What a good conversion rate optimisation agency actually does, the questions that expose a weak one, and the red flags that should end the conversation.
A good conversion rate optimisation agency does three things: diagnoses why visitors do not convert using real data, turns those findings into testable hypotheses, and runs experiments with enough statistical discipline that the results mean something. Everything else, the redesign opinions, the best-practice checklists, the tools, is decoration. Here is how to tell the agencies that do those three things from the ones that sell them.
A declaration of interest before we start: Clear Click runs conversion rate optimisation engagements, so we are not neutral. What we can offer is the view from inside the industry, including the parts that do not flatter it.
What should a CRO agency actually do?
The work divides into three phases, and a credible agency can describe its process across all three without reaching for a slide.
- Diagnosis. Analytics analysis to find where the funnel leaks, plus qualitative research, session recordings, surveys and user interviews, to understand why. The combination matters: numbers without the why produce guesswork, and opinions without numbers produce redesigns.
- Hypothesis. Findings become specific, testable propositions with an expected effect and a reason to believe it. If an agency cannot show you what a written hypothesis looks like, it does not have a process, it has taste.
- Experimentation. Controlled tests, run to statistical significance, with results recorded whether they win or lose. Losing tests are not failures, they are paid-for knowledge about your audience that competitors do not have.
Our own version of this sequence is set out in how to run a CRO audit that finds real revenue, and the wider discipline in our CRO framework guide.
The questions that separate real CRO from redesign theatre
How will you handle our traffic levels?
The single most revealing question. A/B testing needs traffic and conversions to reach significance, and plenty of sites do not have enough of either for a fast testing programme. A strong agency says so, and proposes a research-led programme instead: fixing evidenced friction directly, measuring before and after, and reserving formal testing for the pages that can support it. A weak agency proposes a testing roadmap your traffic cannot power, and six months later every result is inconclusive.
How do you decide what to test first?
The answer should combine traffic, conversion value and evidence strength, because a test on a page with no commercial weight is a test wasted. Listen for prioritisation by revenue impact rather than by ease of implementation.
What does a result need to look like before you call it?
You are listening for statistical confidence, minimum run times, and pre-agreed success metrics. Anyone comfortable calling a winner from three days of directional data will fill your roadmap with false positives.
How will this show up in revenue?
Uplift percentages compound impressively on slides and are meaningless without the revenue line. Ask to see how they report: the good ones connect experiments to transactions or pipeline in your analytics and CRM, which is the same measurement discipline behind our tracking and measurement work.
Red flags worth walking away from
- Guaranteed uplift. Nobody can promise a conversion increase before diagnosing the site. It is the CRO equivalent of guaranteed rankings.
- A redesign in CRO clothing. If the proposal is a new design with no research phase, you are buying opinions with an experimentation invoice attached.
- Tool-first pitches. The testing platform matters far less than the thinking. An agency leading with its tool partnership is telling you where its incentives sit.
- Best-practice audits. Generic checklists, trust badges here, urgency there, ignore the fact that conversion barriers are specific to your audience and offer. If the audit could apply to any site, it diagnosed nothing about yours.
- No losing tests in the case studies. Real experimentation programmes lose tests regularly. A highlight reel with a 100 percent win rate is a highlight reel, not a record.
How CRO fits alongside SEO and paid media
CRO multiplies the channels you already pay for. Every improvement to conversion rate lowers the effective cost of every visit from search, paid or email, which is why we treat it as part of one growth system rather than a standalone service. It is also why the diagnostic phase should look at where your traffic comes from and what it was promised before it landed: a page can convert badly because the page is wrong, or because the traffic is. An agency that only looks at the page is seeing half the problem.
Questions we get asked about choosing a CRO agency
What does a CRO agency cost in the UK?
Sensible engagements are scoped against the value of the conversion improvement, because that is the denominator that makes any fee judgeable. Be wary of pricing that scales with the number of tests run: it rewards volume over judgement, and volume is easy to fake with trivial tests.
How long before CRO shows results?
Research-led fixes can show measurable movement within weeks. Testing programmes take longer because significance takes time: months, not days, for a trustworthy read on most sites. An agency promising transformation inside a month has skipped the diagnosis.
Do we have enough traffic for CRO?
Possibly not for formal A/B testing, and a good agency will tell you where the line sits for your site. You almost certainly have enough traffic for research-led CRO, because interviews, recordings and funnel analysis do not need statistical power to find broken things.
Should we do CRO in-house or hire an agency?
In-house works when you have analytical capability, research skills and the seniority to protect the programme from opinion-driven overrides. The honest agency case is capability on demand: you rent a full diagnostic and experimentation skill set without hiring four people.
Where we fit
Since you are reading this on our site: Clear Click builds CRO programmes measured in revenue rather than uplift theatre, senior-led and connected to the analytics and CRM plumbing that makes results provable. The full picture is on our conversion rate optimisation service page. If you are evaluating agencies, come and ask us the hard questions, including the ones above. If we are not the right fit for your traffic levels or your stage, we will say so, which is perhaps the most useful test of all.
