feel the quality of these results is as good as ordinary search results. That research also indicated consumers are most concerned about spawning ads (49 percent) and pop-up ads (35) compared to paid placement, a scant 2 percent.
Overall, I hope it doesn't come down to a battle between sides but rather cooperation. Paid listings are not evil and in fact arguably take pressure of unpaid editorial results from the heavy third-party optimization they faced in the past. I continue to hear webmasters more and more talk about pursuing "content" as a way to win free traffic (which I've always preached) while paid listings are a way to get guaranteed placement.
Both sides, consumer advocates and search engines, should come together to better understand what's the best way to provide disclosure for those who want it, which is in everyone's interest. The last thing the search engines want people doing is completely skipping paid listings out of fear nor scrolling past good editorial results, simply because they think these are all sold. Both examples are cited in the report.
In fact, I saw this first hand when moderated a Future Of Web Search session last month for the Silicon Valley Web Guild. During Q&A, one of the attendees informed the search engines on the panel that since he knows all the top links are sold, he routinely jumps to the second or third page of results.
In reality, on every major search engine, it's still the case that the majority of listings in the top results are NOT sold. Nevertheless, this person clearly had that perception and ironically might have been jumping right past the best results out of fear.
It was amazing to hear, but I had to cut short my interrogation of him when an audience member in the front row looked really unhappy that I failed to keep all questions to something like 30 seconds. Well, it doesn't always work that way :)
It's also interesting to see the Overture rep on the panel struggling with the concept of editorial results versus paid listings. I understand his point. Overture takes pride in the fact that their paid listings are subject to editorial review. Indeed, as I've written before, paid listings actually come under far more editorial review than ordinary crawler-based results.
Despite this, there is a difference. Editorial results -- however they are devised, through the use of crawlers, human power or whatever -- are results that are on the page because it is felt they should be there regardless of payment, in order to ensure search quality. With paid results, no such assurances are made. If someone stops paying, they disappear.
A search for "white house" is a good example of this. Any US-based search engine worth its salt should provide a link to the US White House web site in response to that search, regardless of whether anyone is paying for it. That's editorial in action -- and good editorial is what helps bring in the consumers who will also consider the paid listings, which may also be relevant to their query.
This is also why I strongly resist the term the financial community is so in love with, "algorithmic" results. When Overture purchased AltaVista and AllTheWeb, it is not because it needed to have an "algorithmic" or crawler-based search solution. Instead, it needed to have a set of unpaid editorial results that it could offer to potential portal partners who want those to compliment paid listings.
At the moment, crawler-based/algorithmic technology is the primary way such editorial results are delivered. But that could change in the future. What won't change is the fact that search engines will have both paid and editorial results.