The Quiet Shift in Enterprise Search
Glean built its reputation on a simple but powerful premise: connect every tool in your company’s stack – Slack, Google Drive, Salesforce, Confluence – and let employees search across all of it with one query. For a few years, that promise was enough to make Glean the default answer when enterprise IT teams asked, “How do we fix our internal search problem?” The company raised at a valuation that signaled serious investor confidence in that thesis. But the thesis is getting complicated.
Notion AI is not positioning itself as an enterprise search tool. That’s exactly the problem for Glean. Notion has spent the past 18 months quietly building AI capabilities directly into a workspace that millions of knowledge workers already live inside every day. No new tab, no new login, no change management project required. The search comes to the user, and the user barely notices it happening.
What Glean Actually Does Well
To understand where Glean is losing ground, it helps to understand where it genuinely excels. Glean’s core product is a universal search layer that indexes third-party tools and surfaces results through a single interface, with permissions respected at the source level. For large enterprises with hundreds of integrated apps and thousands of employees who need to find documents across systems they don’t even directly log into, that architecture is genuinely difficult to replicate. Glean’s connector library is extensive, and its ability to handle permission-aware search – meaning employees only see results they’re authorized to see – is a real engineering achievement.
The sales motion has historically targeted IT leaders and CIOs, not individual contributors. That’s a high-trust, high-contract-value relationship that doesn’t evaporate overnight. Glean’s enterprise customers tend to be large organizations where procurement cycles are long, switching costs are real, and the people who bought Glean are not the same people who are quietly switching their personal workflow to Notion AI. This distinction matters because Glean’s slippage is not yet a cancellation story. It’s an attention story.
How Notion AI Is Winning the Attention War
Notion AI’s strategy is built on a different assumption about where enterprise knowledge actually lives. Rather than indexing everything, Notion is betting that an increasing share of meaningful work documentation flows through Notion itself – meeting notes, project specs, onboarding docs, team wikis. If that assumption holds, then an AI that reasons over your Notion workspace is solving the actual problem for most employees most of the time, without requiring any infrastructure work from IT.
The product experience difference is stark. When a Notion user highlights a document and asks the AI to summarize it, find related pages, or draft a response based on existing content, the interaction takes seconds and requires no setup. When a new employee wants to search for context on a project, they do it inside the same tool where they’re already writing. There is no onboarding friction because there is no separate tool. That frictionlessness is quietly pulling daily search behavior away from dedicated search products.
Notion AI also benefits from a pricing dynamic that enterprise search tools struggle to counter. Notion’s AI features are folded into existing workspace plans, which means buyers are not approving a separate line item for search. The value appears embedded, not additional. Glean, by contrast, is a dedicated spend that requires justification, renewal, and ongoing advocacy from an internal champion. When budgets tighten, dedicated tools with narrow use cases face harder renewal conversations than tools that do ten things at once.
There’s also a product roadmap question that Glean has not fully answered publicly. Notion AI is expanding its ability to connect to external sources, including Google Drive and Slack, through integrations that bring more of the enterprise knowledge graph into Notion’s context window. If Notion closes the connectivity gap – even partially – the argument for maintaining a separate search layer weakens considerably for mid-market companies that aren’t running Fortune 500-scale infrastructure.
The Mid-Market Is the Battleground
Glean’s strongest defense is the enterprise tier – companies large enough to have genuine cross-system search complexity, dedicated IT teams, and the budget to pay for a specialized solution. But the mid-market, the 500-to-5,000-employee range, is where Notion AI is making the most visible inroads. These are companies sophisticated enough to think carefully about knowledge management but not so complex that Glean’s full connector architecture is strictly necessary. Many of them already standardized on Notion as their documentation layer years ago.
For a 1,200-person company where 80% of documented knowledge lives in Notion, the case for adding Glean is genuinely hard to make now that Notion AI exists. The remaining 20% – the stuff in Salesforce, in Jira, in email – matters, but it matters differently for different teams. Sales might need Salesforce search. Engineering might need Jira. But the company-wide search layer that Glean promises to be is harder to justify when the tool where most people work already has AI-assisted search built in.
Glean’s Next Move
Glean is not standing still. The company has invested in building a more conversational, agent-like interface on top of its search infrastructure, positioning itself less as a “search box” and more as an AI assistant that can take actions across systems. This is the right directional move – pure search is a feature, not a product, and Glean clearly understands that. The question is whether the agent-layer vision can be executed and communicated quickly enough to hold the attention of buyers who are already comfortable with Notion AI’s more modest but immediately useful implementation.
Glean also benefits from something Notion cannot easily replicate: deep enterprise security and compliance positioning. For industries with strict data governance requirements – financial services, healthcare, government contractors – Notion’s hosted AI model raises questions that Glean’s enterprise architecture is designed to answer. That’s a real moat, but it’s a narrow one, and it doesn’t protect Glean’s position in the broader market where compliance is a factor but not the deciding factor.
The competitive pressure Glean is facing from Notion AI rhymes with patterns seen elsewhere in the productivity software market – a focused, standalone tool getting squeezed by a broader platform that absorbs the use case without announcing it as a direct attack. What makes this particular situation worth watching is that Glean’s pitch was always about connecting the entire knowledge graph, not just one tool’s content. If Notion keeps extending its integration reach, and if enterprises keep consolidating their documentation in Notion, Glean’s most important customers may start to ask a question that no vendor wants to hear: do we still need this?
