Collab365

Microsoft 365 Search Guide 2026

Q: How do I integrate external databases without duplicating data?

Administrators must deploy Federated Graph Connectors. These connectors use the Model Context Protocol (MCP) to fetch data live from external systems based on the user's credentials, without indexing or storing the data in Microsoft 365.

Q: How do I fix users complaining about 'Zero Results' for common terms?

Review the Query Analytics report in the Microsoft 365 admin center. Identify top failing queries, then create targeted Bookmarks via the Search & intelligence portal to route those queries to correct resources or embedded Power Apps.

C

Collab365 TeamAuthorPublished Apr 8, 2026

37

At a Glance

Target Audience: SharePoint and Microsoft 365 Administrators
Problem Solved: Irrelevant search results, zero-result queries, slow performance, missing external data, and legacy crawl management in modern M365 tenants.
Use Case: Configuring enterprise search in hybrid environments with Copilot, integrating external data via Graph Connectors, and deploying custom PnP search dashboards.

Microsoft 365 uses continuous indexing since 2021, meaning no manual full crawls are needed.1 In 2026 Microsoft 365, site admins handle Search Centre-like customisations via site settings, while service admins use the Microsoft 365 admin center > Settings > Search & intelligence > Connections for crawling, queries, and reports. Semantic search rolled out globally in 2023, providing an 85% relevance boost per MS benchmarks.2

The Collab365 team have updated their research to replace our old SharePoint 2010 guidance. If your users are reporting irrelevant results, slow performance, or missing enterprise data, you are in the right place. Today, the administrative focus spans Microsoft Search, SharePoint Premium (formerly Syntex) classifiers, and Copilot for Search.3 We have shifted from tweaking keyword frequency to managing Generative Engine Optimisation (GEO) and semantic intent.

Grab a coffee. Here is exactly how we set this up last week, and how you can optimise your tenant for the modern era.

TL;DR / Quick Answer

Site-Level vs Service-Level: Site administrators handle interface customisations (like Search Verticals) directly in SharePoint site settings. Service administrators manage enterprise-wide indexing, Graph connectors, and Copilot settings in the Microsoft 365 admin center.5
No More Full Crawls: Microsoft 365 operates on continuous indexing. Legacy scheduled crawls are obsolete for native cloud data.1
Semantic Reranking: The system automatically applies an L2 semantic ranker to initial keyword results, boosting relevance by 85% without manual intervention.2
Graph Connectors: You can now choose between Synced connectors (indexed into the Graph) and Federated connectors (live fetch without indexing) to bring external data into Copilot.7
Modern Best Bets: Legacy Keywords are replaced by Bookmarks and Promoted Results. These are now manageable via the Graph API and support embedded Power Apps.8
Security Shifts: Basic ACLs are replaced by Microsoft Purview sensitivity labels and Just-In-Time (JIT) access policies that dictate what Copilot can display.4

Key Takeaway: Microsoft 365 handles crawling continuously in real-time. Stop worrying about crawl schedules. Shift your focus to optimising relevance through semantic indexing and extending your search reach via Microsoft Graph Connectors.

Who Is This Guide For?

We wrote this specifically for SharePoint or Microsoft 365 administrators with one to three years of experience. You are likely fielding support tickets from end-users who cannot find the files they need, or who complain that Copilot is giving them outdated answers. You need to know exactly where to apply fixes.

To execute the configurations detailed in this post, you need specific administrative prerequisites. You must hold the Global Administrator or Search Administrator role within your tenant.9 You also need access to the primary administrative portals, specifically admin.microsoft.com.9

Furthermore, many of the 2026 updates we discuss—such as Federated Connectors or specific Copilot Studio custom actions—require your administrative account to be enrolled in the Targeted release track. You can configure this within your organisation profile settings in the admin center.10

Key Takeaway: Always verify your Role-Based Access Control (RBAC) permissions before planning a service-level overhaul. Site-level changes only require Site Collection Administrator permissions, but enterprise-wide connector deployments require Global or Search Admin rights.

How Do You Configure Search at the Site Collection Level in Microsoft 365?

We used to spend hours configuring Search Settings for the Search Centre URL, drop-down modes, and contextual scopes back in the SharePoint 2010 era. The interface was clunky, and the results were rigidly locked to whatever we defined in Central Administration.

