At the front of The Subscription Playbook I promised readers a "pricing model recommendation tool: find out which pricing model is the best fit for your business." This is it.
It's a lot more useful in 2026 than it would have been in 2022. Back then most businesses chose between a flat fee and a few tiers. Now AI has made credits, usage meters and outcome pricing everyday choices, and the wrong pick can quietly eat your margin.
How to use the pricing model picker
Answer ten questions about what you sell, what happens to your costs as customers use more, how value shows up, who buys, how much usage varies, how predictable the bill must be, how people buy, whether you can measure outcomes, how fast customers get value, and how big your market is.
The tool starts with an example already filled in (an AI writing tool for small businesses), so you can see a result straight away. Change any answer and the recommendation updates.
You get three things:
- A recommended pricing model from seven: flat-rate, tiered, per seat, usage-based, credits, outcome-based and hybrid. You also get the reasons, real 2026 examples and the main trap to avoid.
- A runner-up. If it's close, the answer is often a combination.
- How customers should start: freemium, a free trial or a reverse trial.
How the scoring works
Each answer adds or removes points from each model. For example, "my costs rise a lot with every use" adds points to usage, credits, hybrid and outcome pricing, and takes points away from flat-rate and freemium, because a flat price with high variable costs is how AI companies end up with negative margins.
Model score = Σ points from each answer Confidence = gap between the top two models (4+ strong, 2–3 good, 0–1 close call)
The weights reflect the book's guidance and 2026 practice: match the price to how value grows, protect your margin where costs are variable, and give buyers the predictability they need.
A worked example
The pre-filled AI writing tool has costs that rise with every use, usage that varies 10x between customers, small business buyers who want a mostly predictable bill, and a self-serve sign-up where people get value in minutes.
The tool recommends credits, with hybrid as the runner-up. That's what much of the AI market has done: Lovable sells Pro from $25 a month for 100 credits, and ElevenLabs runs credit plans from $6 to $990 a month. The recommended start is a reverse trial: give new users the paid plan's credits for a short time, then drop them to a small free allowance.
Now change "who buys" to enterprise and "how do customers buy" to negotiated contracts. Hybrid takes the lead: a committed platform fee with usage on top, which is the book's minimum-commitment idea.
How to read the result
Treat the recommendation as a strong starting point, not a verdict. A "close call" usually means you should combine models: a base subscription with an allowance, plus a meter for the thing that costs you money.
The model is the structure. You still need the unit it counts (your value metric) and the price points. And whatever you choose, check the margin on your heaviest customers before you launch.
What to do next
- Choose the unit your price counts with the Value Metric Finder.
- Read the book's 5 subscription pricing models and the full subscription pricing strategy hub.
- Torn between two? See seat-based vs usage-based pricing and usage-based vs subscription pricing.
- Combining models? Start with hybrid pricing.
- Selling AI? Design a margin-safe rate card in the Credit Pack Designer.
Frequently asked questions
What are the main pricing models for a subscription business?
The seven used in this tool cover almost every business in 2026: flat-rate, tiered (good-better-best), per seat, usage-based, credits, outcome-based and hybrid. The book also describes five revenue models (fixed fee subscription, advertising, e-commerce, transaction processing, and data and APIs). Freemium, free trials and reverse trials are ways customers start, layered on top.
What is the difference between a pricing model and a value metric?
The pricing model is the structure, such as tiers, usage or a hybrid. The value metric is the unit that structure counts, such as users, jobs, messages or resolutions. Two companies can both use three tiers but grow very differently if one tiers by users and the other by transactions. Pick both deliberately.
Which pricing model is best for an AI product?
Most AI products need some link between price and usage, because every request costs money. Credits and hybrid models (a subscription with an allowance, then usage) are the most common in 2026. Outcome pricing suits agents with a countable result, such as a resolved support ticket. Flat unlimited plans are the riskiest.
Should I offer freemium or a free trial?
Freemium brings more sign-ups but converts fewer of them, so it suits huge, shareable markets where free users bring in others. Free trials convert better and suit most other businesses. A reverse trial, where everyone starts on the paid plan and drops to free, is a middle path that only about 7% of products use.
Can I change my pricing model later?
Yes, and most successful companies do. Salesforce went through three pricing models for its AI agents in about a year. Changing models for existing customers is a price change, though: model what each customer would pay, give clear notice, and consider moving new customers first and existing ones at renewal.
Built on The Subscription Playbook
This tool turns frameworks from Robert Coorey's book into something you can use in minutes. Robert is the author of The Subscription Playbook and a Director at BPAG Accountants. More about Robert.