When you search in Viva Engage, a lot happens behind the scenes, all in less than a second. Searching in Engage involves a collection of innovative features to focus search results to the desired goals of your query. The following image shows the workflow of Engage search, which uses a powerful feature called hybrid search.

A flowchart showing the behind-the-scenes processing of a search query for Viva Engage

Hybrid search: Keyword matching + meaning matching 

Viva Engage uses an architecture called hybrid search, which is two fundamentally different search approaches that run at the same time and complement each other: 

  • Keyword matching finds posts that contain the exact words you enter. It uses a proven information retrieval technique that considers not just whether your keywords appear, but how often they appear and how distinctive they are. A rare, specific word like "hackathon" carries more weight than a common word like "team." This is great for specific terms like project names, acronyms, or someone's name. If you search for "FY26 Q3 OKRs," keyword matching finds posts that use those exact terms. Keyword matching also draws from two pools of content: exploration results (discovery-focused, across all content you can access) and affinity results (personalized, weighted toward people and communities you interact with most). These two pools are merged to give you both breadth and personalization right from the start.

  • Meaning matching uses AI to understand the intent behind your query. Your query text is converted into a mathematical representation of its meaning (called an "embedding"), and then compared against embeddings for all threads in the network. Posts with similar meaning are surfaced, even if they use completely different words. For example, if you search for "team morale ideas," meaning matching can detect and show a post titled "Fun Activities to Boost Team Engagement". There's no word overlap, but the meaning is the same. Only results that meet a minimum similarity threshold are included, ensuring quality.

Why use both? Keyword matching is precise and predictable. Meaning matching helps you discover content you might have missed. Together, they cast a wide net, typically evaluating hundreds of candidate posts, before narrowing down the most relevant results. 

Privacy and permissions

Engage applies strict permission checks. You will only see content to which you have access. Posts from private communities you haven't joined, or threads that have been deleted, are never shown. Content from communities you've muted will still appear in search results - muting affects your feed, not search. 

Personalized ranking 

After finding all the potentially relevant posts, Engage uses a machine learning model to rank your search results. The model evaluates each candidate post across over 100 different signals, organized into several categories: 

  • Text relevance: how well the post content matches your query, measured through multiple dimensions including term frequency, term importance, match density, and which part of the post the match appears in (the title, body text, or replies).

  • People affinity: how much you interact with the person who wrote the post, across Engage, Outlook, Teams, and other Microsoft 365 tools. The system computes a personalized affinity score between you and every author in the result set.

  • Community affinity: how active you are in the community where the post was shared, based on your visits, replies, and engagement history with that community.

  • Semantic similarity: three separate AI-computed similarity scores: how close your query meaning is to the post content, to the post author, and to the community where it was posted.

  • Recency and time signals: when the post was created, how much time has elapsed, and time-decay factors that naturally boost newer content.

  • Engagement signals: view counts, reply counts, reactions, and the user's own search and click history help predict what they'll find valuable.

  • Highlight quality: where in the post the matches appear, how concentrated they are, and how prominent the matching sections are.

The end result is that two people searching for the same thing might see different results. If you work closely with someone in the engineering team and they posted about "hackathon," their post naturally ranks higher for you than for someone who has never interacted with them. 

Speed and performance 

All of these capabilities, including hybrid candidate generation, feature computation across 100+ signals, machine learning ranking, and permission filtering, take place in under a second. Several techniques make this possible: 

Parallelism: keyword and meaning matching run simultaneously, not one after the other, so the total time is the duration of the slower search, not the sum of both 

Smart caching: when you view the first page of results, Engage pre-fetches and caches the next page in the background. This means pagination feels instant - clicking to page 2 or 3 serves cached results with no delay 

Batch processing: signals like engagement history and community metadata are fetched and computed in optimized batches rather than one at a time 

The result is a search experience that feels instant while doing sophisticated work behind the scenes. 

Searching for people 

Viva Engage search isn't just for conversations. It's also a powerful way to find people across your organization. When you search for a person, Engage matches against: 

  • Display name: first name, last name, or both (e.g., "Rajesh Jha")

  • Job title: search by role (e.g., "engineering manager" or "principal PM")

  • Email or alias: search by email address or alias prefix

People results appear in both the instant suggestions dropdown and in the People tab on the search results page. Each result shows the person's name, profile picture, job title, and email, so you can quickly identify the right person even when there are multiple matches. 

Tip: If you know someone's email alias, searching for it (e.g., "chrzeng") is often the fastest way to find them. Searching by job title (e.g., "product manager Engage") helps you discover people you may not know by name. 

Searching for communities 

Looking for a community to join? Search matches against community names and descriptions. This means you can search by topic (e.g., "accessibility," "onboarding," "women leadership") and find relevant communities even if the exact word isn't in the community name. 

Community results also appear in the instant suggestions as you type, making it easy to navigate directly to a community without visiting the full results page. 

A few examples of how community search works: 

You search for...

You'll find communities like...

"accessibility"

Engage Accessibility, Accessibility Connected Community, Accessibility Leadership Community 

"women leadership"

SME&C Women in Leadership, Women's Leadership Community (WLC), Technical Women Leaders

"Azure DevOps"

Azure DevOps / 1ES, and related Azure engineering communities

"onboarding"

Onboarding, Alchemy Onboarding, New Employee Onboarding

Keyword highlighting 

When you land on the search results page, your search terms are highlighted in post previews. This helps you quickly scan results and understand why each post was returned. 

Highlighting appears in the following locations: 

  • The thread title (if the post has one).

  • The body text preview: Engage shows the most relevant snippet of the post with your keywords highlighted.

  • Replies: if a reply matches your query, it appears with highlights below the original post. For example, searching for "AI tools and copilot" highlights each matching word in the results, making it easy to see how the post relates to your query.

Note: Keyword highlighting may not appear consistently across all result types. For instance, topic names and community names are indexed and searchable but are not currently highlighted in the search results. We’re actively working to improve highlighting consistency across the search experience. 

Learn more

Search on Viva Engage mobile

Tips for searching in Engage

Facebook LinkedIn Email

Need more help?

Want more options?

Explore subscription benefits, browse training courses, learn how to secure your device, and more.

Microsoft 365 subscription benefits

Microsoft 365 training

Microsoft security

Accessibility center