Microsoft Search 101
Microsoft Search 101
What is Microsoft Search?
Microsoft Search is the evolution of search at Microsoft. Microsoft Search transforms the way people in an organization find the information they need—no matter where they are in their cloud journey. Either integrated with Microsoft 365 or as a standalone solution, Microsoft Search is a secure, easily managed, enterprise search experience that works across Microsoft 365 applications and services to deliver more relevant search results and increase productivity.
Microsoft Search powers the search experience across Microsoft 365 apps and services to include (e.g. Microsoft Search entry points):
- Microsoft Teams
- Microsoft Bing
Relevance and Personalization
How does Microsoft Search determine relevance?
Microsoft Search leverages the global perspective of the Microsoft Graph to deliver personalized search results across information stored in Microsoft 365 and connected systems when using Microsoft Graph connectors.
Relevance is determined based upon a set of signals that exist between an anchor and an object. A signal is an event capturing a user-behavior, either active or passive, that is commonly used to directly power end-user facing experiences.
Microsoft Search uses these signals to deliver high value, personalized results, and ranking.
What’s the relationship between the Microsoft Graph and Office Graph?
The Microsoft Graph and the Office Graph are different things despite similar names. The Microsoft Graph is the API endpoint, that surfaces all Microsoft 365 data (AAD, Exchange, SharePoint, Excel etc.).
The Office Graph is the name for a collective set of services and insights generated on top of the infrastructure developed by FAST, most associated with Office Delve.
In some cases, some REST APIs are referred to as Office Graph APIs when they provide access to insights that are not exposed through other established APIs. For example, the Microsoft Graph Insight APIs expose relationships generated by Office Graph services directly to the caller.
Personalization of results
Whichever app people are working in; Microsoft Search is personal. Microsoft Search uses insights from the Microsoft Graph to show results that are relevant to each person. Each person might see different results, even if they search for the same words. They only see results that they already have access to, Microsoft Search doesn’t change permissions.
Leveraging Microsoft Graph’s insights API, Microsoft Search uses advanced analytics and machine learning to provide the most relevant files people need throughout their workday. This API is sued to power familiar Microsoft 365 experiences, including Office Delve, SharePoint Home, the Discover view in OneDrive for Business, and Outlook on the web.
Context versus Coherence
What is context?
Context is the circumstances that form the setting for an event, statement, or idea, and in terms of which it can be fully understood and assessed. In Microsoft Search, context is the user job, or otherwise where they’re working and the corresponding intent. For example, in Outlook, Microsoft Search prioritizes search results based on the context (user job), interpreting the users’ expectation for search to return the most relevant email conversations, calendar items, people, etc. whereas in SharePoint, results are prioritized according to that context – sites, files, news, and people.
What is coherence?
Coherence is the quality of being consistent or forming a unified whole. In Microsoft Search, the application of it in Microsoft 365 is coherent, meaning that it’s ubiquitous, available across its applications and services; however, intentionally is not coherent in the results provided across those apps and services.
For example, when searching in Outlook (context or otherwise the setting) the expectation would be results are prioritized to that setting, finding emails, etc. that satisfy the intent of the search or in Microsoft Teams, chats, messages, etc. whereas in SharePoint, the expectation would be to find information related to that setting. Several “settings”; however, do provide wide coherence in the scope of tenant-wide search to include Office.com, SharePoint home, Bing, and the NTP in Microsoft Search are all scoped to tenant-wide results by default. In addition, for several core Answers, coherence is provided across search entry points. Learn more about Answers here.
Microsoft Search can be customized in a variety of ways dependent upon the selected entry point.
Verticals and result types can be created and presented in the SERP across SharePoint, Office.com, and Microsoft Bing when configuring external data connections with Microsoft Graph connectors.
Verticals are intended to make it easier for users to find the information that they have permission to see. For example, you could create a search vertical for marketing analysis data from third-party software for people in the marketing department. You can also define result types and customize the layout for this data.
Verticals and result types can be created at both the organizational (tenant-wide search) level in addition to the site level. Organizational search, as previously mentioned, is provided by SharePoint, Office.com and Bing whereas site-level search is provided by SharePoint.
You can define how results are displayed in the vertical by designing the layout using result types. The result layout lets you show important information directly in the search results, so people don’t have to select each result to see if they found what they’re looking for.
A search result type is a rule that causes distinct kinds of search results to be displayed in different ways. It consists of the following:
Custom results pages
Custom results pages allow for redirecting search to a custom SERP.
With a custom results page you can create a new page that can be used to control the layout and design of search results to support your needs. You can use any built-in web parts, open-source search web parts from SharePoint Patterns and Practices community, as well as any custom web parts that you may have developed using SharePoint Framework.
Learn more about custom results pages at https://docs.microsoft.com/en-us/microsoftsearch/create-search-results-pages.
The Microsoft Search API provides a query method to search across your data in Microsoft Search, where you pass a searchRequest in the request body, defining the specifics of your search. You can use the Microsoft Search API to query Microsoft 365 data in your apps.
Learn more about the Microsoft Search API at https://docs.microsoft.com/en-us/graph/api/resources/search-api-overview?view=graph-rest-beta and get samples at https://github.com/microsoft-search.
Microsoft Search is managed through a unified admin center, Search and intelligence, in Microsoft 365.
Microsoft Search is on by default for all Microsoft 365 apps and services that support it (see above). Bing can optionally be enabled for people in an organization to allow for organization wide and web search via bing.com when signed in with a Work Account or when searching with the address bar in supported browsers.
While Microsoft Search doesn’t require setup, using the Search and intelligence admin center, you can improve the overall Microsoft Search experience through some basic administrative tasks such as creating and curating Answers, customization (connectors and result type), and monitoring key metrics to evaluate how search is performing in your organization.
To manage Microsoft Search from Microsoft 365 admin center:
In Microsoft 365 admin center, go to Settings > Search and intelligence.
Microsoft Search uses Microsoft Graph connectors to provide connections to external data sources, making that data searchable across Microsoft 365 entry points in addition to Microsoft Bing.
By default, Microsoft Search indexes all information across Microsoft 365 apps and services. Microsoft Graph connectors allow you to expand the scope of indexing to third-party data, both on-premises and in the cloud, to be available in Microsoft Search.
In addition to connectors provided by Accenture, Adobe, BA Insight, Box, Cognizant, Go1, LumApps and Raytion. Microsoft Search provides a set of out-of-the-box that can be used to index systems to include:
Additional connectors will be available to support Windows File Shares and Oracle Db (on-premises).