A 100-seat floor and a ~$60K ACV starting point are doing more shape-changing to the mid-market buying decision than anyone at Glean admits.
Publisher disclosure: This site is published by ASCENDING, which builds Jarvis AI — a competing agent platform. The figures below come from public reporting (Vendr, GoSearch, Workativ) and procurement conversations through Q1-Q2 2026. Glean has not published a public price list.
Most quotes I've seen for Glean in the first half of 2026 open the same way. The AE asks how many knowledge workers you want covered. You answer with a real number — 62, 75, 110. Then the email lands and the order form says 100 seats. Sometimes 150. The bottom of the page reads Term: 12 months. Minimum commit: 100 users. Annual subscription, paid up front.
That's the floor. It does most of the work in deciding who Glean is for, and it's almost never the line the AE leads with.
The 100-seat floor and what it really means
A 100-seat minimum is a contractual commitment to license at least 100 named users for the term, regardless of how many you actually have. At Glean's reported $45-50 per seat base, plus the $15 Work AI add-on for the agentic features anyone bought Glean to use, the math lands at roughly $60,000-$78,000 in year-one subscription before professional services, the $70,000 paid POC fee, or the 10% support uplift. Vendr's marketplace data, the cleanest third-party source on this, has been consistent on the 100-user floor since late 2025. GoSearch and Workativ both reported the same number in their 2026 pricing teardowns. Glean's own collateral talks about "enterprise pricing" without disclosing the floor; you only learn it from the order form. For the broader pricing teardown, see Glean pricing in 2026 and the cheaper paths to the same outcome.
What that floor really means in practice: if your knowledge-worker count is under about 80, you're paying for empty seats. If it's between 80 and 120, you're paying for some empty seats and the unit economics start to wobble. If it's over 200, the floor is invisible and you're a normal Glean customer.
How Glean's seat floor compares to peers
The minimum is not unusual. Every enterprise-grade vendor has one. But the level matters, and it's instructive to see them side by side.
| Vendor | Reported seat floor | Notes |
|---|---|---|
| Glean | 100 seats | Enterprise contracts only; flex pricing reported through Vendr |
| Microsoft 365 Copilot | 300 seats (E3/E5 tenants) | Effectively no floor below E3 ceiling, but tenant-scale gating |
| Onyx (open source) | None | Self-host; commercial support starts ~20 seats |
| Workativ | 25 seats | Mid-market positioned; smallest reported floor in the segment |
| Guru | 50 seats (Enterprise tier) | Lower tiers available without floor |
Sources: Vendr marketplace listings, vendor sales conversations, GoSearch and Workativ 2026 pricing reports.
The pattern is clear. Glean and Microsoft sit at the top end of the floor distribution because they're enterprise-list products built around enterprise sales cycles. Workativ, Guru, and self-hosted Onyx sit lower because they sell into a segment where 50-200 employees is the target, not the rounding error.
Why the floor exists
The floor is not arbitrary. It's the output of a CAC-payback equation.
A Glean sales cycle costs the company real money. AE time, SE time, a paid 6-8 week POC, security-review hours, legal back-and-forth on the order form. Vendr's data and Workativ's reporting both put the all-in cost of closing a Glean enterprise deal somewhere in the $30,000-$50,000 range from a vendor-cost perspective. Below 100 seats and roughly $60,000 ACV, the payback period stretches past the point where the CFO's spreadsheet stops working. Glean's Series F round in 2025, at about $7.2B valuation per public reporting, added pressure on per-customer ACV, not less. The floor is how they protect gross margin on a high-touch sale. None of this is published anywhere by Glean directly. It's reconstructed from procurement data, sales-cycle benchmarks reported by Workativ and Vendr through 2026, and the consistent pattern of order forms that customers have shared with us through Q1 and Q2. Treat the math as third-party reporting, but treat the floor itself as confirmed.
You can argue the model is rigid. You can't argue it's irrational. Glean would say the floor exists because the product genuinely takes 100+ users to demonstrate value across enough connectors. There's some truth to that — search-quality and ranking depend on usage signal. The rebuttal is that "we need scale for ranking signal" is also a convenient story for "we need scale for our CAC math."
Three scenarios where the floor blocks you
Scenario one: the 175-person legal SaaS at $40M ARR
You're a fast-growing legal-tech company. Head count is 175. The actual knowledge-worker population, lawyers, paralegals, product, customer success, the people who would touch Glean weekly, is 65. Glean quotes 100 seats. You're paying for 35 ghosts.
The first-pass math: 100 × $45 × 12 = $54,000 base, plus the $15 Work AI add-on on the 100 = $18,000, plus the 10% support uplift = $7,200. Year-one subscription lands around $79,000 for value being delivered to 65 humans. That's $1,215 per active user. A comparable Workativ deployment at 65 actual seats would cost roughly $30,000-$40,000 total in the same band, depending on negotiated rate.