In 2026, site collection configuration is streamlined but offers incredible depth. You no longer build dedicated "Search Centres". Instead, you customise the search experience directly where the users work.

Step-by-Step: Managing Modern Site Verticals

Search verticals are the tabs that appear across the top of the search results page (for example, All, Files, People, News). They filter results down to specific silos.5

Here is how you configure them today:

Navigate to your specific SharePoint site.
Select the gear icon in the top right corner to open the settings panel.
Select Site information, followed by View all site settings.5
Under the Microsoft Search section, select Configure search settings.5
In the left navigation pane, locate Custom experience and select Verticals.5
Select an existing vertical to edit, or select Add to create a brand new one.5

When creating a vertical, you use Keyword Query Language (KQL) to define its exact scope. For example, if you want a vertical that only shows PDF contracts and excludes web pages, you apply the KQL query NOT(FileType:htm) AND FileType:pdf.5

You can also use profile query variables to create dynamic verticals. By inserting {Profile.accounts.userPrincipalName}, the vertical automatically filters results to only show items created by or assigned to the currently logged-in user.5 This provides a highly personalised experience without needing custom code.

Key Takeaway: Use KQL query variables in your verticals to create dynamic, personalised search tabs. A "My Documents" vertical takes five minutes to build and saves users hours of manual filtering.

Implementing PnP Modern Search in 2026

When the out-of-the-box Microsoft Search verticals are not enough, we highly recommend deploying the PnP Modern Search solution. According to Collab365 research, deploying PnP web parts is the fastest way to recreate complex legacy search portals in a modern SharePoint environment.

Currently, version 4 is the industry standard, providing highly modular web components that communicate with each other on the page.11 You simply edit a modern SharePoint page and drop these web parts in.

Here is a breakdown of the four main PnP components you need to know:

1. The Search Box Web Part This replaces the old contextual search box. You can configure the placeholder text, decide whether it sends queries to a new page via a URL fragment (#), and enable query suggestions based on SharePoint static search suggestions.11 We usually set the input field height to 32px and use the Theme primary colour for the search button to keep it looking native.11

2. The Search Verticals Web Part This conditionally renders results based on the tab the user clicks. The best feature here is Audience Targeting. Unlike legacy 2010 scopes, you can bind an Entra ID security group to a specific vertical. If a user is a member of the "Legal Team" Entra ID group, they see the "Contracts" tab. If they are not, the tab is entirely hidden.11 This requires User.Read.All permissions for the SharePoint Online Client Extensibility Web Application Principal.11

3. The Search Filters Web Part These are your modern refiners. You connect the Filters web part to the Results web part. You can display filters as checkboxes, date ranges, or combo boxes.11 A massive quality-of-life improvement is the ability to merge filters. If you connect two different data sources that both have a "FileType" filter, the web part combines them into a single, seamless drop-down for the user.11

4. The Search Results Web Part This is the engine. It retrieves data and renders it. You can define how the data looks using Handlebars templates (.html) or Adaptive Cards.11

If the default layouts do not work for you, you can write custom web components. These act as HTML wrappers for React components, allowing you to pass complex JSON objects directly from your Handlebars template into the React component using the JSONstringify helper.11 It is incredibly powerful.

Key Takeaway: If you need to build a complex, audience-targeted search dashboard, do not write custom code from scratch. Deploy the PnP Modern Search web parts and use two-way connections to link your search box, filters, and results.

What Are the Modern Equivalents of Search Keywords and Best Bets?

In 2010, you likely managed "Search Keywords" and "Best Bets" in site settings. When the CEO wanted the new HR portal to appear at the top of the page whenever someone searched "vacation", you manually added a Best Bet.

Today, these concepts are replaced by Bookmarks, Q&As, and Promoted Results managed centrally in the Microsoft 365 admin center.8

Managing Bookmarks in 2026

Bookmarks are highly structured, visually appealing answers triggered by specific queries. As of March 2025, Microsoft retired Microsoft Search in Bing, meaning all your recommended bookmarks now live entirely within the M365.cloud.microsoft and SharePoint environments.8

To create a great bookmark today, you need four elements:

Title: Keep it under 60 characters. Avoid clickbait. "This week's cafeteria menu" performs better than "You won't believe what is on the menu".8
Description: A succinct 300-character summary of the resource.8
Keywords: Add at least five user-friendly terms. You can also set a "Reserved Keyword" to guarantee the bookmark triggers exclusively for a specific phrase, overriding all other search logic.8
Categories: Use these for your own internal administrative sorting; end-users never see them.8

