Audits and builds

Buyer page experiments: test a change to your buying pages against a control

Buyer Page Experiments test whether a change to your pricing page, demo form, comparison page or a page template works before you roll it out. Each 8-week cycle runs one test against a control, with its sample size and stop rule set before launch. It costs $6,300 per 8-week cycle.

Prices in US dollars, as published on 9 November 2026.
Item	Price	Terms	How you buy it
One 8-week cycle	$6,300 per 8-week cycle	Agreed in writing before payment, with no automatic enrollment.	Agreed in writing

How you buy it: Agreed in writing before payment, after the 30 minute conversation.
Deadline: An 8-week cycle from recorded scope confirmation: design in week 1, build in week 2, run in weeks 3 to 7 and report in week 8.
Refund and terms: Agreed in writing before payment, with no automatic enrollment.

Talk to us Every offer and price

What happens next: in the 30 minute conversation we confirm the scope, the delivery date and the terms with you. The written agreement then fixes the price shown here, and no payment is taken before it.

What we deliver and when

One test of a buying-flow change, such as a pricing page, a demo form or a comparison page, against a defined outcome such as demo requests per 100 visits, with the sample size and stop rule set before launch.
Or one search split test across a page group, for example 40 product pages split into control and variant, with a stated design, a metric from Search Console fixed in advance and a fixed duration.
A review of reading and use on the tested pages (headings, tables, phone layout and accessibility) and the changes that go into the variant.
A results report with the decision rule applied, including a no-difference result.

Who it is for

A company with enough traffic to test that wants evidence before changing its buying pages, rather than rolling out a change and hoping.

Included

A traffic check before purchase. If the site cannot reach the sample size in five weeks, we say so and do not sell the cycle.
The variant built with publishing access, or handed to your developer.

Not included

Changes to the design, metric or stop rule after launch.
Answer measurement on ChatGPT, Claude, Perplexity, Google AI Overviews and Google AI Mode, which our measurement offers cover.

Method and receipts

Every count we report comes from our published method, version 1.1: the base and the date beside each figure, failed captures listed apart from the valid base, and a receipt kept for every observed answer. Read the method.

Terms

Agreed in writing before payment, with no automatic enrollment.
We do not promise any ranking, placement, traffic or revenue result. What the engines say is recorded on stated dates.

Read our terms.

Questions buyers ask

What if our traffic is too low?

We check before you buy. If the site cannot reach the sample size in five weeks, we say so and recommend Fix Sprint instead.

What if the result is no difference?

That is reported as the result, with the evidence, so you do not roll out a change that did nothing.

Who builds the variant?

We do, with publishing access, or we hand the variant to your developer.

Two ways to start, if this is not the right size

Start free

Claude answers ten buyer questions for your category in about a minute. No email.

Run the free check

Have us do it

We find where you are missing, fix the pages and sources, and measure again.

Talk to us

Fix Sprint, $3,900. Program, $5,500 a month.