Google’s Giant New Search Results Banner Ads

Google recently began testing branded search results pages. When users search the name for one of the participating advertisers, there is a small percentage chance that instead of the normal results page, they will end on a page with all of the normal results … and also a very large branded banner ad directly above the results.

That’s no moon

Marketing Land has kindly collected screenshots from users who have had the luck (good or bad) to be presented with some of the test results:

Google’s traditional search results are divided into two kinds: There are the “organic” results of its search technology, and the “paid” results that often appear above or to the right of the results, and are indicated primarily by an impossibly faint background color. These test ads place a “Sponsored” link above the banner ad and draw a faint gray box around both the banner ad, and, confusingly, the associated organic links for the advertiser that naturally appear at the top of the results. And the advertiser will always legitimately appear at the top of the results.

These ads are not being tested for use with random keywords. A search for “shoes” won’t show one; but a search for “Nike” might. Since Google’s search results are generally quite good, any search for a major company will always return that company’s own site as the top results. For larger sites, Google also often conveniently returns links to major sections from the site in a two-column section immediately below the primarily result. Without paying penny, a search for “Nike” will return a top result of Nike’s own site. So why buy banner ads?

Fear will keep the local systems in line

Let us start with a thought experiment: You want to search for flights for an upcoming trip. You could just enter in a search term of “flights,” whereupon you would greeted by some very expensive keyword ads for, in my recent test, CheapTickets.com and Orbitz, the latter of which is the sixth organic result on the page, and the former of which doesn’t even appear. This sort of advertising may very well be quite worth it, especially for CheapTickets.com, with its low organic rank. But what if you know you want to fly on Southwest Airlines? Just what is the address? southwest.com? southwestair.com? southwestairlines.com? You start typing “southwest” into your browser’s unified search and address bar and Google conveniently suggests a search for “southwest airlines” after the first several letters, so you click on that. The top organic result is for the airline’s own site (along with links to online checking and flight booking, among others). You immediately click either on the main page or, perhaps, on the booking link. Here is the key distinction between this and the search for “flights”: you already know exactly where you want to go. The brief stop through Google is merely a two- or three-second interlude because you don’t know the exact address already.

If there were a giant Southwest Airlines banner ad, or paid text ads for Southwest, above the top-ranked organic Southwest link, you might click on one of those instead. Google can claim to advertisers that buying ads for their own name is important because, just look how many people click on them! But what if they weren’t there? You already know where you want to go; you’re just clicking on the first link that takes you there. Advertisers could not spend money on the ads and you’d still end up in the exact same place seconds later, and for free.

As a thought experiment alone it should be obvious that this is indeed exactly how such branded searches work—that there is no benefit to advertisers who buy ads for their own top-ranked names, merely duplicating search results. And yet we don’t have to merely rely on the obviousness of this, because eBay actually conducted a scientific experiment to test the effectiveness of search advertising.

The study, “Consumer Heterogeneity and Paid Search Effectiveness: A Large Scale Field Experiment,” (PDF) makes for a surprisingly entertaining read. They conducted a series of tests that included stopping use of brand keyword advertising:

The results show that almost all of the forgone click traffic and attributed sales was immediately captured by natural search. That is, substitution between paid and unpaid traffic was nearly complete. Removal of these advertisements simply raised the prominence of the eBay natural search result.

This is, of course, exactly what you’d expect to happen. Advertisers are essentially paying to put their own link above their own link. The only winner here is Google, which is conning advertisers into pay for clicks they’re essentially guaranteed to get anyway. The advertisers still get a fully branded experience after the searcher (almost inevitably) clicks on either link, paid or free. So why do advertisers continue to buy ads for their own brands? Fear. Companies bid on their own name, and around one-third of the people click on that link instead of the free one below it. If they stopped bidding on their name, wouldn’t that cut their traffic by one-third? Most companies are simply too scared to try it. In eBay’s study, the actual change was half a percent, which is within the margin of error for the experiment. My own theory is that most big companies simply already have a set marketing budget and they’re content to spend a certain portion of it buying ads for their own name even if they can’t actually determine any return on investment. It’s just a small amount, so even if it doesn’t help, it doesn’t hurt that much, right? And yet these relatively small ad buys from countless companies on demonstrably ineffective ads collectively constitute the majority of Google’s profits. I’ll leave you with my favorite quote from the study:

“Bluntly, search advertising only works if the consumer has no idea that the firm has the desired product.”