CRO
What CRO is and how to start without a data team
A definition of conversion rate optimisation, how to prioritise what to test when traffic is low, and what to do when you lack the volume for an A/B test.
Short answer
CRO (Conversion Rate Optimization) is the work of getting more results from the same traffic by changing a page, a message or a flow. It does not need a data team: at low volumes you progress with qualitative diagnosis and large changes, not with A/B tests of small details.
CRO is usually sold as a discipline of A/B tests and expensive tools. In most businesses, 80% of the value sits before any of that.
What it actually is
CRO is the work of getting the same traffic to produce more results. If 100 visits currently bring 2 enquiries and you get to 3, you have increased opportunities by 50% without spending another peso on advertising.
It is the cheapest lever available, because you are already paying for the traffic.
The mistake of starting with tests
An A/B test is a precision instrument: it decides between two similar options when you have enough volume to tell the difference from noise. If your site gets few visits a month, a test takes months to conclude and will probably conclude "cannot tell".
At low traffic the path is different: large changes, measured before and after, with qualitative diagnosis telling you what to change.
How to prioritise what to test
Before touching anything, answer three questions with data:
Where are people lost? You need the funnel measured step by step. Without it you are guessing.
Why are they lost there? Numbers say where, not why. The why comes from session recordings, one-question exit surveys, or listening to your own sales team.
What is it worth fixing? A stretch few people reach does not deserve the same effort as one everybody passes through.
Changes that usually move the needle
These are not tricks; they are frequent problems:
- Costs revealed late. Shipping, tax, or a "from $X" that turns out to be another number.
- Forms that ask for too much. Every field has a cost. If you will not use the data, remove it.
- The offer is not clear in five seconds. If a stranger cannot say what you sell, no button saves it.
- Too many actions per screen. If everything stands out, nothing does.
- No proof. With no cases, no verifiable data and nothing backing the claim, the visitor has to take you on faith.
- Speed. A page that lags is a page abandoned before it is read.
A realistic working cycle
- Measure the current funnel and leave it running.
- Pick the stretch with the largest absolute loss.
- Find out why with qualitative material.
- Make a change big enough to be noticed.
- Let enough time pass and compare against the previous period.
- Document what you changed and when, so the comparison means something.
Step 6 is the one almost everybody skips, and without it you cannot attribute anything to anything.
What to expect
CRO is not a run of consecutive wins. Many changes move nothing and some make things worse. The value is in accumulating knowledge about your specific audience, not in applying a best-practice checklist.
In short
CRO is making the same traffic yield more. Start by measuring the funnel, find out why people leave where they leave, and make large changes if your volume cannot support tests. Document every change and its date.
Want us to review where your traffic is lost? Tell us about your project.