The AI budget conversation has shifted. For the past two years, companies bought AI tools on the promise of productivity. Work faster, automate the repetitive tasks, save time. Finance teams largely went along with it. Now those same teams are reviewing the line items, and the numbers are not always flattering.
Glean has turned that scrutiny into a sales opportunity. The company's product is an AI-powered search layer that connects to everything a company uses internally: email, documents, project management tools, customer data, HR systems. Employees ask questions in plain language and get answers drawn from across all of those sources.
The revenue growth is significant. Glean went from $100 million in early 2025 to $200 million by December 2025, and now reports $300 million. That is a tripling in 15 months, a pace that most established software companies do not come close to.
The cost-cutting angle is the key to understanding why. When AI tools query a company's data directly, they process enormous amounts of irrelevant content to surface a single useful answer. Glean acts as a filter. It builds a structured map of how a company's information is organised, what belongs to which project, who works on what, which documents are current. The AI then only processes what it actually needs. Glean claims this reduces AI processing costs by around 30 percent.
For a company running hundreds of employees on AI tools, 30 percent is not a minor saving. It is the kind of number that gets a CFO to take a meeting.
The usage numbers support the adoption story. More than 85 percent of Glean customers use it across five or more departments, meaning it has spread beyond a single team's experiment. Nearly half of monthly users return daily, which is more than twice the typical engagement rate for enterprise software.
The competition is serious, though. Microsoft, Google, Salesforce, OpenAI, Anthropic, and Atlassian are all building tools that do overlapping things. Microsoft and Google in particular can bundle AI search directly into the productivity software their customers already pay for, making Glean an additional line item rather than a replacement.
Glean's response to this is platform neutrality. It supports more than 15 different AI models from providers including Anthropic, Google, and OpenAI, accessible through major cloud platforms. Customers can use models they have already contracted for, or models that meet their data rules in regulated industries. No single AI provider controls how Glean works, which reduces the vendor lock-in concern that procurement teams in manufacturing, finance, and insurance tend to raise.
The $300 million figure also needs some context. Part of Glean's revenue comes from usage-based pricing, where customers pay per query rather than a fixed monthly fee. That means a portion of the reported number is an estimate based on current usage rather than a locked-in contract value. It does not make the growth less real, but it does mean the revenue could move in either direction as usage patterns change.
Pricing is not transparent. Contracts reportedly start at $50 or more per user per month, with a minimum seat count, and advanced features are often separate add-ons. Implementation and integration work adds further cost. For a mid-sized organisation evaluating Glean, the total bill will be higher than the base license.
The underlying trend is durable regardless of what happens to Glean specifically. Companies have accumulated dozens of software tools over the past decade. Information is scattered across all of them. Employees spend time hunting for things they know exist but cannot find. Any tool that genuinely solves that problem, and can show measurable savings on the AI processing costs that are now on every finance team's radar, is selling something real.