Scenario two: the 90-person fintech in a SOC 2 push
Head count 90, all knowledge workers. You want Glean for the audit-prep workflow specifically — the 6-week SOC 2 sprint where everyone needs faster retrieval over the policy library. Glean still asks for 100 seats and a 12-month term. You don't have 100 employees, and you don't need a year of it. The floor pushes you into either a self-hosted alternative (Onyx) or a shorter-term Workativ deployment.
Scenario three: the 240-person manufacturer with a 60-person pilot
You're large enough to clear the floor on paper. Head count 240. But the buying committee wants a 60-person pilot in operations before rolling site-wide. Glean's pricing structure makes the pilot the same as the full deployment — 100 seats minimum either way. The "pilot then expand" motion that works on most SaaS doesn't apply here. Either commit to the 100 immediately or run a $70,000 paid POC for the 60 people and pay for the production licenses later regardless.
What 50-250-employee orgs should evaluate instead
If you're sized out of Glean, the shortlist looks different from the one Glean's competitive deck shows. Three options worth a real evaluation:
- Workativ. Mid-market positioned with a 25-seat floor. Lighter connector library than Glean (the gap is real — fewer SaaS sources out of the box), stronger on conversational automation. Pricing scales linearly from the floor.
- GoSearch. Search-first product with no published seat minimum and per-seat pricing that targets the 50-300 employee band specifically. Connector coverage is narrower than Glean's; if your stack is Google Workspace plus three SaaS apps, this is fit-for-purpose.
- Onyx. Open-source, self-hosted. Zero seat floor. You pay in DevOps hours and infrastructure instead of license, which only makes sense if you have a platform team to run it. Commercial support is available from the Onyx team starting around 20 seats.
For a deeper side-by-side that includes Microsoft Copilot, Guru, and the regulated-industry angle, see the Glean alternatives 2026 shortlist.
If you decide to push past the floor anyway
Sometimes the floor is annoying and the product is still right. If Glean genuinely fits your roadmap and the 100-seat math only mildly overshoots, push the AE on three levers before signing.
Get the unused-seat reclaim language into the order form. Standard Glean contracts don't let you reclaim mid-term; some have been negotiated to allow a true-down at month six on documented under-utilization. The lever you have at the negotiation table is "show me the unused-seat reclaim clause or I walk to Workativ." Customers who walked into the conversation with a written alternative quote got better outcomes than customers who hadn't. Push for a feature-credit if the agent SKUs in the quote aren't GA in your region on day one, same pattern as the renewal checklist we published last month: see the Glean renewal checklist. And get the data-egress format in writing before signing, so the floor doesn't quietly become a switching cost twelve months from now. For the full pricing teardown across deployment sizes, the Glean pricing teardown covers the math past the floor.
The 100-seat floor isn't the reason Glean is or isn't right for you. It's the reason the conversation takes a different shape at 175 employees than at 1,750. Knowing that going in is the difference between a procurement cycle that lands and one that ends in three months of wasted AE calls.
FAQ
Can you negotiate Glean below 100 seats? Rarely, and almost never on first quote. We've seen Glean accept 75 seats in two specific cases — both were strategic-logo deals where the customer was a recognizable brand in a target vertical and the AE could justify the exception to RevOps. Neither was open-market mid-market.
What's the practical ACV floor on a Glean deal in 2026? Roughly $60,000-$78,000 year-one subscription on the 100-seat minimum, depending on whether the Work AI add-on is included. Add $70,000 if you go through the paid POC and another 10% for support. The all-in first-year exposure for a mid-market buyer typically lands between $150,000 and $220,000.
Is the 100-seat floor industry-standard? No. Workativ and GoSearch operate without it. Microsoft 365 Copilot has tenant-scale gating that effectively imposes a much higher floor for E3/E5 buyers. The 100-seat number sits in the middle of the enterprise-AI-search distribution, not at the bottom.
What happens if we drop below the seat count mid-term? You pay the contracted amount regardless. Glean does not pro-rate down. Some negotiated deals include a true-down right at six months, but you have to ask for it before signing — it's not in the standard order form.
Does Glean's flex pricing waive the minimum? The reported flex pricing tier still carries a seat floor in the same band. It changes payment cadence, not the commitment. Verify against the current order form; Glean's pricing has been moving fast through 2026.
About ASCENDING
ASCENDING is an AWS Advanced Consulting Partner founded in 2018 in Fairfax, Virginia. We build Jarvis AI, a governance-first, MCP-native agent platform, and publish this site as an independent industry resource. When an article compares Jarvis to a competitor, the disclosure stays on the page.