Resources

Here you can find a collection of our Newsletters, Videos, and Blog posts. We hope you find them interesting, educational, and they keep you up to date on what’s happening with digital marketing.

GOOGLE REVEALS HOW AUTOCOMPLETE PREDICTIONS WORK

Dec 2, 2020 | 2020, December

Google recently revealed how their Autocomplete predictions function. Google’s Danny Sullivan explains how Autocomplete isn’t based on popularity alone, as most people assume.

“Autocomplete is a complex time-saving feature that’s not simply displaying the most common queries on a given topic.”

In addition to recent Google searches, there are several other factors taken into account:

  • Common and Trending Queries
  • Language and Location
  • Freshness
  • Intuitive Suggestions Based on Similar Types of Queries

Danny adds that there are certain predictions you are less likely to see:

  • Unexpected or shocking
  • Untrue and mistaken for fact
  • Violent, sexually explicit, hateful, disparaging, or dangerous
  • Likely to return unreliable content
  • Offensive, hurtful, or inappropriate queries in relation to named individuals

He states that, “people can still search for such topics using those words … Nothing prevents that.” However, Google tries to prevent generating predictions that people might not be expecting to see.

He concludes by saying:
“While our automated systems typically work very well, they don’t catch everything. This is why we have policies for Autocomplete, which we publish for anyone to read. Our systems aim to prevent policy-violating predictions from appearing. But if any such predictions do get past our systems, and we’re made aware (such as through public reporting options), our enforcement teams work to review and remove them, as appropriate. In these cases, we remove both the specific prediction in question and often use pattern-matching and other methods to catch closely-related variations.”

Google understands that their protective system may keep some useful predictions from showing up, and they admit taking “a particularly cautious approach when it comes to names and might prevent some non-policy violating predictions from appearing.” They also know that even if a prediction isn’t autogenerated, the user can still type in the desired keywords related to their query and view the search results.

If you wish to read Danny’s original post, you can find it here!