Trackum

Conversion rate optimisation: check measurement, then test

If you want to raise your conversion rate, check measurement first. Here is how to find bottlenecks in the funnel, prioritise changes and spot typical brakes on small websites.

Guide · · 4 min read · Written with AI, AI-assisted review on

Conversion rate optimisation: check measurement, then test
📷 Eyestetix Studio · Unsplash

Conversion rate optimisation (CRO) means getting more of your existing visitors to take an action, such as an enquiry or a purchase. The first step is not changing the page but checking the measurement. After that you find the point where most visitors drop off, remove the cause and check whether the change worked.

Why must you check measurement before optimising?

Every optimisation rests on numbers. If the numbers are wrong, you optimise into the void. A form that sends enquiries but is not counted in the tracking looks like a page with no success. The opposite also happens: if a conversion is counted twice when the thank-you page is reloaded, the page looks better than it is.

So check three things before the first intervention. Submit a test form yourself and check that exactly one event arrives. Compare the number of enquiries in Google Analytics 4 (GA4) with the emails or CRM entries that actually came in. And test the page with consent rejected and with consent accepted, because without consent, depending on the setup, data can be missing.

How do you find the bottleneck in the funnel?

A funnel shows the steps on the way to the conversion and how many visitors move on at each step. In GA4 there is the funnel exploration under Explorations for this. According to Google Help, you can define up to 10 steps, set up the funnel as open or closed and break the table down by a dimension, for example by device category.

For a business with a contact form, four steps are often enough: landing page or home page viewed, service page viewed, form started, form submitted. The breakdown by device category shows whether mobile users drop off at a different point from desktop users. If the majority drop off between “form started” and “submitted”, the problem is in the form. If visitors drop off even before the service page, it comes down to the promise, load time or navigation.

A worked example with assumed figures: 1,000 visitors come to the page. At each step you divide the number of users by the number at the previous step and see where the biggest share is lost.

Example funnel with assumed figures
StepUsersShare of previous step
Home page viewed1,000100%
Service page viewed45045%
Form started27060%
Form submitted5420%

In this example, 1,000 visitors give a conversion rate of 5.4 percent (54 divided by 1,000). The biggest relative loss is in the form: four out of five users who start it do not submit it. That is where you start first, for example with fewer fields, and not at the home page.

Form abandonment needs events of its own. GA4’s enhanced measurement can record form interactions, but it does not work reliably with many forms. Check in Google Tag Manager whether the start event is really triggered on first filling in and the submit event only after the success message.

→ Cookie banner with Consent Mode v2: compliant, with no monthly licence

Which typical brakes are there on small companies’ websites?

On small company sites the same causes keep recurring. The following table sorts them by how to recognise them and the obvious countermeasure.

Typical conversion brakes, how to recognise them and what to do
BrakeHow you recognise itCountermeasure
Slow mobile pageMany mobile visitors bounce before doing anythingShrink images, reduce scripts, test on mobile on a mobile network
Form too longMany start the form, few submit itAsk only for fields needed for the reply
Unclear offerHigh bounce rate on the entry pageAlign the headline with the benefit, state the price or the next step
No visible next stepVisitors read but click nothingOne clear button with a call to action, visible without scrolling on mobile
Lack of trustVisitors see the page but do not get in touchReal references, photos of your own work, contact details and legal notice visible
Broken formEnquiries in the tracking fall, emails are missingTest the form yourself, also after plugin updates
Cookie banner in the wayBanner covers the button or formMake the banner smaller, check the position on mobile

How do you prioritise changes?

You will have more ideas than you can implement. Rate each idea by three questions and rank them accordingly.

  • How many visitors does the spot affect? A change on the most visited page works more often than one on a minor page.
  • How big is the expected effect? A broken form weighs more than a different button colour.
  • How big is the effort? Changing a headline takes minutes, a shop rebuild is planned in weeks.

Begin with hurdles that are obvious and need little effort. Only once these are removed are tests worthwhile in which two variants run against each other. Always change one thing and let the test run long enough that chance does not decide.

→ Setting up Google Consent Mode v2 correctly: no warning in Tag Assistant

How do you check whether a change has worked?

Calculate the conversion rate as conversions divided by visitors, separated by source and device. Suppose your page had 12 enquiries from 400 visitors from Google Ads before the change, which is 3 percent. After the change, 400 visitors bring 18 enquiries, which is 4.5 percent. With figures this small, part of the difference may be chance. So wait a few more weeks before you decide. Then see whether the direction stays the same.

A clean comparison matters: the same source, the same period in weeks, no big changes to the campaign at the same time. Anyone who changes page and ads at the same time does not know in the end what worked. Note every change with its date so that you can attribute changes later.

Which numbers conversion optimisation needs

Trackum does not do the optimisation itself, but it does supply the numbers it relies on: at a fixed price, Google Tag Manager, GA4 and Consent Mode v2 are set up, forms and clicks on phone and WhatsApp are measured, and everything is tested before going live. With the subscription (€19 per month) the page is checked daily. Whatever breaks after a plugin update is repaired. Whether your measurement is right today is shown by the free check at trackum.app/check.

Frequently asked questions

What is conversion rate optimisation?

Conversion rate optimisation (CRO) means increasing the share of visitors who carry out a desired action, such as sending an enquiry or buying. It consists of measuring, finding bottlenecks, making changes and checking the results.

How can I increase my conversion rate?

Start by making sure conversions are counted correctly. Then find the point with the most drop-offs and remove the cause there, for example a form that is too long or a slow mobile page. After that, check every change with data.

Which conversion rate is good?

There is no value that applies to everyone. The rate depends on industry, offer, price and where visitors come from. So compare with yourself: the same page and the same source over several weeks.

When is conversion optimisation worthwhile with few visitors?

With few visitors, it mainly pays to remove hurdles you can spot without a test: broken forms, slow pages, unclear offers. Fine-grained tests such as colour variants need more visitors than small websites have.

Sources

  1. Google Analytics Help: Funnel exploration support.google.com
  2. Google Analytics Help: Key events support.google.com
  3. web.dev: Core Web Vitals web.dev
Help