$0.99 per AI resolution. That is what Intercom charges for Fin, its AI chatbot (Intercom Pricing, 2025). At first glance, it sounds fair. You pay for results. No results, no bill.
The problem is not the first month. The problem is month six.
The Pricing Trap Explained
When you first deploy an AI chatbot, your knowledge base is small. Maybe 10 articles covering the basics. Your AI resolves a modest number of conversations. The bill is low.
Then you do what any good team does. You write more articles. You fill knowledge gaps. You improve your documentation. Your AI gets better. It resolves more questions. More customers get instant answers.
Here is the trap: your bill goes up because your AI works better.
You are financially penalized for improving your support system. The better your knowledge base, the more questions AI answers, the higher your monthly cost. This is the opposite of how software pricing should work.
The Real Numbers: Month by Month
Let's walk through a realistic scenario for a growing SaaS company using per-resolution pricing at $0.99/resolution.
Month 1: You launch with 10 KB articles. AI resolves 200 conversations. Bill: $198.
Month 2: You add 10 more articles. AI resolves 400 conversations. Bill: $396.
Month 3: You fill knowledge gaps from zero-result queries. AI resolves 700 conversations. Bill: $693.
Month 4: Your team gets serious about documentation. AI resolves 1,200 conversations. Bill: $1,188.
Month 5: Your KB is comprehensive. AI resolves 1,800 conversations. Bill: $1,782.
Month 6: Your AI is working great. It resolves 2,000 conversations. Bill: $1,980.
In six months, your monthly AI bill went from $198 to $1,980. Not because you chose a more expensive plan. Not because you added features. Because your AI improved.
Over those six months, you paid a total of $4,437 in per-resolution fees alone.
The same volume on a flat-rate plan at $49/month? $294 total. You save $4,143 in the first six months.
Why $0.99 Sounds Cheap But Is Not
The psychology of per-resolution pricing is clever. $0.99 feels insignificant. Less than a dollar. Barely noticeable.
But support volume is measured in hundreds or thousands, not units. Here is what $0.99 looks like at scale:
| Monthly Resolutions | Cost at $0.99 Each |
|---|---|
| 100 | $99 |
| 500 | $495 |
| 1,000 | $990 |
| 2,000 | $1,980 |
| 5,000 | $4,950 |
| 10,000 | $9,900 |
At 10,000 resolutions per month, you are paying $9,900 for AI alone. That does not include the base platform subscription, your team seats, or any other features.
For context, a flat-rate Scale plan covers 40,000 AI answers per month for $249. That is the same cost as 251 Intercom resolutions.
The Compounding Problem
Per-resolution costs do not just grow because your AI improves. They grow because your business grows.
More customers means more support questions. More support questions means more AI resolutions. Your customer base doubles, and your AI bill doubles with it.
Meanwhile, your flat-rate subscription stays the same. Whether you have 100 customers or 10,000 customers, a flat-rate Starter plan covers 2,500 AI answers for $49/month. The Pro plan covers 10,000 for $99/month.
The Growth Tax
Think of per-resolution pricing as a tax on growth. Every new customer increases your support volume. Every improvement to your documentation increases your resolution rate. Both increase your AI bill.
With flat-rate pricing, growth does not change your AI costs until you exceed your plan's limit. And when you do, the upgrade is modest: $49 to $99, or $99 to $249. Not $990 to $9,900.
What Per-Resolution Vendors Don't Tell You
"Resolution" Is Broadly Defined
A resolution is counted when the AI handles a conversation without human escalation. But this includes:
- Conversations where the customer gave up and left.
- Conversations where the AI gave a wrong answer but the customer did not complain.
- Short exchanges ("what are your hours?" "We're open 9-5") that barely qualify as "support."
You pay $0.99 for each of these. Whether the customer was truly helped or simply stopped asking.
You Cannot Predict Your Bill
Try budgeting for per-resolution pricing. You cannot. A product launch brings extra questions. A bug brings a support spike. A viral social post sends traffic to your chat widget. Each event increases your AI resolutions and your bill.
