Google wants you to have the best search experience possible. This means that they want and expect search results to be different from person to person and that even people searching in the same office may see different search results.
This also means that the same person may see different results between his or her work computer and home computer. Here’s why two people in the same room may (and probably will) see different results when running the same keyword search using Google.
Google attempts to automatically detect a user’s location and provide customised results based on that location. They use your ip and location data gleaned from your Google toolbar to estimate this. Users can also specify their preferred location, using a street address. Manually-set locations are saved in a user’s browser cookie.
Your location based results will trigger normally when location specific keywords are typed in. Try searching for “italian food” using Google and you’ll likely get a map of nearby Italian restaurants and takeaways on the right side column as well as some local results appearing at the top of your list.
Google provides personalised results based on previous searches. When you’re signed in to any of Google services, Google personalises your results based on your account’s web history. For example, if you often click on Amazon’s links in your search results, Google will see that as a preference and start to show you Amazon links more often and in higher placements in your search results.
In addition to results based on your web history, Google now shows “Personal Results” when you’re logged into your Google+ account. These are linked from the top of a standard search results page and are based on your personal search history as well as searches from your Google+ connections. So if your friend likes a particular Italian restaurant, it’s likely to come up higher in your search if you’re logged into your Google account while searching.
Even when you’re not signed in, Google still provides personalised search results based on information stored in your browser cookies. Google stores your non-logged in search queries and clicked results for up to 180 days via cookies.
Based on a user’s location, a keyword search is performed by one of Google’s many data centres. Google doesn’t say exactly how many data centers it has, but there are at least 30 reported around the world.
While sharing the same search algorithm, Google’s data centres may have some variance between them during updates. One data centre may already have the newest search algorithm while other data centres still have the older one running while propagation takes place.
And no, you won’t always have your search performed by the same data centre, even if you search from the same location every time. Google routes searches between data centers with their load balancers. While the out-of-sync algorithms are not as common of a factor, it does occur and can cause results to be different for hours or potentially days between data centres as the algorithm update occurs.
Google is constantly testing algorithm changes on search users. Potentially up to 40% of all searches on Google are testing some algorithm variation. The way users respond to the search and subsequent browser history will factor into how such algorithm tweaks are used, understood and adopted.
If you’re the subject of one of these random “bucket tests,” you will see different results than your coworker – even if you both are logged out of Google, have just cleared your cookies, and are searching from the same IP address.
While you don’t have control over which of Google’s data centres your searches are run through and if your search results are part of their algorithm testing, you do have control over your location and personalization – you can change your localisation settings and even turn off search history personalisation, unless you’re not bothered that is.