AI Helpdesks Without Per-Resolution Pricing: What They Really Cost

AI Helpdesks Without Per-Resolution Pricing: What They Really Cost

Intercom Fin costs $0.99 every time AI answers a customer question. That pricing model works great for Intercom's shareholders, but it creates a perverse incentive for your support team: the better your AI works, the more you pay.

Run the numbers on a busy month. At 4,200 AI resolutions, Fin alone costs $4,158, before a single seat is paid for. Nothing went wrong in that month. The AI did its job, customers used it instead of emailing, and the invoice grew with every answer. That's not a billing error. That's the business model.

This article is about the money only: how per-resolution billing works, how it grows with volume, the fees that sit around it, and what the same support load costs on flat-rate tools. If you're comparing tools on features, knowledge base quality and migration effort, our guide to Intercom alternatives for AI support covers each option in depth.

How Intercom Charges for AI Support

An Intercom bill has two layers:

• Seats: $29-$132 per seat per month depending on plan. The Advanced plan most growing teams land on is $85 per seat.
• Fin resolutions: $0.99 for each conversation Fin resolves, on every plan.

On top of those sit optional add-ons such as Proactive Support Plus ($99/month) and unlimited Copilot usage for agents ($35 per seat per month).

What Counts as a Resolution

A resolution is counted when Fin answers from your Help Center content and the customer either confirms the answer helped or leaves the conversation without asking for a human. If the customer comes back and asks for a person, it isn't counted. The definition depends on customer behavior you don't control, which makes the monthly count hard to predict and occasionally hard to audit.

Why Intercom Prices It This Way

Per-resolution billing aligns cost with outcomes: if the AI solved a problem, Intercom argues that outcome is worth $0.99. From the vendor's side it also protects revenue. If AI deflects enough tickets, a customer might need fewer seats, so charging for each resolution keeps AI usage generating income. For the customer it creates three problems:

1. Uncertainty: The bill moves month to month with support volume.
2. Disincentive: Every article you add and every flow you improve raises the bill if it works.
3. Misaligned incentives: You want AI to handle more; each extra answer is new revenue for the vendor.

How the Bill Grows With Volume

Resolution fees scale in a straight line with AI success. Here is the AI line alone, before seats or add-ons:

AI resolutions per month	Fin cost per month	Fin cost per year
300	$297	$3,564
500	$495	$5,940
1,000	$990	$11,880
2,000	$1,980	$23,760
4,200	$4,158	$49,896

Seats add a second growth curve. A 3-person team on Advanced pays $255/month in seats; at 5 people it's $425; at 10 it's $850. So a growing company pays more for every hire and more for every answer the AI gives.

Estimate Your Own Resolution Volume

Start with last month's ticket count and the share of tickets that are repeat, tier-1 questions (password resets, billing lookups, "how do I" questions) your docs already answer. Multiply the two to get a realistic resolution count, then multiply by $0.99. For example, 500 requests a month where 60% are answerable from docs is 300 resolutions, or $297/month in Fin fees. Do the same calculation at double your current volume, because that is where the model hurts.

The Pricing Models You'll Run Into

"No per-resolution fees" can mean several different things. Know which one you're buying:

1. Per-resolution: A fee for each AI-resolved conversation, on top of seats. Intercom Fin ($0.99) works this way, and so does Help Scout's AI Answers ($0.75).
2. Per-seat with AI included or as a seat add-on: You pay per agent, and AI is either bundled into higher tiers or sold as a per-agent add-on. Freshdesk, Zendesk, Zoho Desk and HubSpot Service Hub work this way. The AI bill doesn't move with conversation volume, but it moves with headcount.
3. Per-workspace: One price for the team regardless of seat count. Crisp charges $95/month for Essentials with unlimited agents and its AI chatbot included.
4. Conversation tiers: A fixed price up to a conversation limit, then a jump to the next tier. Tidio and Gorgias price this way. It's predictable within a tier, but check where the next tier starts.
5. Flat plan with support included: Support, chatbot and CRM in one plan priced per team size. This is how PipeCrush prices (details below).

Hidden Costs to Check Before You Switch

A tool can avoid per-resolution fees and still surprise you. Look for:

• "Unlimited" with a ceiling: "Unlimited AI up to 1,000 conversations per month" is a cap. Find out what happens when you hit it: an automatic upgrade, an overage charge or a paused bot.
• AI sold per agent: Freshdesk's Freddy AI runs $29-$35 per agent per month on top of base plans, and Zendesk's advanced AI is roughly $50 per agent per month on top of Suite pricing. Five agents turns a modest add-on into a real line item.
• Onboarding fees: HubSpot Service Hub Professional ($90/seat/month) carries a mandatory $1,500 onboarding fee.
• Seat minimums: Some plans require buying more seats than you use.
• Annual contracts: A low monthly figure that is only available on a 12-month commitment isn't a monthly price.
• Integration spend: If the helpdesk doesn't include a CRM, add the CRM subscription and whatever you pay Zapier to keep the two in sync.

Total Cost Comparison: Same Team, Same Volume

Scenario: a 3-person support team whose AI resolves 500 conversations a month, then the same team at 1,000. Figures are list prices billed monthly, before taxes; confirm current rates on each vendor's pricing page before you decide.