You can manage these manually in the admin center, or use the Microsoft Search content creator extension for Chrome and Edge to add bookmarks directly while browsing your intranet.8

Advanced Administration via Microsoft Graph API

If you have hundreds of bookmarks, clicking through the admin center is slow. We manage these programmatically using the Microsoft Graph API (POST /v1.0/search/bookmarks).8

The API payload allows for advanced targeting that we could only dream of in 2010. You can configure:

availabilityStartDateTime: Perfect for scheduling a new IT policy bookmark to go live at midnight on Monday.8
groupIds: Restrict the bookmark so only specific Entra ID security groups see it. The finance team gets a different bookmark for "expenses" than the marketing team.8
targetedVariations: Deliver a different URL based on the user's device (iOS vs Windows) or geographic region. A user in London gets the UK benefits portal, while a user in New York gets the US portal.8
powerAppIds: This is the best feature. You can embed a Power App directly into the search result. Users can submit expenses or log IT tickets right from the search page without ever navigating away.8

Key Takeaway: Stop manually adding single links. Utilise the Graph API to deploy targeted Bookmarks mapped to Entra ID groups, and embed Power Apps directly into the search interface for immediate user action.

Table: Legacy Keywords vs 2026 Promoted Results

Feature Category	SharePoint 2010	Microsoft 365 (2026)
Terminology	Best Bets / Visual Best Bets	Bookmarks / Promoted Results / Answers 8
Targeting	Basic Audience Targeting	Entra ID Groups, Device OS, Geographic Region 8
Creation Method	Site Settings GUI	M365 Admin Center, Browser Extension, Graph API 8
Interactive Elements	Static Links and Images	Embedded Power Apps directly in search results 8
Analytics	Basic Web Analytics	Copilot Dashboard, connection analytics, click-through rates 13

Service-Level Admin: Setting Up Content Sources and Crawling in 2026

The transition from on-premises to the cloud fundamentally altered crawling administration. We used to spend our Fridays configuring content sources, mapping server drives, and scheduling incremental crawls to run overnight so the servers would not crash.

Continuous Indexing Replaces Scheduled Crawls

In Microsoft 365, SharePoint and OneDrive content is governed by Continuous Indexing. The moment a user uploads, modifies, or deletes a document, the search index updates almost instantaneously. There is no administrative interface to schedule a "Full Crawl" for native Microsoft 365 data.1 The system handles the load balancing and indexing behind the scenes.

Microsoft Graph Connectors: Extending the Reach

If you need to search data that lives outside of Microsoft 365—like ServiceNow tickets, Jira issues, or an old on-premises file share—you use Microsoft Graph Connectors.

You manage these by navigating to Microsoft 365 admin center > Search & intelligence > Data sources.7 By 2026, the architecture offers two very different pathways for connectors.

1. Synced Connectors (The Standard Approach) Synced connectors physically crawl your external systems and ingest the data into the Microsoft Graph index.7

Access Model: Configured at the organisation level by administrators.7
Setup: You can map up to 128 custom properties. You must assign semantic labels (e.g., Title, Created By, Last modified by) to your external data fields. This is crucial because it tells Copilot how to read and render the data.5
Schedules: Because this data is external, you do have to configure sync schedules. A typical best practice is to run an incremental crawl every 15 minutes and a full crawl once a day.10

2. Federated Connectors (The Modern Live Fetch) Introduced in early access preview for Targeted release members, Federated connectors solve the problem of highly sensitive or rapidly changing data.14

Architecture: Instead of pulling data into the Graph index, these connectors use a Model Context Protocol (MCP) to fetch data live at the exact moment the user submits a query.14
Access Model: Federated at the user level. The system uses the individual user's credentials to authenticate against the external system in real-time.14
Data Residency: No external content is synced or stored within the Microsoft Graph, making compliance teams very happy.15

Key Takeaway: Use Synced Connectors for static external data like wikis or standard file shares. Use Federated Connectors for highly sensitive, volatile data where you absolutely cannot store a copy of the information in the cloud.

