GoTo Sells Positions

GoTo Sells Positions

From The Search Engine Report
March 3, 1998

Search engine GoTo debuted its new pay-for-placement service on Feb. 21. It allows web site owners to bid for placement. Those willing to pay more can appear higher in the search results.

Pay-for-placement is not new. Open Text experimented with it in mid-1996, and numerous small search directories currently offer the ability to pay for a better ranking.

Despite this, the major search engines have avoided such a service. This is mainly because they see their results as akin to editorial copy. They aren't something that advertisers are supposed to influence. By allowing pay-for-placement, search engine users may begin to mistrust the results they receive. Such complaints helped cause Open Text to give up its experiment.

Does that mean GoTo's results are bad? Not necessarily. Founder Bill Gross makes an interesting argument on how money produces better results.

"This is an opening shot of changing the search engines from a white pages to a yellow pages," Gross says.

What he means is that with a white pages phone book, numbers are organized alphabetically. Money does not get you higher in the results, though some people use company names such as "A Typographical Service" to try and jump ahead of others.

With a yellow pages phone book, money is central. The more you pay, the bigger ad you can have. That doesn't make the yellow pages any less useful. It's a very convenient way to find businesses, and often those with the bigger ads are the better ones.

Of course, there are plenty of good businesses without big budgets, and there are plenty of bad businesses that can afford large ads. So the ability to spend money does not guarantee quality. But it can be a useful ranking tool, and one that Gross says makes more sense than leaving it to search algorithms.

"All you're really getting back is how good people are at tricking the search engines, not how good people are, Gross said."

It's overstated to say only people who know how to trick the search engines get top listings. However, it's not far off the mark when it comes to some popular phrases or one word search terms. The more general and more popular the search, the more likely you'll see people who are aggressively fighting for a top placement.

A search for "monica lewinsky" illustrates this rather easily. A recent survey I did revealed that on average, one-third of the top sites listed in the major search engines were "clickthrough" spam sites.

These are sites that earn referral fees from people clicking through from their sites to other sites. The site owners make pages that the search engine algorithms think are relevant but which most people would probably consider junk.

Beyond this opportunistic activity, there's continuing competition to do well for all sorts of product related terms. While it's heartening to still see small sites rise to the top "naturally," the reality is that people are spending more and more time clawing their way to the top. The days of being listed for popular terms without expending effort aren't gone, but they are receding.

Since time is money, it's easy to argue people are already buying their listings. For those without time or skill, they can turn to search engine placement services. These services will oversee the submission of bridge pages and monitor them for you constantly. They get paid by the click, so if you've got the budget, they can seem a saving grace to the otherwise confusing world of improving your placement. That's assuming you get the listing you are looking for, of course.

Given all this activity, it might be advantageous to search engine users if the search engines took back control of the process. By selling placement for the most popular terms, they might help defuse the arms race of spamming and aggressive placement that exists. Results would certainly benefit, if not just for the fact that opportunistic spam would lose some of its incentive.

It would also come as a great relief to web marketers, who scratch their heads trying to keep up with what's the best way to get listed. Rules about hidden text, repetition in meta tags and the need for HTML text in a graphic-intensive site are confusing and complex. In contrast, pay-for-placement is simple and makes sense.

What about those with little or no budget? Would material at universities be locked out, for example? Again, this material is probably already poorly ranked for many popular and general searches. Pay-for-placement wouldn't hurt it, and the material would still remain accessible for longer and more specific searches.

For example, pay-for-placement might mean a realtor in Newport Beach, California would no longer be listed "accidentally" for "real estate." However, they may still do very well for "newport beach homes for sale."

So there are many reasons why pay-for-placement makes sense. There's also a big reason against it. It just doesn’t feel right.

If you go to a search engine and enter something like "microsoft," many people would expect the service should probably list the Microsoft web site. That's a relevant result, and you want it regardless of whether Microsoft has made an ad buy to ensure it happens.

Likewise, if you do a search for "real estate," you might appreciate having a human editor point you toward a few sites about real estate, not just toward major realtors. You might also appreciate some mechanism that helps you be more specific about what you are looking for.

