Price points are easy to adjust. The unit — what you're charging for — requires internal and external overhaul once it's set. Get the unit wrong and no amount of price optimization fixes it.
Here's how to get both right, whether you have six weeks or six days.
The Ideal Scenario
Start at alpha or beta — you have an idea and an MVP. You're putting it in front of customers, getting feedback on how closely it solves their challenge or replaces their current workflow. Once you've validated you're solving a real problem, move into willingness to pay.
Start with 10–15 conversations primarily focused on monetization. Three to five gives you initial signal. By ten you have a trend — and enough to start segmenting buyers.
Begin with 5–7 questions on value and market
Then move into pricing. These three question types give you anchors from three different directions — each revealing something the others miss:
| Question Type | What to ask | What it reveals |
|---|---|---|
| Alternative | If our solution didn't exist, how would you solve this problem?
This is the alternative to position against in your messaging.
|
Your competitive baseline — what buyers fall back on and what they're comparing you to. |
| Relative | What's a software you currently pay for, and what would you rank our value relative to that software as a %?
Example: they pay $100/user for GitHub and rank your value at 50% = $50/user.
|
A concrete price anchor grounded in real spend — not hypothetical willingness. |
| Value-based | Quantify your product's value to business outcomes — revenue up, cost down, or risk reduced — then apply a value-to-price ratio.
Example: alleviates workload of 5 FTEs at $100k each. $500k value at 10× ratio = $50k contract.
|
Your ceiling. What the business outcome is actually worth — and what share of that value you can credibly capture. |
The Imperfect Scenario
Sometimes it's crunch time. You decide to launch new pricing in under 30 days. Before you do anything: you can always launch without pricing. Announce it's coming. Cap usage, comp the cost for 30–60 days up to the limit you set. Then go back and run the ideal scenario.
If that's not an option, get three things right:
Cost
Do you know your costs on a per-customer basis and how they scale? For AI products especially, you want to understand what drives costs up — primarily to set usage limits and avoid surprise bills.
Competition
What does the competition charge for the same offering? Have you decided whether to follow the market or break from it?
Fattmerchant offered a flat fee for transaction processing when every vendor was on a per-transaction basis. That flat fee gave SMBs predictability they couldn't get anywhere else — and that's how they won. The unit was the differentiation. But if you have product differentiation, you don't need pricing differentiation. A new model means enrolling customers in a new way of buying, budgeting, and tracking usage — unnecessary complexity if the product can carry the deal.
Growth
Does your chosen metric grow at the rate you want your business to grow?
I worked with a vendor that offered tabletop exercises on a per-domain basis — a clear way to measure demand and quote customers. But even the largest organizations had a finite number of domains and rarely added new ones. When they priced lower than the competition to win deals, there was no way to grow the accounts they landed. The unit capped the ceiling.
Pick your unit. Set a conservative price. Cap at 30 days comped usage. Launch. You'll know within a month what's working and what to adjust.
Every pricing model has tradeoffs. Your product evolves. Your buyer evolves. Your pricing should too.
The best companies re-evaluate pricing every quarter. Multi-product companies push changes every month. The question isn't whether your model is right — it's whether you're asking often enough.