The SharePoint Server Connector (Hybrid Configuration)

Many of the organisations we work with still maintain legacy on-premises servers. For these scenarios, the SharePoint Server Microsoft 365 Copilot connector is essential.10 It supports SharePoint Server 2016, 2019, and Subscription Edition (SPSE).10

To set this up, you install the Microsoft Graph connector agent on a machine with access to your farm.10 You must configure an Entra ID App Registration and add the user_impersonation scope.10

Next, you run a PowerShell script in the SharePoint Management Shell to establish the OIDC trust. It looks like this:

$t \= Get-SPTrustedIdentityTokenIssuer \-Identity "\<TrustedIdentityTokenIssuerName\>"
$uri \= New-Object System.Uri("\<SharePointSiteUrl\>")
$t.ScopedClientIdentifier.Add($uri, "\<EntraIdAppIdentifierUri\>")
$t.Update()

You must run that .Add() command for every unique URL you intend to crawl.10

A critical prerequisite here is Active Directory synchronization. You must use Entra Connect Sync to mirror your on-premises identities to the cloud. If you skip this, security trimming fails, and users might see documents in their Copilot results that they do not have permission to read.10

Table: Legacy Crawling vs 2026 Graph Connectors

Crawl Characteristic	Legacy SharePoint Server	Microsoft 365 Graph Connectors (2026)
Execution	Manual Schedules (Full/Incremental)	Continuous (Native) / Synced or Federated (External) 7
Protocol	BCS / Custom Search Connectors	Microsoft Graph API / Model Context Protocol (MCP) 15
Security Trimming	Active Directory ACLs	Entra ID Sync, Live User Credential Federation 10
Property Mapping	Crawled to Managed Properties	Direct mapping to Semantic Labels (e.g., title, url) 5

Optimising Queries and Results: Authoritative Pages, Scopes, and Exclusions Today

If users complained about irrelevant results back in the day, our immediate response was to adjust search scopes or manually elevate Authoritative Pages in Central Admin. Today, relevance is handled algorithmically via artificial intelligence, but as administrators, we still need to guide the system.

Semantic Search and the Reranker

The most significant upgrade to search administration is the introduction of the semantic ranker (L2 ranking). According to Microsoft benchmarks, this capability provides an 85% relevance boost.2

Here is how it actually works. When a user types a query, the system first executes a traditional keyword retrieval—using BM25 or Reciprocal Rank Fusion (RRF)—to gather a broad set of candidate documents.2

Then, the semantic ranker takes over. Using deep learning language models adapted from Microsoft Bing, it evaluates the intent and contextual meaning of the query against the retrieved documents.2 It computes a new relevance score (@search.rerankerBoostedScore) and reorders the results entirely.16

Furthermore, the ranker automatically extracts semantic captions. These are verbatim sentences pulled from the documents that best summarise the content, highlighting the key passages right on the search page so users can scan them quickly.2

You do not need to build complex query rules to achieve this. It is a fundamental part of the Microsoft 365 architecture, billed by usage in Azure AI Search but built into the core M365 experience.2

Key Takeaway: Stop trying to manually tweak keyword weights. Ensure your content is properly tagged, rely on the Semantic Ranker to understand user intent, and let the AI do the heavy lifting for relevance.

Managing Exclusions

There are times when data must be completely hidden from the semantic index and search results. Think of draft financial models, archived project sites, or sensitive HR sandboxes.

To exclude a site:

Navigate to the specific SharePoint site settings.
Select Search and offline availability under the Search category.17
Select No for "Allow this site to appear in search results".17

This action completely removes the site from both Microsoft Search and the Copilot semantic index.17 You cannot exclude a site from Search but keep it available for Copilot; the exclusion applies strictly to the entire tenant-level index.17

Custom Filters and Property Aliasing

To refine results on the search page, you configure Custom Filters in the Microsoft 365 admin center.5

A major optimisation technique we use constantly is merging properties across different content sources using a common alias. For example, if your organisation uses both ServiceNow and Jira, the system will naturally output two different author fields.

You can map the Owner property from ServiceNow and the Publisher property from Jira to a single alias called Author.5 This creates a unified, single drop-down filter for the end-user. Instead of clicking through three different filters to find who wrote a ticket, they click one. It drastically improves the search experience.5 Note that you can only configure filters for text and dateTime properties, and dropdowns max out at 50 values.5

Server Name Mappings and Security: What's Replaced Them?

In SharePoint 2010, Alternate Access Mappings (AAM) and server name mappings were critical. If you got them wrong, internal users could not resolve links, and external users were met with broken search results.

