Active-User Pricing: Examples & Companies

What is it

Active-User Pricing is a billing unit where customers are charged per monthly or daily active user rather than per provisioned seat. It is the seat model with the waste squeezed out: nobody pays for the licenses sitting unused after a big-bang rollout, because the meter counts engagement, not entitlement.

Dust is the corpus’s cleanest specimen. Its self-serve Pro tier is a conventional €29/user/month seat, but Enterprise is quoted on active users — and the active user runs through the whole commercial design: programmatic usage overage is capped at $50 per active user (and $1,000 per cycle), so the same unit that meters the bill also bounds it. Glean runs the hybrid version: per-user Enterprise Flex seats deployed “to every employee,” with a pooled FlexCredits layer absorbing the variance in how intensively each of them uses premium AI.

The contact-center pair shows the unit’s enterprise ancestry: Observe.AI licenses per agent (reportedly from ~$69/agent/month for a single module via AWS Marketplace, with ~100-seat minimums on direct deals) and Uniphore blends per-agent licensing with per-interaction consumption — in both cases the “user” is a deployed call-center agent whose headcount the buyer already tracks daily.

How it works

Worked example — what inactivity is worth. A 1,000-employee company rolls out an assistant and 320 people use it in a given month. On provisioned seats at €29, the bill is €29,000; on an active-user basis at the same rate it’s €9,280 — a 3.1x difference that is pure adoption risk, transferred from buyer to vendor. That transfer is why active-user vendors add floors and minimums underneath: Dust starts Enterprise at 100 members, and the contact-center vendors set seat minimums that assume the workforce is active. The usage-metric guide treats this as the canonical case of choosing the metric the buyer already trusts — every company audits its own active headcount.

Companies using this

6 in-corpus companies bill on active users: Dust and Glean among horizontal enterprise assistants, Observe.AI and Uniphore in contact-center AI, and Character.AI and the discontinued Phind on the consumer/prosumer end.

Patterns observed

The unit never travels alone. Every enterprise vendor here pairs the active-user meter with a second mechanism that bounds variance: Glean’s pooled FlexCredits absorb intensity differences between users, Dust’s credit ladder and per-active-user overage caps bound the programmatic bill, and Uniphore rides a per-interaction consumption component on top of the agent license. Pure presence-based billing — pay only when someone shows up — exists nowhere in the corpus without a floor, a pool, or a cap attached, because neither side can plan around a naked fluctuating denominator.

The second pattern is that the unit thrives where the buyer already counts heads operationally: contact centers audit agent rosters daily, which is why Observe.AI and Uniphore can anchor six-figure contracts to per-agent counts that nobody disputes at renewal.

Counterexamples & variants

Character.AI marks the consumer boundary: with no organization in the loop, “active user” collapses into the subscriber themselves — a $9.99/month c.ai+ plan where activity gates experience (peak-hour priority, no ads) rather than the bill. Phind is the cautionary variant: a flat ~$20/month prosumer subscription and ~$40/user Business tier that captured nothing extra from heavy users — and the company shut down weeks after raising ~$10.4M, an arc the corpus reads as flat freemium’s weakness under cost pressure rather than an indictment of user-based billing itself. And the broader counter-model is everywhere in the adjacent cohorts: observability and billing platforms deliberately make users unlimited and meter ingestion instead, betting that seat friction — active or not — fights adoption.

What this means for buyers vs vendors

For buyers

Get the activity definition in writing — daily vs monthly active, and what action counts — because it moves the denominator more than the rate. Check what sits underneath: minimums and floors (Observe.AI’s ~100 seats, Dust’s 100-member Enterprise threshold) can make the “pay for actual usage” story mostly notional at small scale. And use the model’s fairness in negotiation: if a vendor insists on provisioned seats for an uncertain rollout, an active-user clause with a floor is the natural compromise position.

For vendors

Active-user pricing is an adoption-risk transfer you should price deliberately: take it on when your product’s engagement is genuinely sticky and auditable, and bound it with the mechanisms the corpus validates — floors for your downside, per-user overage caps for theirs (Dust’s $50/active-user cap is the template), and a pooled credit layer where usage intensity varies more than headcount. Define activity events precisely in your metering pipeline from day one; an active-user invoice that the customer’s own analytics can’t reproduce is a dispute, not a bill.