There's more than one way to test a price. The straight A/B price test gets the cleanest answer, but it carries the most risk. Other approaches trade some precision for simplicity or safety. Whichever you use, judge the result by gross profit per visitor, and work out the break-even conversion rate before you start. The price test calculator handles both.
1. Straight A/B price test
Half of visitors see the current price, half see the new one, at the same time.
- Pros: cleanest causal answer; controls for seasonality and campaigns.
- Cons: customers may notice different prices; legal rules on price display and personalized pricing vary by market; needs consistent prices across product page, cart, checkout and emails.
- Best for: stores with enough traffic and a clear policy for handling customer questions.
2. Sequential (before-and-after) test
Run price A for a few weeks, then price B for the same length of time.
- Pros: every shopper sees one price at any moment; simple to run.
- Cons: anything else that changes between the periods (season, ads, competitors) gets mixed into the result.
- Make it better: alternate several times (A, B, A, B) in whole-week blocks, and compare with products you didn't change over the same weeks.
3. Product-level test
Raise prices on a random half of similar products (say, half your candle scents) and leave the rest as a control.
- Pros: no shopper sees two prices for the same item.
- Cons: products differ, so you need many comparable products; shoppers can substitute between test and control items.
4. Bundles and quantity breaks
Keep the unit price, test "buy 2, save 10%" or a three-pack against single units.
- Pros: low risk; can lift order value; same price for everyone.
- Cons: tells you about bundle appeal more than the right base price.
5. Shipping thresholds and fees
Test a free-shipping threshold of $50 against $65, or a flat fee against free shipping with a higher product price.
- Pros: shipping costs are a common reason for abandoned carts, so these tests often have large effects.
- Cons: interacts with order value; analyze revenue and profit per visitor, not just conversion.
6. Research before testing
Survey methods such as Van Westendorp's price sensitivity questions, or interviews with recent buyers, can narrow the range worth testing.
- Pros: fast, cheap, no live risk.
- Cons: stated willingness to pay isn't actual behavior; use it to choose test prices, not to set them.
Choosing a strategy
| If you... | Try |
|---|---|
| Have high traffic and a clear customer policy | Straight A/B |
| Want zero visible price differences | Product-level, bundles or shipping tests |
| Have low traffic | Sequential with alternation, or research first |
| Suspect shipping cost is the real problem | Shipping threshold test |
Always do the break-even math
A price rise from $40 to $45 on a product that costs $15 can lose up to 16.7% of conversions before profit per visitor falls. A cut from $50 to $45 on a $20 cost needs 20% more conversions just to break even. Knowing those numbers before the test tells you what result to look for. The full method is in the price A/B testing guide.
Common questions
What is the safest way to test prices?
Testing the offer rather than the headline price: bundles, quantity discounts, shipping thresholds or guarantees. Every shopper sees the same base price, so there's less risk of upsetting customers, and many of the same lessons about willingness to pay come through.
How big a price change should I test?
Big enough to measure, small enough to reverse easily. Changes of 5% to 15% are common. Calculate the break-even conversion first, so you know what result would make the new price worthwhile.