Bill Baer /bɛːr/

As we continue to ease into working remotely, search has evolved beyond its traditional perception of providing simple utilitarian value, to now serving as a proxy to person-person interactions, quick hallway conversations, and unstructured meetings – effectively search has become the “digital water cooler” for many businesses.  The ability to find information in context has become increasingly important, not only because we’re more distributed, but so is information; however, you need to be able to not only discover information in context but have the assurance that the results are the most relevant to you and your task.

Now in Microsoft Teams when searching for messages, the top 3 messages are generated and ordered by a newly developed relevance model – previously messages were ordered chronologically with the most recent messages at the top of the search results. 

Similar to how broad relevance is determined in Microsoft Search, Microsoft Teams learns the relevant characteristics of a message that matters most to you based on a vast array of active and passive signals, for example, who you interact with most frequently or on which teams and channels you are most active on. In addition, the model can also differentiate between human content and automatic generated content as well as learned preferences such as the age of the message, its attachments, its message content, or its subject.

These improvements to search are available today in Microsoft Teams worldwide.