Traditional search engines, such as Google, control what search results are seen and how they are ranked. Although there are hundreds—or even thousands, depending on how you count—of different factors that search engines use to determine the most relevant results, different results may be relevant to different users.
As the dominant search engine on the market, Google has always followed the paradigm that to provide users with the most relevant results, one has to collect as much data about each individual user and then tailor the results to their derived preferences. And this approach makes sense; that data can, of course, shed light on what a user really wants when they just type in one or a few keywords.
However, some new search engines take a different approach: They allow users to take control over their results. Instead of personalizing the results, they allow users to state preferences for certain types of results explicitly. This approach should not be confused with taking control of the results returned by formulating complex queries or using advanced search interfaces. While information professionals can, of course, use these to get the maximum out of search engines, the bitter truth is that lay users use these features very, very rarely. And while these techniques allow us to control the result sets, we cannot control the result ranking.
New search engines
This is where new search engines like Brave Search, Neeva and You.com come in. Although their approaches differ, they allow us to influence their ranking by stating our preferences. What preferences can be expressed and how they are selected varies from one search engine to the other:
You.com allows you to rate apps (meaning sources) that should be shown less or more often in the search results. For example, a simple thumbs-up will show more results from a source. You.com also allows you to add apps and vote on the frequency with which results from that app should appear, both directly on the search engine result pages.
Neeva follows a similar approach. In this search engine, you can state your preferences in areas such as shopping, media, ratings or recipes, also by selecting sources from a list and giving them a thumbs up or down.
With Brave Search, users can use a feature called “Goggles” (not to be confused with that big search engine!). Goggles are collections of sources that, if chosen by a user, will be preferred in their searches. This means that if a user, for instance, added the “news from the left” in future searches, results from these sources will be ranked higher. A nice thing about Goggles is that anyone can compile their own and access the ones provided by other users.
New approach to search
While everyone should decide on their own whether selecting sources from which results should be shown more frequently is to their taste, the overall approach is interesting because it differs from what users have been used to in search engines for many years. While one may question whether it is a good idea that users should choose news sources to be displayed based on their political preferences – and thereby ignoring other sources from the political spectrum, re-ranking results is what really differentiates these search engines from Google and similar search engines. A major learning users can take from these new search engines is that different result rankings are possible. This may sound like a no-brainer, but consider what users have seen over many years: just one set of results, ranked in a standard way. Even if results are personalized, users cannot see whether and how this is done. And often, they are simply not interested in how the results are generated as long as they seem relevant.
In contrast, these new search engines show that different interpretations of the web’s content are possible. And that it’s not only the job of the search engine to interpret this content but also the users’ choice what type of results they prefer. We can only hope that many users will try out regaining control over the results, whether with one of the search engines mentioned or others. Maybe even Google or its main competitor Bing will see the value in letting users decide on sources according to their preferences. In that sense, we may be at a turning point in the history of search engines – and all without fancy artificial intelligence. Giving users control over search results may benefit them more than slight improvements in the overall relevance of results.
Dirk Lewandowski (dirk.lewandowski@uni-due.de; www.searchstudies.org/dirk) is an interim professor for computer science at the University of Duisburg-Essen and a professor of information science at the Hamburg University of Applied Sciences, Germany. He has published extensively about search engines. His latest book, Understanding Search Engines, will be published by Springer in March, 2023.