Total cost of ownership for support AI
The license fee is the smallest part of what support AI really costs. Total cost of ownership includes the per-resolution meter, the seats for tickets the AI does not resolve, the admin staffing to keep it configured, the setup time before it earns anything, and the churn cost of confidently wrong answers. This page adds up the full picture so you can compare tools on their real cost, not their sticker.
The five layers of TCO
- Platform and per-resolution fees. The base plan plus any per-resolution meter (Fin near $0.99, Zendesk near $1.50).
- Seats for the gap. Every ticket the AI does not resolve needs a paid human. At a real 45 to 53 percent resolution rate, that gap is large.
- Admin staffing. Tools like Zendesk are described as needing a dedicated support-ops person, a salaried cost that never appears on the plan page.
- Setup and ramp. Painful onboarding and a 1,000-ticket AI cold start are staff time before any value.
- Trust failures. A confidently wrong answer can drive cancellations, as with Cursor's invented policy, or legal liability, as with Air Canada. That is a real, if variable, cost.
Where the hidden money is
The two layers teams most often forget are the seat cost for unresolved tickets and the admin staffing. A low resolution rate quietly means you are paying for the AI and a full human team at once. And a tool that needs a dedicated operator adds a salary that dwarfs the license. Both are real TCO, and both favor an AI-first, zero-admin platform that resolves most tickets and needs no dedicated operator.
One caution on any cost-per-resolution math: a "resolution" is only worth paying for if it actually solved the customer. Resolve reports verified resolution (confirmed by survey, confirmation, or no 72-hour recontact) separately from deflection, and treats reopen rate as the lie detector. A per-resolution bill that counts deflections, where the customer simply gave up, charges you for tickets that were never really solved. Grounding the price in verified resolution is the only honest way to compare cost.
How Resolve lowers TCO
How Resolve prices instead
Resolve decouples price from resolutions entirely. You pay a flat monthly rate for a band of inbound contact volume, with unlimited seats and unlimited AI resolutions inside that band. The published tiers are Launch at $250 a month for roughly 750 inbound contacts, Growth at $900 for roughly 3,000, and Scale at $2,500 for roughly 10,000, with a custom flat Enterprise tier above that. If you go over a band, you bump to the next band, never a per-unit surprise.
The promise is one sentence: your bill never goes up because the AI did its job. It only changes when your business grows. That is the opposite of a per-resolution meter, where the better the automation performs, the more you owe.
Resolve compresses every layer. Flat pricing removes the per-resolution meter and per-seat growth tax. A design intent of 80 to 90 percent auto-resolution shrinks the seat cost for the gap. Zero-admin operation removes the dedicated-operator salary. Day-one resolution from existing docs removes the cold-start ramp. And the ground-or-refuse trust stack, with a supervisor check and business logic computed in code, is built to prevent the confidently-wrong failures that carry the largest tail-risk cost.
Frequently asked
What is total cost of ownership for support AI?
The license plus per-resolution fees, seats for unresolved tickets, admin staffing, setup and ramp time, and the tail-risk cost of trust failures. The sticker price is usually the smallest layer. Compare tools on the full stack, not the plan page.
What is the most-forgotten cost?
The seat cost for tickets the AI does not resolve, and the admin staffing to keep the tool configured. At a real 45 to 53 percent resolution rate you pay for the AI and a full human team at once, and a tool that needs a dedicated operator adds a salary the license never shows.
How does Resolve reduce total cost of ownership?
Flat pricing removes the meter and the per-seat tax, high auto-resolution shrinks the seat gap, zero-admin removes the operator salary, day-one resolution removes the ramp, and the trust stack prevents the confidently-wrong failures that carry the biggest hidden cost.