Price testing methods that do not upset existing customers

Aug 22
3 min read

Updated: 4 days ago

Introduction

Pricing is the fastest lever available and the one least often tested, for an understandable reason: showing different prices to different people feels risky and can be genuinely unfair.

There are ways around that. The constraint is real and it does not mean pricing has to be set once and guessed at forever.

1. Price testing methods for small businesses avoid split-testing the same product

Large retailers show different prices to different visitors. A small business should not.

Customers talk, compare receipts, and screenshot. Being caught charging two prices for the same thing at the same time costs more trust than any pricing insight is worth.

So the useful methods all separate the test by something legitimate: a new product, a new customer group, a different location, or a different time. Everything below is a variation on that.

2. Test on new products first

The cleanest test is a product with no price history, because nobody has an expectation to violate.

Launch at a higher price than instinct suggests. It is far easier to reduce a price later, or to introduce a lower tier, than to raise one that customers have anchored on.

Every new item is a free pricing experiment, and most businesses waste them by pricing consistently against their existing range.

3. Use new customers as the test group

Raising prices for new customers while existing ones keep their rate is normal, expected and easy to explain.

It also gives a clean read: conversion rate among new enquiries before and after the change, holding everything else constant.

Grandfathering existing customers costs some revenue and buys the ability to test without damaging relationships. For most small businesses that trade is worth making.

4. Separate by location or channel

If you operate in more than one place, or through distinct channels, those are legitimate boundaries.

Different prices by location are understood — costs differ. Different prices between your own site and a delivery platform are standard, since the commission differs.

Keep the reason honest and stated. A price difference customers can explain to themselves does not generate complaints.

5. Test sequentially, and control the season

With no way to split the audience, the alternative is time: run the new price for a defined period and compare with the previous one.

The weakness is that everything else moves too. Guard against it by choosing comparable periods, avoiding holidays and promotions, running at least a month, and comparing against the same period last year as a sanity check.

Sequential tests are weaker evidence than split tests. They are usually the only honest option, and a weak test read carefully beats no test.

6. Test the structure, not only the number

Some of the largest gains come from changing shape rather than price.

Introduce a tier above your current offer. Bundle two items. Change the unit — per person instead of per group, monthly instead of annual. Add a minimum. Move something from included to chargeable.

These often produce more than a straight increase, and they do not require anyone to pay more for exactly what they already bought.

7. Read the result on profit, not volume

The instinct is to watch whether sales fell. That is the wrong number.

A 10% price rise that costs 10% of unit volume usually leaves you with more profit and less work, because your costs fell with the volume. Judge on total gross profit across the period, and on profit per transaction.

Also look at who left. Losing your most price-sensitive, most demanding customers is a different outcome from losing your best ones.

8. Give it long enough, and write down what happened

Immediate reaction is not the result. Some customers buy on a cycle and will not appear for weeks.

Run for at least a full purchase cycle, and expect a short dip as regulars notice. Judge after that settles rather than in week one, which is when most price rises get reversed unnecessarily.

Record the test: what changed, when, the before and after figures, and the conclusion. Pricing decisions get revisited every year, and a written history means the next round starts from evidence instead of from whoever remembers most confidently.

Conclusion

Do not show different prices for the same thing at the same time. Test on new products, new customers, separate locations or channels, or sequentially over comparable periods.

Test structure as well as price, read the outcome on total gross profit rather than unit volume, look at which customers you lost, allow a full purchase cycle before judging, and write the result down.