In that light, what's probably needed is more human intervention in results, not commercialization of them. Search engines might create pages of links for these topics, carefully chosen for relevance. There would probably be a place for advertising on these pages, yet done right, it shouldn't offend anyone concerned over the results being "influenced."

This move could please everyone. Those with money could pay to be placed. Those with quality sites could request being listed on the merit of their content. Users would be given more guidance than they get sometime from raw search results. Spamming should decrease naturally, as incentive disappears.

Raw search results would remain, of course. You can't create pages for everything, and even then, people may still want more choices. However, these popular topic pages would likely satisfy many users.

You can see these types of pages growing already. Excite's Entertainment channel or the Lycos Web Design guide, both which can be found from the home page of either service, provide more organized and relevant information that a blind search for "entertainment" might yield. At some point, it is likely they will make the leap to supplant raw results, in some cases.

Ironically, GoTo had these precise type pages. A search for "travel," for example, brought up "The Travel Page." It was nicely organized, with sites listed in side-by-side columns under topics such as "car rental" and "air travel." At the bottom were raw search results. Now, only raw results are presented.

Enough about the pros and cons of paid listings. How about GoTo itself? For one, GoTo is not a new search engine. It was launched in 1997 and incorporates the former University of Colorado-based World Wide Web Worm. However, its results have been algorithm ranked until now.

To make a purchase, advertisers open an account and note how much they wish to "bid" per click. For example, Toyota currently comes up tops on a search for "cars," and the amount of $0.02 is displayed next to the listing. That means Toyota has agreed to pay 2 cents for anyone clicking on its link.

Honda could open an account and choose to pay 3 cents per click. If so, it would rise above Toyota. In the case of a tie, the site which users have voted best gets the priority.

The way users get to vote is probably the most annoying thing about GoTo. Selecting a listing launches a frame, with the actual destination page embedded in the frame. The frame allows users to vote on the site, similar to how Lycos sets up a frame when people browse links in its community guide. It can also be easily turned off. But it seems too intrusive, and certainly anyone who actually paid to be listed might prefer to have the screen all to themselves.

What impact will GoTo have on the other search engines? Probably little. A quick call to representatives at Excite and Lycos found minimal interest.

"It will be interesting to see how this plays out. My feeling that the consumer wants something more cleaner than commercialism," said Brett Bullington, Executive Vice President of Strategic and Business Development at Excite

Likewise, Lycos search manager Rajive Mathur said, "I'm not sure it's really providing value to the user, in the long term. I think they want some independent sorting"

The closest Lycos is likely to come to pay-for-placement for the time being are the Bullseye ads it has been experimenting with for the past month or so. They put advertisers just above the raw search results section, though they remain physically separated from it. They should be available to all advertisers within the next week or so, at $75/CPM.

How about the impact GoTo will have on marketers. Will they suddenly flock to the service? Perhaps if it grows more, but it has a way to go.

GoTo has well below the traffic of the major search engines. It's at 5 million page views per month, about one-fifth of what Excite generates. But Gross is a well-known Internet entrepreneur. His company, idealab, is backing GoTo and about to launch a new advertising campaign. Awareness may grow, which in turn may help make pay-for-placement more acceptable than it has been in the past.

Gross stresses that he can experiment with the model more easily. "Since we have no brand to worry about, we can get away with it easier than most," he said. In fact, Gross hopes that his service fills a niche that the other services won't be able to immediately, either because of consumer opposition they would face, or due to lack of interest.


GoTo Advertising Placement Page

Engines Battle Irrelevance of Results
Internet World, Feb. 23, 1998

GoTo Searches With Capitalistic Engine
Wired, Feb. 24, 1998

Pay-for-placement gets another shot, Feb. 19, 1998,4,19307,00.html

New Search Engine Goes Commercial, Feb. 18, 1998,4,19281,00.html

Lycos Adds New Features,
Reorganizes Suggested Links

The Search Engine Report, Jan. 9, 1998

Talks about the "Bullseye" links advertisers can purchase to appear above the raw search engine results, along with other changes.