Price and willingness to pay
How do I pick a price before launching?
Updated 19 August 2026 · Thresholds read live from the verdict engine
Short answer
Take the price from what the category already charges, pick the plan shape before the number, then put that exact price on a funnel and see whether strangers commit. Split testing two prices needs thousands of visitors per arm to separate them, so at validation budgets you test one price, read the result, and change it only if the result asks you to.
Start from what the category already charges
Your buyer has a price in their head before they reach your paywall, set by whatever they already pay for something similar. Pricing against that anchor is far more reliable than pricing against your costs, which the buyer neither knows nor cares about.
List five competitors, write down every plan and every price, and note which one their paywall preselects. That last detail tells you what the category has learned about which plan people accept.
| Category | Common monthly | Common annual | Notes |
|---|---|---|---|
| Habit and fitness | $9 to $15 | $40 to $80 | Annual is preselected almost everywhere |
| Meditation and sleep | $10 to $15 | $50 to $90 | Long free trials are standard |
| Language learning | $8 to $13 | $60 to $120 | Heavy discounting on annual |
| Personal finance | $8 to $15 | $50 to $100 | Trust matters more than price |
| Creator and productivity tools | $8 to $25 | $70 to $200 | Weekly plans are rare and read as predatory |
| Small business B2B | $25 to $99 per seat | 10 to 20% off monthly | Pilots and annual invoices are normal |
Pick the shape before the number
Shape changes conversion more than the number does. A $59 annual plan and a $4.99 weekly plan cost roughly the same over a year and behave nothing alike.
- Annual, monthly shown as a comparison. The default for consumer subscriptions. Highest cash up front, lowest churn exposure, and it needs the most trust.
- Monthly only. Easier first yes, more churn to manage, and it tells you less at test time because the commitment is small.
- Weekly. Converts well and ages badly. Refund rates and complaints are higher, and store reviewers watch it.
- Free trial then charge. Lifts checkout conversion and moves the real decision to day seven, where 30% to 50% of trials typically convert. Useful once the product exists, distorting before it does, since you are measuring willingness to start a trial.
- One-time purchase. Honest, simple, and it caps your revenue per customer. Works where the job is finite.
Test one price, not five
The instinct is to run three prices at once and pick the winner. The arithmetic says otherwise. Separating a 2% conversion rate from a 3% one at 95% confidence with a reasonable chance of detecting it takes roughly 3,800 visitors per price, so a three-way test is over 11,000 visitors. At $1 a click that is $11,000 to answer a question you could answer another way.
- Put your intended price on the funnel and run it. One price, 300 to 500 visitors, one country.
- Read the paywall stage, not only the final number. If people reach the paywall and leave, price is a live suspect. If they never reach it, price is not your problem yet.
- Change the price only if the paywall is where they die. Then rerun on fresh traffic at the new price.
- Compare revenue per visitor across the two runs. A price that halves conversion and triples revenue per visitor is the better price.
Sequential tests are contaminated by seasonality and by audience fatigue, so they are weaker than a clean split. They are also affordable, which a clean split is not at this stage. Take the weaker method and know that you took it.
Read revenue per visitor, not conversion rate
Conversion rate is the right metric for a build decision and the wrong one for a pricing decision. Multiply the rate by the price to get revenue per visitor, which is what your ad budget is actually buying.
| Price | Conversion | Revenue per 500 visitors | Revenue per visitor |
|---|---|---|---|
| $29 a year | 3.2% | $464 | $0.93 |
| $59 a year | 2.1% | $620 | $1.24 |
| $99 a year | 1.1% | $545 | $1.09 |
Worked example
When the cheaper price was the wrong one
A journalling app tested $29 a year first because it felt safe, and converted 3.2% on 500 visitors. Clicks cost $1.40, so each customer cost about $44 against $29 of revenue in year one.
They reran at $59 on fresh traffic. Conversion fell to 2.1%, still above the build line, and the cost per customer rose to $67 against $59 in year one plus renewals. Revenue per visitor went from $0.93 to $1.24. The lower price looked friendlier and could not pay for its own traffic.
What to do when the price is the problem
- Move the shape before the number. Offering annual with monthly beside it often beats cutting the annual price.
- Add a reason for the price. What it replaces, and what that costs. A price with a comparison attached is judged against the comparison.
- Change the audience. The same price that stalls with students clears easily with freelancers. Price problems are often audience problems.
- Do not race to the bottom. Under about $20 a year, paid acquisition stops working for almost everyone, and you are left needing a free channel you have not built.
The numbers
Whatever price you land on, the funnel carrying it gets judged against the same bands.
- Funnel starts before any verdict
- 30
- Below this, don't build
- 1.0%
- At or above this, build
- 2.0%
- Visitors before a segment counts
- 50
A useful discipline: write your intended price and the conversion rate you would accept for it before you launch. If the price needs 2% to work and you get 1.2%, you have a decision to make rather than a number to reinterpret.
How VerifyToLaunch does this
Your price goes into the project settings and appears on the paywall the funnel generates, in the plan shape you choose. Rerunning at a different price is a settings change and a fresh batch of traffic, and the dashboard keeps the runs separate so you can compare revenue per visitor rather than only conversion.
There is no automatic price split test here, and that is deliberate: at the traffic volumes a validation budget buys, a split test cannot separate two nearby prices, so offering one would be selling you a false answer.
Common questions
- Should I test a price before the app exists?
- Yes, and it should be the price you intend to charge. Testing an unrealistically low price produces a number you cannot use, since the whole point is to learn whether the price you need is a price people accept.
- How many visitors does a price A/B test need?
- Roughly 3,800 per price to separate a 2% rate from a 3% one at 95% confidence with reasonable power. That is why validation budgets test one price at a time and compare sequential runs on revenue per visitor.
- Is a free trial a good idea in a pre-launch test?
- It inflates the number you are measuring, because starting a trial is much easier than paying. Either test the direct charge, or discount the result yourself using a 30% to 50% trial-to-paid rate before comparing it to any benchmark.
- What if nobody buys at any price?
- Then the price was never the variable. Look at where people leave: if they never reach the paywall, the promise or the audience is wrong, and no price fixes that.
Read next
- How do I find out if people will actually pay?Three ways to collect a paying signal before you build, and the CAC maths behind the rate.
- What is a web2app funnel and what converts?Why apps sell on the web first, the screens in order, and the rate each one holds.
- What conversion rate should I expect?Benchmarks by traffic type and ask, plus how to diagnose a bad rate stage by stage.
- How do I validate an app idea?The whole method: pick a commitment that costs the visitor something, buy cold traffic, count who commits.
Put your price in front of strangers
Set the price and the plan shape, generate the funnel that carries them, and read the paywall stage separately from the rest so you know whether price is actually the problem.
One-time licence, 30-day money-back guarantee