Conversion rate optimisation is usually presented as a testing discipline: hypotheses, variants, statistical significance. For sites with substantial traffic that is correct and productive.
For a business site receiving a few thousand visits a month, it is the wrong frame entirely. You will never accumulate enough data to test your way to an answer, and while you wait the obvious problems go unfixed.
What works at that scale is finding and removing the specific friction that is losing enquiries, which is usually visible to anyone who looks at the site honestly on a phone.
What to do when conversion is already reasonable
Once the obvious friction is gone and conversion has stopped responding to structural fixes, the remaining gains come from a different kind of work and it is worth knowing when you have reached that point.
The signal is that the five common fixes are genuinely done, the path metrics look sane, and someone unfamiliar can complete the main task on a phone without hesitating. At that stage further layout changes produce noise rather than improvement.
What still moves the number is proof and specificity. Adding a named client where you previously had a generic claim. Publishing a price range where you previously had none. Replacing a stock photograph with a real one of your premises or your work. Each of these is a content decision rather than a design one, and each reduces the uncertainty a buyer is trying to resolve.
The other lever is speed of response, which is not on the page at all. For enquiry-led businesses, the proportion of enquiries answered within an hour affects revenue more than most on-site changes, and it is entirely within your control. Measure it before assuming the website is the bottleneck.
Beyond that, the honest advice is usually to stop optimising conversion and go and get more of the right traffic, because you have taken the cheap gains and the next ones are expensive.
Where enquiries actually leak
Measure the path rather than the total. A single conversion rate tells you there is a problem and nothing about where.
The useful breakdown is how many visitors reach a page describing what they want, how many of those reach the contact page or a contact affordance, and how many of those complete an enquiry. Each ratio points somewhere different.
If few visitors reach a service page, the problem is navigation or the homepage failing to say what you do. If many reach a service page and few reach contact, the service page is not convincing or does not ask. If many reach contact and few complete, the form or the contact options are the problem.
Most businesses have never looked at this and assume the problem is at the end, when it is usually in the middle. A service page that describes what you do without addressing price, timeline or proof leaves the visitor unconvinced, and no amount of contact-form optimisation recovers them.
In practice
The five fixes that usually work
These recur across almost every business site we audit, and none require a testing programme.
Why A/B testing rarely helps small sites
To detect a meaningful difference between two variants you need enough conversions in each to distinguish signal from noise, and for a site producing thirty enquiries a month that takes many months per test.
By the time a result arrives, seasonality has shifted, the market has moved and you have tested one button colour. Meanwhile the form still has eleven fields.
So at this scale, act on judgement and observation rather than on testing. Watch someone unfamiliar with your business use the site on a phone and narrate what they are thinking; twenty minutes of that produces more actionable findings than a quarter of testing.
Look at session recordings if you have them, specifically for people who reached the contact page and left. Read the enquiries you do receive for the questions people ask, because every question repeatedly asked is something your site failed to answer.
Testing becomes genuinely useful once you have enough volume that the obvious problems are already fixed. Getting there first is the point.
The twenty-minute exercise
Hand your phone to someone who does not know your business, on mobile data rather than wifi, and ask them to find out what you charge and get in touch. Say nothing while they do it. Almost every business that runs this discovers two or three problems they had stopped being able to see, and it costs nothing.
Key takeaways
- Doubling conversion is usually cheaper than doubling traffic and compounds with everything after.
- Measure the path rather than a single rate; each ratio points at a different problem.
- The leak is usually in the middle: service pages that do not convince, not the form.
- Five fixes recur everywhere: one-tap contact, clear positioning, proof beside claims, price, shorter forms.
- A/B testing needs volume most business sites do not have. Use observation instead.
- Watch an unfamiliar person use your site on a phone. It finds what you can no longer see.
Frequently asked
It varies enormously by industry and traffic source, so benchmarks are mostly unhelpful. The useful comparison is against your own site last quarter. More useful still is the path breakdown: what proportion reach a service page, what proportion of those reach contact, and what proportion of those enquire. That tells you where to act.
Only if you have enough conversions to reach significance in a reasonable period. For a site producing thirty enquiries a month, a single test takes months and by then the context has changed. Below that volume, act on observation and judgement: the obvious problems are usually still unfixed and fixing them beats testing button colours.
Usually the middle of the path rather than the form. Visitors reach a service page that describes what you do without addressing price, timeline or proof, and leave unconvinced. Check what proportion of service page visitors even reach your contact page; if it is very low, optimising the form will not help.
It usually decreases total enquiries and increases qualified ones, which is the trade most businesses want. People who cannot afford you stop enquiring, saving everyone a meeting, and people who assumed you were out of reach discover you are not. The businesses that regret publishing prices are usually those competing purely on being cheapest.