Conversion rate optimisation (CRO)

We first find where people get stuck on the site, using data and session recordings; then we decide by testing changes. If your traffic is not enough for an A/B test, we say so at the start and suggest another route.

What is CRO and how is it done?

Conversion rate optimisation (CRO) means getting more sales or enquiries from the same traffic by finding where a site loses people and testing changes. It has four steps: diagnosis from funnel data and recordings, a prioritised hypothesis backlog, an A/B test with its length set by sample size, and a written result. Low-traffic sites compare before and after instead.

What it covers

How it runs

  1. Measurement and traffic inventory.

    In the first week we look at what data is collected, whether measurement is reliable and whether the pages get enough traffic for an A/B test. The result is shared in writing.

  2. Diagnosis.

    Funnel data, heatmaps and session recordings show where people stop. The findings are listed with the data and the recordings they rest on.

  3. Hypothesis backlog and written decision.

    We agree with you what will be tested and what will be fixed without a test. A button that does not work is not a test subject; it gets fixed.

  4. First test.

    The hypothesis at the top of the list runs in full weeks for the length set by the sample size calculation. No other changes are made to the page while the test runs.

  5. Result and archive.

    Win or lose, the result goes into the archive and the next test is chosen. The status of running tests and the results of finished ones are in the Monday report.

What you hold at the end

  • An assessment of whether your traffic is enough to test
  • Funnel and behaviour analysis findings
  • A prioritised hypothesis backlog
  • A written plan and result for every test
  • A test results archive

When this is not the right thing to buy

  • If your pages do not get enough traffic for a reliable A/B test, do not buy a testing programme; it is better to start with heatmaps, session recordings and fixing the obvious usability problems.
  • If nobody trusts your measurement, the GA4 setup has to be audited first; you cannot test on a setup that counts wrongly.
  • If you expect a promise of a specific rise in conversion rate, we cannot give one: some tests show no difference at all, and we write that down too.

Questions

The questions we hear most about this work. If yours is not here, write to us and we will answer within one working day.

How long does CRO take, and when do results show?

The first week goes on the measurement and traffic inventory; diagnosis and the hypothesis backlog follow. The sample size calculation tells us when the first test will end: if a page’s traffic and current conversion rate are low, the test runs longer. Because obvious faults found in the diagnosis are fixed without a test, some improvement can show early. A testing programme should not be seen as a one-off job, though; its value grows as the archive grows.

What do you need from us for CRO work?

Access to GA4 and Tag Manager, access to the site so that a heatmap and session recording tool can be installed, the system that holds your sales or enquiry records, and a line to a developer who can publish test variations. We also ask you for two decisions: which number you make decisions by, and that no other changes are made to the page while a test runs. The questions customer service hears most often help the diagnosis too.

What determines the fee for a CRO service?

Whether measurement has to be fixed first, the number of pages and funnels to examine, how many tests will run each month, who designs and builds the variations, and whether results will be tracked in a custom dashboard. Testing and session recording tools may have their own subscriptions; these are paid to the provider from your own account. The first meeting is free; 2 to 3 days later we send a proposal that lists each item.

How much traffic do you need for an A/B test?

There is no single number that holds for everyone. Three things set the number of visitors needed: the page’s current conversion rate, the size of the difference you want to detect and how confident you want to be in the result. A low conversion rate, or a search for a small difference, calls for far more visitors. In the inventory week we run this calculation for your pages, and if a result cannot be reached in a reasonable time we say so in writing.

What can be done if our traffic is too low for A/B testing?

We do not set up a testing programme. We study session recordings and heatmaps instead, look at where people give up in form fields and fix the obvious problems: a button that does not work, a field that overflows on mobile, delivery information that appears too late. We track the effect of changes by comparing before and after. This method cannot speak as firmly as an A/B test, because seasons and promotions move the numbers too; we say so in the report.

Why not stop a test early if the result looks good?

Because a difference in the first few days is often chance. Checking a test’s numbers every day and stopping the moment they look good raises the odds of declaring a winner that is not really there. We set the length with a sample size calculation before the test starts, and we keep to it. Running in full weeks has a similar reason: people behave differently on weekdays and at weekends, and half a week distorts that.

How much will CRO increase our conversion rate?

We cannot know in advance and we do not promise a figure. Some tests show a clear difference, some change nothing, and some reveal that the new version is worse. All three are useful: a losing test has stopped that change being rolled out to everyone. The promise we can make is about method: every test starts with a hypothesis, its length is fixed beforehand and its result is written down as it is.

Further reading

From the same service

Tell us about
your project.

Fill in the form or write to us directly, and we will agree on a day to talk.

Your first meeting is with Deniz Sezer, founder of Detmore.

First meeting free · Reply within one working day · Written proposal in 2 to 3 days · NDA ready

What do you need? pick as many as you like

Name, email and phone are required.

Contact

We reply within one working day.