# A practical CRO review for a small business website > Start a conversion review by choosing the action that matters to the business. Then examine the visitor’s questions, the evidence supporting your offer and the friction in the next step. Measure completed outcomes as well as clicks, and test changes against a clear hypothesis. By Dom England. Published 2026-11-04. Canonical page: https://doms-lab.com/insights/conversion-focused-website-design A website can look polished and still leave a potential customer wondering whether the business is right for them. They might understand the headline but miss the service area. They might like the work but have no idea what happens after an enquiry. That is where I’d start a conversion rate optimisation review: follow a real customer’s decision from arrival to outcome. The work can include copy, layout, performance and forms. Changing a button colour is one possible experiment, but first we need to understand the decision around it. ## Define the outcome you want Choose a primary outcome that means something operationally. For a service business, that might be a relevant enquiry that the team can follow up. For a shop, it might be a completed purchase. A button click can help diagnose a journey, but it does not establish that the person reached the end. Google Analytics uses key events to represent actions important to a business. Decide which event should count, where it is confirmed and who checks it. For an enquiry form, the success signal should follow confirmed acceptance by the receiving system, with retries considered so one submission is not counted repeatedly. Sources: [Google Analytics: about key events](https://support.google.com/analytics/answer/9267568?hl=en) ## Answer the arrival questions A first-time visitor needs enough context to decide whether to continue. What does the business do? Who is it for? What is available in their location or situation? What evidence supports the offer? The opening should give a useful answer in language a customer would recognise. I’d compare the page with the promise that brought the visitor there. An advert about a specific service should lead to a page that explains that service clearly. If every route lands on a broad company introduction, people have to reconstruct the connection themselves. A focused page can make that connection explicit. ## Make the evidence specific A project story is more useful when it explains the problem, the work and the resulting product. Name the contribution clearly, show genuine material and qualify the project’s stage. A pilot and a live commercial product can both demonstrate skill; they answer different questions for a prospective customer. The same care applies to testimonials, credentials and performance figures. Use information you can substantiate and explain what it refers to. If you have not measured a sales improvement, describe what was built and how it is intended to help. That gives a reader something they can assess. ## Walk through the next step Try the enquiry journey on a phone, including an invalid email address, a dropped connection and a double tap on the submit button. Check that labels remain visible and that errors tell someone how to recover. A reassuring success message needs to reflect what the system actually confirmed. Ask only for information the business needs at this stage. If a prospect cannot describe a budget or a technical specification yet, decide whether those fields belong in the first conversation. Explain what happens next and give a practical alternative, such as a direct email link, when the form is unavailable. ### An example review hypothesis Visitors cannot tell whether a short initial enquiry is enough. Explain the first conversation beside the form, then compare completed, relevant enquiries with the previous experience. ## Choose a way to learn Observed usability sessions can reveal where people misunderstand a page or struggle with a task. GOV.UK’s guidance describes moderated testing as watching participants attempt tasks and using what happens to understand usability problems. For a small site, that can provide useful direction before you attempt a large experiment. An A/B test needs enough observations and a plan for interpreting the result. Agree the main measure, the comparison period and what would make the experiment inconclusive before starting. Keep other changes in view, including campaign traffic and seasonality. A handful of extra enquiries is a reason to investigate, rather than automatic proof of causation. Sources: [GOV.UK: using moderated usability testing](https://www.gov.uk/service-manual/user-research/using-moderated-usability-testing) - Record the current journey and its known measurement gaps. - Write one explanation for the problem and one proposed change. - Watch for effects on enquiry quality, errors and follow-up effort. - Keep a record of what changed and when. - Retain an inconclusive result as learning, rather than reporting an uplift. ## Make the review repeatable The first review should leave the business with a clearer page and a better way to learn. I’d keep a short record of the question, the evidence, the change and the result. That makes the next decision easier and prevents the same debate returning without context. Good CRO connects the visitor’s experience with the work after the website. If enquiries become clearer but the team cannot respond promptly, the next useful improvement may be a handoff or workflow. Follow the actual journey far enough to see where the value is being lost. ## Sources & further reading - [Google Analytics: about key events](https://support.google.com/analytics/answer/9267568?hl=en) - [GOV.UK: using moderated usability testing](https://www.gov.uk/service-manual/user-research/using-moderated-usability-testing) [Discuss your project](https://doms-lab.com/#contact)