The Death of Alternate Access Mappings

In Microsoft 365, AAM is entirely obsolete. The cloud architecture relies on host-named site collections and unified routing.18

If an organisation needs to change a URL, you use the SharePoint admin center or the Start-SPOSiteRename PowerShell cmdlet.19 The cloud infrastructure automatically generates a permanent redirect from the old URL to the new one. This ensures your search indexes, browser favourites, and existing sharing links do not break.19

You must note a few limitations: renames are not supported for sites that have the Publishing feature activated, or sites containing more than 15 million documents.19 During the rename process, which takes about 10 minutes, the site is read-only. You will also need to manually republish any SharePoint 2013 Workflows or Power Apps connected to the site.19

Key Takeaway: Never attempt to manually build URL redirects. Use Start-SPOSiteRename. Let Microsoft 365 handle the routing automatically to keep your search index intact.

Modern Security: Microsoft Purview Integration

Security trimming in 2010 was limited to basic Active Directory access control lists (ACLs). If a user had read access to a folder, they saw the files in search.

In 2026, Microsoft 365 Search is deeply integrated with Microsoft Purview.20 You do not manually connect your M365 workloads to Purview; they are automatically integrated once you enable Information Protection.20

When data is crawled by the semantic index, the system strictly respects Microsoft Purview Information Protection policies:

Sensitivity Labels: If a document is labelled "Highly Confidential," the search engine enforces those encryption and access rules at the query level.4 Even if Copilot reads the file, it will not summarize it for an unauthorised user.
Data Loss Prevention (DLP): Search results will not expose snippets or semantic captions of documents containing sensitive data (like credit card numbers or passport details).4
Just-In-Time (JIT) Protection: You can configure JIT access. If a user finds a sensitive document via search, they cannot open it immediately. They must request temporary elevation, which triggers an approval workflow.4
eDiscovery (Premium): Legal and compliance teams use the eDiscovery portal to search across all mailboxes, sites, and connected apps (via Graph Connectors) simultaneously. They can place legal holds on data directly from the search interface without needing IT intervention.21

Reporting and Monitoring Microsoft 365 Search Performance

Without robust analytics, you cannot identify why users are failing to find data. The legacy web analytics reports from 2010 were clunky and hard to export. Today, they have been completely replaced by the Search usage reports and the Copilot Dashboard in Viva Insights.22

Microsoft Search Usage Reports

Accessible via the Microsoft 365 admin center, the usage reports provide a 28-day or annual view of search activity across SharePoint, Office.com, and Windows Search.22

The dashboard presents several critical metrics:

KPI Cards: The top level of the dashboard provides high-level figures, including Total Queries, the Zero-Result Rate, and Active Copilot Users.22
Top 5 Zero-Result Queries: This is presented as a bar chart. It plots the exact query terms that users searched for but received absolutely no results.24
Connection Analytics Click-Through Rate: Presented as a line chart over a 28-day period. This tracks how often users actually click on results generated by your third-party Graph Connectors (like Jira or ServiceNow).13

The Zero-Result Queries metric is your daily to-do list. The Collab365 team recommends reviewing this list every Friday. If you see that 50 users searched for "2026 Holiday Schedule" and got zero results, you immediately create a Bookmark for that phrase pointing to the HR portal.

Viva Insights Copilot Dashboard

With the rise of generative AI, monitoring Copilot usage is just as important as monitoring traditional search. Global Administrators use the Copilot Dashboard to track deep adoption metrics.23

You can track Active Copilot Users, which counts anyone who interacted with Copilot in the last 28 days.23 More importantly, you track Total Copilot Actions Taken. This measures actual engagement—such as a user expanding an AI-generated document summary—rather than just passive queries.23

If you need to build custom reporting for your executives, you can export de-identified, user-level metrics spanning six months. This allows you to build custom Power BI dashboards to analyse adoption rates across different departments, proving the ROI of your search and AI investments.25

Key Takeaway: The Zero-Result Queries report is your most valuable administrative tool. Review it weekly. Use it to deploy targeted Bookmarks and fix knowledge gaps before users open support tickets.

How Does Copilot and Semantic Search Change Everything?

