When conversion slows, the list of possible fixes can grow quickly. Rewrite the headline. Shorten the form. Add another trust signal. Redesign the landing page. Each suggestion may have merit, but the team still needs to understand which problem is actually limiting performance.
A conversion rate optimisation audit brings structure to that decision. It combines measurement checks, customer research and journey analysis to identify where suitable demand is being lost, what may explain it, and which changes deserve attention first.
The value sits in the connection between evidence and action. An audit should leave the team with a clearer view of the customer problem and a practical way to investigate or resolve it.
The conversion needs to mean something commercially
Before reviewing the website, it helps to agree what a valuable conversion actually means. The action recorded in an analytics platform may be several steps removed from the outcome the business needs.
For an ecommerce business, purchases matter, but so do margin, returns and repeat behaviour. For a B2B business, a form submission may be only the beginning of a much longer qualification process. For a property business, a completed enquiry and an attended appraisal are different outcomes.
Choose a primary measure and a small set of guardrails. A lead-generation team might prioritise qualified opportunities per eligible visitor, while monitoring enquiry volume and sales handling effort. An ecommerce team might use contribution per visitor, with purchase completion and refund rates alongside it.
That choice changes the audit. A shorter form may look attractive when the goal is submissions. Its value is less obvious if it creates extra work without producing more suitable opportunities. This is the distinction explored in Lead Volume vs Lead Quality: the journey needs to support the commercial outcome beyond the initial enquiry.
Reliable measurement comes before interpretation
A sudden drop-off can look like a customer-experience problem when the underlying issue is how the journey is being recorded. Data integrity matters because the explanation is only as useful as the evidence supporting it.
Test the key actions yourself. Confirm that a successful submission is distinguished from a button click, duplicate events are handled, and transactions or enquiries can be reconciled with the underlying business systems. Record known gaps, including consent-related differences and journeys that move between domains.
GA4 funnel exploration can show completion across defined steps, but its configuration affects who is counted. Open and closed funnels use different entry rules, and the required sequence matters. A reporting definition can therefore create an apparent gap that needs investigation before any design change. Google Analytics funnel documentation.
Locate the problem in a meaningful segment
The sitewide conversion rate is a starting point. Segmentation helps reveal whether the problem is shared or concentrated in a particular audience, device, acquisition source or product category.
Consider a hypothetical business whose mobile conversion rate has fallen. If it also increased mobile prospecting activity, the audience mix may have changed. Comparing the same campaign and landing-page combination over time provides a more useful starting point than treating every mobile visit as equivalent.
Avoid dividing the data into so many groups that each contains too little information to interpret. Use segmentation to sharpen the question, then collect evidence capable of answering it.
Investigate what prevents progress
Analytics helps locate a problem. Customer research helps explain it.
Observe people attempting the task. Review relevant support questions and sales objections. Use session recordings to investigate interaction patterns, while recognising that a recording cannot tell you everything a person intended or understood.
Organise the review around customer decisions:
| Customer decision | What to investigate |
|---|---|
| Is this relevant to me | The relationship between acquisition message, landing page and offer |
| Can I evaluate it | Product or service detail, comparisons, price and constraints |
| Can I trust it | Evidence, terms and clarity about the business |
| Can I complete the task | Forms, checkout, error handling and accessibility |
| What happens afterwards | Confirmation, fulfilment expectations and sales follow-up |
A broad checklist can help maintain coverage. The finding still needs to describe a specific obstacle for a specific customer in a specific journey.
Separate observations from explanations
Imagine that mobile users repeatedly encounter a validation error in an enquiry form. That is an observation. The explanation might be that the field accepts only one telephone-number format. The proposed change might be to accept appropriate alternatives and make the input requirements clear. The hypothesis connects that change to the behaviour the team expects to improve.
Those are three distinct statements. Keeping them separate prevents a plausible explanation from becoming an assumed fact.
A useful finding records the affected audience, evidence, likely mechanism, proposed action, expected outcome and remaining uncertainty. It also identifies an owner and the conditions for evaluating the change.
Prioritisation is where the audit becomes useful
Assess opportunities using commercial reach, evidence strength, expected value, delivery effort and risk. Avoid presenting a subjective score as a precise revenue forecast.
A reproducible form failure deserves a direct fix and verification. An unclear proposition may require message research before a design is chosen. Competing layouts may warrant a controlled experiment when the available traffic and expected effect make that practical.
Estimate opportunity with transparent assumptions. The number of people dropping out is not the number of customers you can recover. Some were browsing, comparing or unsuitable for the offer.
Choose a manageable first set of actions that the team can implement and evaluate. A recommendation without a delivery route is unlikely to change performance. As we explore in The Risk of Generic Optimisation, an abundance of ideas makes prioritisation and implementation more important.
What a useful CRO audit should deliver
The final output should connect findings to execution: a trustworthy baseline, a prioritised set of opportunities, supporting evidence, implementation ownership and a measurement plan.
For agencies and internal teams, that means agreeing who owns research, design, development and analysis before the audit ends. The work becomes more useful when the next decision is clear. If the findings suggest that problems span the whole experience, the next question is whether targeted optimisation or a redesign is the right investment.
At Foster & Funnel, our Growth & Optimisation work connects research, analytics and experience design to the decisions that improve commercial performance. If growth has stalled, talk to us about where to focus next.