Flat-rate pricing gives you a number you can put in a spreadsheet and forget about until renewal. Per-resolution pricing gives you a number that changes every month.
There Are No Meaningful Volume Discounts
Some platforms offer small discounts at higher volumes. But "small" is the key word. Going from $0.99 to $0.89 per resolution does not change the fundamental math. At 5,000 resolutions, you save $500/month. You still pay $4,450 instead of $99-$249.
The Flat-Rate Alternative
Flat-rate pricing flips the incentive. As your AI improves, your cost per resolution decreases. Here is the math.
Month 1: 200 resolutions on a $49/month plan. Cost per resolution: $0.245.
Month 6: 2,000 resolutions on the same $49/month plan. Cost per resolution: $0.0245.
Your AI got 10x better. Your bill stayed at $49. Your cost per resolution dropped from $0.245 to $0.0245. You are rewarded for improvement instead of punished.
What Flat-Rate Plans Include
Helpable's plans include more than just AI answers:
- Starter ($49/month): 2,500 AI answers, 2 team members, knowledge base, live chat, widget customization.
- Pro ($99/month): 10,000 AI answers, 10 team members, Smart FAQ Creator, advanced analytics.
- Scale ($249/month): 40,000 AI answers, unlimited team members, Custom Context RAG/PDF uploads, priority support.
Compare this to per-resolution platforms where AI is an add-on cost on top of per-agent platform fees.
When Per-Resolution Actually Makes Sense
To be fair, per-resolution pricing has one valid use case: extremely low volume.
If your AI chatbot resolves fewer than 50 conversations per month, the per-resolution cost is under $50. That is comparable to the cheapest flat-rate plans.
But here is the question: if you only get 50 AI resolutions per month, do you need AI at all? At that volume, a simple FAQ page might handle the job.
The moment you cross 100 resolutions per month (and if your AI is any good, you will cross it quickly), flat-rate pricing wins on pure cost. At 500 resolutions, the savings are significant. At 2,000 resolutions, they are dramatic.
Making the Switch
If you are currently on per-resolution pricing and the bill is growing, switching to flat-rate is straightforward.
- Check your current resolution count. Most platforms show this in your dashboard. Look at the last 3 months to find your average.
- Compare the cost. Multiply your average resolutions by your per-resolution price. Compare that to flat-rate alternatives at the same volume.
- Factor in everything. Include platform fees, team seats, knowledge base, and live chat. Get the full picture.
- Migrate your knowledge base. If your documentation lives in your current platform, export it. Most flat-rate tools let you import content or offer migration help.
- Test before switching. Set up the new tool alongside your current one. Verify the AI quality meets your standards.
Frequently Asked Questions
Is $0.99 per resolution really that expensive?
At low volumes (under 100/month), it is manageable. At 1,000 resolutions, it is $990/month for AI alone. At 5,000 resolutions, it is $4,950/month. The cost scales linearly without volume discounts that change the math meaningfully.
Can I limit my per-resolution spending?
Some platforms offer spending caps. When you hit the cap, AI stops responding. This limits costs but also means your customers stop getting AI help once the budget runs out. It defeats the purpose of having AI available 24/7.
How does flat-rate pricing handle overages?
With flat-rate tools like Helpable, your plan includes a set number of AI answers. If you exceed it, you upgrade to the next tier. Starter (2,500 answers) to Pro (10,000 answers) is $49 to $99. There are no surprise overage charges.
Does the AI quality differ between per-resolution and flat-rate tools?
The pricing model does not determine AI quality. What matters is the underlying language model, the RAG implementation, and your knowledge base quality. A $49/month flat-rate tool with a well-maintained knowledge base can outperform a $0.99/resolution tool with poor documentation.
How fast do AI resolutions typically grow?
Teams that actively maintain their knowledge base typically see a 3-5x increase in AI resolutions over the first 6 months. This is a good thing for support quality. On per-resolution pricing, it means a 3-5x increase in your monthly bill.