The three numbers that matter
For each price the calculator shows:
- Conversion rate: orders divided by visitors.
- Revenue per visitor (RPV): conversion × units per order × price.
- Gross profit per visitor: conversion × units per order × (price - variable cost per unit).
Gross profit per visitor is the number to decide on. It already balances the two forces in a price test: fewer buyers, more profit from each.
Break-even conversion
You can work this out before the test starts:
Going from $40 to $45 with a $15 cost: the margin goes from $25 to $30, so conversion can fall by up to 1 - 25/30 = 16.7% before the new price earns less. A price cut works the other way. Dropping from $50 to $45 with a $20 cost needs conversion to rise by 20% just to break even.
That second example is why discounts disappoint so often.
Worked example
The default inputs: A sells at $40 to 300 of 10,000 visitors; B sells at $45 to 270 of 10,000 visitors. The cost is $15 per unit.
| A ($40) | B ($45) | |
|---|---|---|
| Conversion | 3.00% | 2.70% |
| Revenue per visitor | $1.20 | $1.215 |
| Gross profit per visitor | $0.75 | $0.81 |
Revenue barely moves (+1.25%), but profit per visitor is up 8%. B's conversion of 2.70% sits comfortably above its break-even of 2.50%.
There's a catch. The conversion difference isn't significant (p = 0.20), so the data is also consistent with B converting at, say, 2.4%, which would lose money. That's normal for price tests on modest traffic. Run longer, or decide whether the downside risk is acceptable.
What to put in "variable cost"
Include everything that scales with each unit sold: product cost, packaging, payment processing fees, pick-and-pack, and the shipping cost you absorb on average. Leave out fixed costs like rent or software. If you offer free shipping above a threshold, a price change can push orders over or under that line, so check that effect separately.
For the bigger picture, read the price A/B testing guide.
Common questions
How do I know if a price increase worked?
Compare gross profit per visitor, not conversion rate. In the example, raising a price from $40 to $45 cut conversion from 3.0% to 2.7%, but profit per visitor rose from $0.75 to $0.81 because each sale earns $5 more.
What is the break-even conversion rate for a price change?
It's the conversion rate the new price needs to earn the same gross profit per visitor as the old one. The formula is old conversion × (old price - unit cost) / (new price - unit cost). Going from $40 to $45 with a $15 unit cost, conversion can fall by up to a sixth (from 3.0% to 2.5%) before you lose money.
Why doesn't the calculator test profit for significance?
Testing revenue or profit properly needs the spread of order values, which you only get from order-level data. The calculator tests the conversion difference, which needs only counts, and reports revenue and profit per visitor as point estimates.
Is it legal to show different prices to different visitors?
Rules vary by country and sometimes by state, and platforms have their own policies. In general, random price tests are treated differently from prices based on personal characteristics, but consumer protection, price-display and anti-discrimination laws can still apply. Check the rules where you sell before running one.