Conversion Optimisation
You are probably not short of traffic. Most businesses that come to us wanting more visitors already have enough. What they do not have is a website that turns those visitors into enquiries — and buying more traffic to pour into a leaky funnel is the most expensive way to solve a conversion problem.
A measurement audit establishing whether your data can be trusted, and the fixes if it cannot · Funnel data, heatmaps, session recordings, form analytics and a device and browser breakdown · A structured review of the pages that matter against clarity, friction, proof, speed and calls to action · Enquiry form analysis, field by field
On this page
- Is it cheaper to double your traffic or double your conversion rate?
- What does a conversion audit cover?
- Where do enquiries actually get lost?
- Does my website have enough traffic for A/B testing?
- How should you read an agency's uplift claim?
- Why does it matter that the same team builds the website?
- What drives the cost of conversion optimisation?
Is it cheaper to double your traffic or double your conversion rate?
Doubling the conversion rate is almost always cheaper, and it improves the return on every other channel at the same time. Take a business getting 2,000 visitors a month and converting 1% — twenty enquiries. It wants forty, and there are two ways to get there.
Option one: double the traffic. Another 2,000 visitors a month. At a realistic UK cost per click for a competitive service term, that is a substantial ongoing monthly cost, every month, forever.
Option two: double the conversion rate, from 1% to 2%. A one-off piece of work. The same forty enquiries, from traffic already being paid for, and every future visitor is worth twice as much.
The second is also less fun to sell, which is part of why it gets neglected. It is the reason we will tell a business running Google Ads and paid social to fix the page before raising the daily budget.
What does a conversion audit cover?
Eight areas, starting with whether your measurement is telling the truth, because everything after it depends on that. Then where people leave, what the form costs you, how the site behaves on a phone, whether the page is clear, what proof exists, what the call to action asks for, and what happens to the enquiry afterwards.
- Analytics and tracking accuracy — first, always. Are form submissions actually tracked? Are phone clicks recorded? Are conversions imported into Google Ads? Is any of it double-counting? Broken tracking is the norm rather than the exception, and everything downstream is guesswork until it is fixed. We have seen businesses make budget decisions for a year on numbers that were wrong from the start.
- Funnel and drop-off analysis. Where people leave. Which pages they arrive on, which they abandon, where in a multi-step process they give up. The answer is frequently one screen nobody had looked at.
- Form and enquiry friction. The most reliable win available. Every additional field costs you submissions. Most business enquiry forms ask for things nobody needs at first contact — company size, budget, how they heard about you — and every one of those is a reason to close the tab. Three fields and a message box will out-perform eleven fields almost every time.
- Page speed and mobile experience. Speed is a conversion factor before it is a ranking factor. And most B2B sites are still designed on a desktop by people who then check them on a desktop, while a majority of first visits arrive on a phone.
- Messaging and clarity. Whether a visitor understands what you do within five seconds. Whether the page addresses the objection they arrived with. Whether it says anything a competitor could not also say — “quality, reliability, customer focus” is not positioning.
- Proof and trust. Reviews, named testimonials, case studies with numbers, accreditations, a real address, real photographs of real people. The absence of these is the most common reason a good page fails to convert.
- Calls to action. Whether there is one obvious next step, whether it is visible without scrolling, whether it is phrased as something the visitor wants rather than something you want, and whether there is an option for people not ready to commit.
- What happens after the enquiry. The part almost every CRO engagement ignores. An enquiry that is answered in two days instead of two hours converts far less often. If your response process is the bottleneck, no amount of landing page work fixes it — which is why we look at the CRM the enquiry lands in as part of the same piece of work.
Where do enquiries actually get lost?
In six places, and only one of them is the page design everybody argues about. An enquiry is lost when the page does not match what brought the visitor there, when the offer asks too much, when the form asks too much, when there is nothing to trust, when the phone experience is poor, or when the reply is late.
| Stage | What usually goes wrong | The cheapest fix |
|---|---|---|
| The landing page | Someone clicks an advert or a search result promising one specific thing and lands on a general page that mentions it in passing. They came for emergency boiler repair and arrived at a heating services overview. | Send each ad group and search term to a page that repeats its exact promise in the first heading. Usually a wording change rather than a new page. |
| The offer | The only thing on the page is a request for a meeting. Someone comparing suppliers is not ready for a meeting, so they leave and are quietly compared out of the running by whoever asked for less. | Add a smaller step alongside the big one — a question answered, a drawing reviewed, a guide sent — so the people who are not ready yet still tell you who they are. |
| The form | It asks for company size, budget and how they heard about you before it asks anything that would let you reply. Every field is another reason to close the tab, and on a phone it is worse. | Cut it back to what you genuinely need in order to respond, and ask the rest in the reply. Whoever maintains the site can normally do that in a single change. |
| Trust | Nothing on the page proves the business is real. No named testimonials, no address, no photographs of actual people, no work shown — just adjectives any competitor could also have written. | Put the proof beside the decision instead of on a separate page. A named quote and a job you can point at next to the form does more than a testimonials page nobody visits. |
| Mobile | The page was designed on a desktop and checked on a desktop. On a phone the next step sits below several screens of scrolling, the phone number is not tappable, and the page is slow on a mobile connection. | Open it on a real phone on a real connection and use it the way a customer would. Most of what you find is a page weight and layout problem rather than a redesign. |
| After the form is sent | The enquiry arrives in a shared inbox and waits. The buyer sent the same message to two other firms, and whoever answers first is often the one who gets the conversation — nothing on the website changes that. | Route enquiries to a named owner and make the delay visible to yourselves rather than invisible. An enquiry that lands somewhere with an owner is worth more than another round of page changes. |
Does my website have enough traffic for A/B testing?