Tool	How AI is billed	500 resolutions/mo	1,000 resolutions/mo
Intercom + Fin (3 seats at $29-$85)	$0.99 per resolution	$582-$750	$1,077-$1,245
Help Scout + AI Answers (3 users at $20)	$0.75 per resolution	$435	$810
HubSpot Service Hub Professional (3 seats at $90)	Included in seat price	$270 + $1,500 onboarding	$270 + $1,500 onboarding
Freshdesk Growth + Freddy AI (3 agents)	$29-$35 per agent add-on	$132-$150	$132-$150
Crisp Essentials	Included, per workspace	$95	$95
PipeCrush Growth (up to 5 users)	Included in plan price	$69	$69

Two things stand out. First, only the per-resolution rows change between the two columns; everything else is flat. Second, the gap widens with success: at 2,000 resolutions a month the Intercom AI line alone is $1,980, while every flat-rate row stays where it is.

The cheapest row isn't automatically the right tool. Crisp and Freshdesk are support tools; you'd still pay separately for a CRM and email marketing. HubSpot bundles CRM data but prices each Hub separately. For a line-by-line look at how Zendesk's per-agent pricing compares, see our Zendesk vs PipeCrush pricing breakdown.

When Per-Resolution Pricing Actually Makes Sense

Per-resolution billing isn't always the expensive choice:

• Very low volume: At 10-20 AI resolutions a month, $10-$20 in fees is cheaper than most flat plans.
• Heavily seasonal businesses: If volume spikes for a few weeks a year and is near zero otherwise, paying per resolution keeps the off-season cheap.
• Testing: Pay-as-you-go is reasonable while you find out whether AI support works for your customers at all.

For a B2B company with steady support volume, flat pricing is more predictable and usually cheaper within a few months.

How PipeCrush Prices AI Support

PipeCrush charges a flat monthly price per plan, sized by team, with no per-seat add-on fees and no per-resolution fees. The AI support chatbot, knowledge base and ticketing are in every plan, alongside the CRM, email sequences and AI email marketing.

Plan	Monthly	Annual (per month)	Users	AI chatbots	Voice AI minutes/mo
Solo	$29	$24	1	1	60
Growth	$69	$58	5	5	300
Pro	$119	$99	10	Unlimited	750
Business	$179	$149	20	Unlimited	1,500

What the plan limits and what it doesn't:

• Chatbot answers aren't billed per resolution. The plan limits how many chatbots you run, not how many questions they answer.
• Metered items are hard caps, not overage. Voice AI minutes, phone minutes, SMS, marketing emails and email validations have monthly allowances. You get a warning at 80%, and at 100% that feature pauses until the next period or an upgrade. There is no overage billing, so the invoice can't spike.
• Team growth means a plan change, not seat fees. Going from 5 to 10 users is a move from Growth to Pro, a $50/month difference.
• No contract lock-in. Cancel anytime with no cancellation fee. Annual billing saves about 17%.

The chatbot answers from your knowledge base articles, website pages and uploaded documents, links customers to the source article, and can open a ticket or hand the conversation to a person when it can't help. Because support shares a database with the CRM, the team working a ticket sees the contact's deals and email history in the same place.

We offer a 7-day free trial for qualified businesses, which is enough time to load your top 20 support questions and see how the bot answers them before you pay.

Conclusion

Per-resolution pricing isn't inherently wrong, but it bills you more as your AI gets better, which is the opposite of what you want from automation. Before your next renewal, take your average monthly resolution count, multiply by $0.99, add seats and add-ons, and set that against a flat-rate plan sized for your team. For most teams with steady volume, the flat plan wins within months, and the gap keeps growing as the AI improves.

---

FAQ

Why does Intercom charge per resolution?

Intercom's per-resolution model ties price to outcomes: if the AI solves a customer problem, Intercom treats that outcome as worth $0.99. The side effect is that AI success raises your monthly cost, which makes the bill harder to predict and discourages expanding AI coverage.

What exactly counts as a "resolution" in Intercom Fin?

A resolution is counted when Fin answers from your Help Center content and the customer either confirms the answer helped or leaves without asking for a human. If the customer follows up and requests a person, it isn't counted. Because the count depends on customer behavior, it can be hard to forecast and occasionally leads to questions about what qualified.

Is flat-rate AI as good as per-resolution AI?

Pricing model and answer quality are unrelated. Fin and most flat-rate alternatives use large language models with retrieval from your documentation (RAG) to ground answers. The difference is the billing structure, not the technology, so test each tool on your own top support questions.

How much can I save by switching from per-resolution to flat-rate?

It depends on volume. At 500 resolutions a month, Fin fees are $495; at 2,000 they are $1,980. Multiply your average monthly resolution count by $0.99, add your seat costs and add-ons, and compare that against a flat plan sized for your team. The table above shows the same 3-person team costing $582-$750 a month on Intercom at 500 resolutions against $69-$150 on flat-rate tools.

What's the catch with flat-rate AI pricing?

Some "flat-rate" tools cap conversations and move you to a higher tier or charge overage when you pass the cap. Others sell AI as a per-agent add-on, so the bill grows with headcount instead of volume. Read the fine print for what happens at the limit. On PipeCrush, chatbot answers aren't metered; metered items such as voice AI minutes pause at their cap rather than billing overage.

Does PipeCrush charge extra for more support agents?

No per-seat add-on fees. Each plan includes a set number of users (1, 5, 10 or 20), and a larger team moves to the next plan. Business, the largest plan, covers 20 users for $179/month.