So, what criteria does Google use to pick the results that appear in the auto-suggest dropdown when you type a query . . . and how can a political campaign help us understand it better?
While most SEO insiders I’ve talked to agree that auto-suggest results have much to do with current search term popularity, others think that additional factors carry more weight in the algorithm. When a client’s name shows up low in auto-suggest, for example, they suspect that optimizing that company’s website and beefing up the link authority might bump its name up a few spots on the list.
Personally, I’m in the camp that thinks auto-suggest results are almost exclusively influenced by current search popularity. But a little test I did recently left me surprised at just how quickly Google updates its auto-suggest results to reflect current search trends.
On Tuesday night, much of the nation (including yours truly) was following a special election in Massachusetts to fill the senate seat left vacant when Ted Kennedy passed away. I suspected that a large number of people following the political campaign across the country were querying Google to find out what times the polls closed in that state, and decided to see if the auto-suggest results would reflect that trend.
At 7:30 PM EST, just a half hour before the polls closed in the Bay State, I started typing the query “what time do the polls close in Massachusetts” in the Google search box. If figured the auto-suggest would present my intended query eventually, but was surprised to see it pop up in the second spot after only typing the six letters “what ti.” To see what I mean, check out the screenshot that I took below.
The small number of characters that I typed to produce the correct auto-suggest, and the fact that it appeared so high on the list, suggests to me that Google was indeed handing back a list of results based on what people were searching for at that time. (Apparently more searchers were interested in what time it was; do that many people really use Google to check the clock?)
The second part of my experiment really convinced me that close-to-real-time search term trends play a big role in auto-suggest results. This part was simple: a few hours later, when it became clear that candidate Scott Brown had defeated Martha Coakley and won the campaign, I went back to Google and started typing the same query again.
The result? Suddenly, “what time do the polls close in Massachusetts” was nowhere to be found in the auto-suggest list, even when I had typed out most of my query. As soon as people across the country stopped asking Google that question, the algorithm yanked it from the list.
Of course, there are a few other variables that I could be overlooking: a higher click-through rate for specific auto-suggest results could end up bumping it to a higher spot on the list, I suppose. But the results of my Massachusetts election test make me even more certain that search engine optimization does absolutely nothing to affect auto-suggest results.
So, what do you think? Leave a comment below and let us know any observations you’ve made about the Google auto-suggest feature that might shed further light on the variables they use.