Most small business websites do not. Reaching statistical significance on a modest improvement needs more conversions per variant than many sites produce in a quarter. Running a test on 200 visitors and declaring a 40% uplift is measuring noise. This is a section that costs us work and is worth including anyway.
So what should you do instead? Fix the things that are known to be wrong. Broken tracking, an eleven-field form, a page that takes six seconds on mobile, no proof, no clear call to action, a phone number that is not clickable. These are not hypotheses requiring a test — they are defects. Fix them, measure the before and after over a decent period, and move on.
Testing becomes worthwhile when you have enough traffic that the remaining questions are genuinely uncertain — usually a few thousand conversions a year rather than a few hundred.
The point
Testing becomes worthwhile when you have enough traffic that the remaining questions are genuinely uncertain — usually a few thousand conversions a year rather than a few hundred.
How should you read an agency's uplift claim?
“We increased conversions by 340%” is a sentence that should prompt three questions: over what period, from what baseline, and how many conversions was that in absolute terms? An increase from 3 to 13 enquiries is 333% and is also within normal monthly variation.
Why does it matter that the same team builds the website?
Because most CRO consultants can tell you what to change, and far fewer can change it. We build the websites, so a recommendation becomes a deployed change in the same sprint rather than a ticket sitting in someone else's backlog waiting for a priority it will never be given.
And because we build CRM systems, we can follow the enquiry past the form — into how quickly it is answered, whether it is followed up, and where in your sales process deals are actually lost. That last part is where the largest gains usually sit, and it is outside the scope of almost every CRO engagement on the market.
It also means the honest answer is available when the answer is unwelcome. If the site's structure is the limiting factor, a redesign or rebuild is the recommendation rather than another round of polishing. If the constraint is the checkout, that is a job for the people who build online stores. And if the traffic itself is wrong, it is the channel and targeting work that needs attention rather than the page.
The point
If the site's structure is the limiting factor, a redesign or rebuild is the recommendation rather than another round of polishing.
What drives the cost of conversion optimisation?
Scope drives it: how many pages and funnels we review, how much measurement has to be rebuilt before it can be trusted, and whether we implement the changes or write specifications for your own developers. For most businesses an audit and a round of fixes is the right start; ongoing work only earns its place where traffic justifies it.
| Service | What it is |
|---|---|
| Conversion audit | Tracking validation, funnel analysis, page review, prioritised recommendations with reasoning, walk-through session. |
| Audit and implementation | The audit plus the changes, scoped after the findings so you are not committing to work before anyone knows what is needed. |
| Landing page build | A single high-intent page built to convert, for a campaign or a core service. |
| Ongoing programme | Continuous work for sites with enough traffic to justify it. Usually delivered inside a marketing retainer. |
How We Work
Measurement audit
Establish whether the data can be trusted. Fix it if not. Nothing else is worth doing first.
Analysis
Funnel data, heatmaps, session recordings, form analytics, device and browser breakdown. Where people actually leave, rather than where we assume.
Review against what is known to work
A structured assessment of the pages that matter, covering clarity, friction, proof, speed and calls to action.
Prioritised recommendations
Ordered by likely impact against effort, with our reasoning shown so you can disagree.
Explore Related Services
Frequently Asked Questions
What is a good conversion rate for my industry?
Be wary of benchmark figures — they average wildly different businesses, traffic sources and definitions of a conversion. A better question is what your rate was last quarter and whether it is moving. That said, if a business services website is converting well under 1% of relevant traffic, something is usually wrong rather than merely unoptimised.
How much traffic do I need before A/B testing is worthwhile?
More than most small sites have. Proper A/B testing needs enough conversions per variant to reach significance, which for many businesses means months per test. Below that threshold, fix known problems and measure before and after instead. We will tell you honestly which situation you are in.
How much does conversion rate optimisation cost?
It depends on how many pages and funnels are in scope, how much of the measurement has to be rebuilt before anything can be trusted, and whether you want the changes implemented or specified for your own developers. For most businesses an audit plus a round of fixes is the right starting point.
How long before we see a difference?
Some changes show immediately — fixing a broken form or a tracking error produces a visible step change within days. Messaging and layout improvements need a few weeks of data to assess fairly. Anything measured over less than a full business cycle is probably noise.
Do you make the changes or just recommend them?
Both are available, and most clients want the changes made. We build websites, so implementation is our ordinary work rather than a hand-off to somebody else's backlog. Where you have your own developers we write the specifications for them instead, in a form they can pick up without a translation layer.
Can you work with our existing website?
Yes, on any platform. Occasionally the honest finding is that the site's structure is the limiting factor and no amount of optimisation will fix it — in which case we will show you the reasoning rather than take a retainer to keep polishing something that cannot work, and talk to you about a rebuild instead.
What if the problem is not the website?
Then we will tell you. Sometimes the traffic is wrong — the wrong keywords, the wrong audience, the wrong ad targeting — and the site converts perfectly well for people who were never going to buy. Sometimes the problem is how long the reply takes. Neither is fixed by redesigning a page.
Is conversion optimisation a one-off or ongoing?
For most small and medium businesses, a one-off audit and round of fixes captures the great majority of the available gain. Ongoing programmes make sense when you have enough traffic for continuous testing, which is a smaller number of businesses than the industry implies.
Let's Build Something That Matters.
Tell us what you're trying to achieve and we'll help you find the right solution.