Become an affiliateStart your program

Measurement and metrics

Conversion rate optimisation (CRO)

Also called CRO, conversion rate optimization, conversion optimisation, konverteringsoptimering.

Short answer

Conversion rate optimisation (CRO) is the systematic work of getting a larger share of visitors to take a desired action – a purchase, a sign-up or a click. The method is data-driven: measure the conversion rate, form a hypothesis, test the change and keep only what demonstrably works.

Definition of Conversion rate optimisation (CRO).

Conversion rate optimisation is about getting more out of the traffic you already have instead of buying more. A web shop with 10,000 visitors and a 2 % conversion rate gets 200 orders; lift the rate to 2.5 % and the same traffic gives 250 orders – 25 % more revenue without spending anything extra on ads. That is why CRO is often the cheapest growth there is: the traffic is already paid for, the work lies in converting it.

The work follows a fixed loop: measure where users drop off, form a hypothesis about why, build a variant and A/B test it against the original. Typical levers are the headline, the call to action, load time, the number of form fields, visible prices and delivery information, and social proof. Only changes that win the test stay – gut feeling does not count as evidence.

In affiliate marketing, both sides work with conversion rate optimisation. The store optimises its own site so that more of the referred clicks end up as sales – this lifts the program's EPC and makes it more attractive to affiliates. The affiliate optimises their own part: link placement, choice of landing page and the credibility of the recommendation itself. The same volume of clicks can earn twice as much commission if both parts convert better.

CRO vs. conversion rate

AspectCROConversion rate
What it isa working method – systematic improvementa metric – the share of visits that convert
Expressed asa process: hypothesis, test, decisiona percentage, e.g. 2.4 %
Answershow do we move the number?where are we right now?

The conversion rate is the thermometer – CRO is the treatment. Measure the rate first, optimise afterwards.

How it goes wrong

  • 01

    Testing without enough data

    At a conversion rate of 2 %, it typically takes more than 10,000 visitors per variant to demonstrate a 20 % improvement with statistical confidence. Stop the test after 300 visitors and you are measuring chance – not effect.

  • 02

    Optimising for the wrong number

    More clicks through to the store is not a goal in itself. Push the click rate up with exaggerated promises and the approval rate drops at the other end, because the traffic does not match the product. Optimise for approved commission, not for clicks.

  • 03

    Changing everything at once

    Change the headline, image and price at the same time and an improvement cannot be traced back to anything specific. One change per test – otherwise you learn nothing from the result.

At Affilyflow

Conversion rate optimisation (CRO) in practice

Conversion rate optimisation needs data about what did not work. Affilyflow logs every tracking event – including the clicks that never turned into a sale – so the conversion rate can be calculated on actual numbers: clicks in, conversions out. Public program pages also only show metrics when the data is above the minimum threshold of 25 clicks and 3 sales, so an impressive EPC is never built on 4 clicks.

Frequently asked questions

What is the difference between CRO and SEO?

SEO brings in more visitors from search engines, while CRO gets a larger share of those visitors to convert. They work together: SEO fills the funnel, CRO makes the funnel less leaky. More visitors to a page that does not convert does not solve the problem.

What is a good conversion rate for a web shop?

There is no single answer – the conversion rate depends on the industry, price level and traffic source, and visitors from a recommendation typically convert better than cold ad traffic. So compare with your own history rather than with other people's averages.

How many visitors does an A/B test need?

Enough that the difference cannot be down to chance. As a rule of thumb, it takes more than 10,000 visitors per variant to demonstrate a 20 % improvement at a conversion rate of around 2 %. Small sites should test bigger, more radical changes – they cannot measure small tweaks.

Can I do conversion rate optimisation as an affiliate without access to the store's site?

Yes. You optimise your own part: link placement, choice of landing page via deep links, the wording of the recommendation and your own page's load time. Measure the effect in EPC – earnings per click – and you capture both your click rate and the store's conversion in one number.

Read on