The Hidden Cost of Per-Resolution Pricing: Why Intercom's $0.99 Fee Can Cost You $5,000+ a Month

IN THIS ARTICLE

1. 01What is per-resolution pricing in customer support?
2. 02Why "per resolution" is not the same as "pay for results"
3. 03The math at three different ticket volumes
4. 04Peak-season risk: the ecommerce problem
5. 05What to look for in a flat-pricing alternative

Intercom charges $0.99 per AI resolution. That number sounds manageable — less than a dollar, surely it cannot add up to much. But run the math at real-world ticket volumes and the picture changes fast. A support team handling 5,000 tickets per month, with 50% resolved by AI, pays $2,475 in resolution fees alone. Every month. On top of whatever per-seat subscription they are already paying. At 10,000 monthly tickets, that number doubles to $4,950. Before a single agent salary, before infrastructure, before any other tooling.

This article explains exactly what per-resolution pricing is, why it creates problems beyond cost, and what the alternatives look like for teams who want predictable helpdesk bills.

What is per-resolution pricing in customer support?

Per-resolution pricing charges you each time a support ticket is closed — typically when an AI agent resolves it without requiring human escalation. The idea is that you only pay for value delivered: if the AI resolves the ticket, you pay; if it fails to resolve it and passes to a human, you do not. In theory, this aligns vendor incentives with your outcomes. In practice, it creates a set of problems that per-seat models avoid entirely.

There are three main helpdesk pricing models in use today. Understanding the differences between them is essential for making a fair cost comparison when evaluating tools.

Why "per resolution" is not the same as "pay for results"

The per-resolution model is often marketed as outcome-based pricing: you only pay when a ticket is actually resolved. But the definition of "resolved" is set by the vendor, not by you or your customer. In most implementations, a ticket is considered resolved when the customer does not respond within a defined window — commonly 24 to 72 hours. Whether the customer got a complete answer is not measured. Whether they came back with the same question the following week is not deducted.

This creates a misalignment of incentives that is worth understanding clearly. The vendor profits most when resolutions are declared fastest — not when customers are most satisfied. An AI that gives a confident but incomplete answer and waits for the non-response window to close is, by the vendor's billing definition, a successful resolution. A human agent who spends 20 minutes solving a complex problem thoroughly is, by their billing model, a cost centre with zero revenue contribution.

The math at three different ticket volumes

Let us calculate the actual total cost of a per-resolution model against Delyt's per-seat model at real ticket volumes. We will use Intercom's published pricing ($39/seat base, $0.99/AI resolution) and Delyt Growth ($69/seat/mo, 10 seats included, $690/mo) for a like-for-like 5-agent comparison. All figures assume a 50% AI resolution rate.

Delyt's per-seat advantage compounds at scale. At 1,000 monthly tickets, the difference is $345/month. At 5,000 tickets, it is $2,325/month. At 10,000 monthly tickets, it is $4,800/month — over $57,000 per year. And Delyt's price stays the same regardless of volume. For an ecommerce brand that handles 2,000 tickets in a quiet month but 8,000 in November and December, per-resolution pricing means their support costs spike exactly when their operational costs are already highest.

Peak-season risk: the ecommerce problem

For ecommerce brands, per-resolution pricing carries a specific risk that is worth naming directly. Support ticket volume is not uniform across the year. Black Friday, seasonal sales, and product launches routinely generate 3x to 5x normal ticket volume in a single week. Under a per-resolution model, your helpdesk bill spikes in direct proportion to that volume — at exactly the time your revenue is highest but your margins are under the most pressure.

A brand handling 2,000 tickets per month normally, running a Black Friday campaign that generates 8,000 tickets in one week, pays an unexpected resolution fee bill for that week alone. This is not a theoretical concern: it is the direct consequence of pricing tied to volume with no ceiling.

What to look for in a flat-pricing alternative

Not all per-seat models are equally transparent. Some tools offer a flat monthly fee but cap the number of conversations included, then charge overage fees beyond that cap — which functionally recreates the same problem as per-resolution pricing. When evaluating alternatives, ask these specific questions before signing up.

• Is the monthly price per seat or per account? Per-seat pricing scales with headcount, which can be expensive for growing teams.
• Are AI resolutions included in the base price or billed separately? Get this in writing.
• What happens when you exceed the included conversation volume? Flat overages or per-unit fees?
• How is "resolution" defined for billing purposes? Ask for the exact definition in writing.
• Is WhatsApp included or an add-on? Channel add-ons are a common place where "flat" pricing stops being flat.

Delyt's pricing is a per-seat monthly rate starting at $39/seat/mo that includes unlimited conversations across all channels — email, WhatsApp, live chat, and social — with no per-resolution fees, no overage charges, and WhatsApp included from the entry plan. The bill is the same in January as it is in November.