In 2026, search administration is no longer just about matching keywords; it is about Generative Engine Optimisation (GEO).26

When an end-user queries Microsoft 365 Copilot, the system does not just display a list of "blue links." Instead, it uses a massive Semantic Index to synthesize a conversational answer. The Semantic Index maps the complex relationships between users, files, meetings, and external data sources.17

This architecture is powered by advanced reasoning models like GPT-5.2 and cross-encoding AI like Rerank 4.27 These models are designed to understand long-context queries and solve multi-step problems.27 For example, if a user asks Copilot, "how do I fix a slow database," the system will retrieve and summarize a document titled "Optimising Database Performance", even if the exact words "fix" and "slow" do not appear anywhere in the text.28 It understands the concept of the problem.

For administrators, this means your focus must shift from manipulating search algorithms to ensuring your underlying data is clean, well-structured, and correctly labelled. If your organisation's metadata is a mess, the LLM will synthesize inaccurate answers.

The Collab365 team stresses that implementing Microsoft Purview lifecycle management to archive outdated content is the single most effective way to improve Copilot accuracy. If you leave ten-year-old HR policies in an active SharePoint site, Copilot will eventually read them and give an employee the wrong advice. You must declutter your data sources.29

Furthermore, you can use Copilot Studio to build custom declarative agents. Using schema version 1.6, you author a JSON manifest that gives the LLM specific instructions, knowledge, and actions.10 You can define capabilities that allow the agent to search specific SharePoint sites, query Graph connectors, or even read specific shared mailboxes to ground its answers in your enterprise truth.10

Comparison: On-Premises SharePoint Search vs Microsoft 365 in 2026

To understand the magnitude of the architectural shift, review this comprehensive comparison between legacy on-premises environments and the modern cloud ecosystem.

Search Component	Legacy SharePoint (2010/2013)	Microsoft 365 (2026)
Indexing Engine	Scheduled Full/Incremental Crawls	Real-time Continuous Indexing 1
Relevance Model	BM25 / Keyword Frequency	Semantic Ranker (L2), Neural intent matching 2
External Integration	Business Connectivity Services (BCS)	Microsoft Graph Connectors (Synced & Federated) 7
Promoted Results	Best Bets / Keywords	Bookmarks with Graph API & Power App integration 8
User Interface	Search Centre Site Collections	Copilot, SharePoint Home, PnP Modern Search 11
URL Management	Alternate Access Mappings (AAM)	Native redirects via Start-SPOSiteRename 18
Security Enforcement	Basic AD ACLs	Microsoft Purview, Sensitivity Labels, JIT access 4
Analytics	Web Analytics Web Part	Viva Insights Copilot Dashboard, Connection Analytics 13

Structured FAQ

1. Do I still need to configure full and incremental crawls?

No. For native Microsoft 365 content (SharePoint, OneDrive, Teams), the system utilises continuous indexing in real-time.1 You only configure manual sync schedules when ingesting external data via Synced Graph Connectors.10

2. How do I integrate external databases without duplicating data?

Administrators must deploy Federated Graph Connectors. Currently in early access preview, these connectors use the Model Context Protocol (MCP) to fetch data live from external systems based on the user's credentials, without indexing or storing the data in Microsoft 365.7

3. How do I enable the Semantic Ranker for my tenant?

Semantic indexing and ranking are enabled automatically for tenants with Microsoft 365 Copilot licenses.17 The system processes the background indexing to map relationships, requiring no manual activation of the L2 ranker.2

4. Users complain about "Zero Results" for common terms. How is this fixed?

Administrators must review the Query Analytics report in the Microsoft 365 admin center.24 Identify the top failing queries, then create targeted Bookmarks via the Search & intelligence portal to explicitly route those queries to the correct internal resources or embedded Power Apps.8

5. Can we keep our on-premises data but search it from the cloud?

Yes. Administrators must install the SharePoint Server Microsoft 365 Copilot connector via a Graph connector agent.10 This requires Entra Connect Sync for security trimming and OIDC trust configuration using PowerShell (Get-SPTrustedIdentityTokenIssuer) to ensure on-premises permissions are respected in the cloud.10

Next Steps

The search landscape in 2026 is driven by AI, continuous indexing, and deep integrations across the Microsoft Graph. Administrators must transition away from legacy mindsets involving manual crawls and rigid site scopes.

Test these configurations in your tenant now. Begin by reviewing your zero-results report, deploying a custom PnP Search dashboard for a specific department, and auditing your existing data for Copilot readiness using Microsoft Purview.

For continuous learning check out the Microsoft 365 Space on Collab365 Spaces.