CRO
Fix the leak before buying more water.
Conversion rate optimisation is finding where people who were interested enough to visit stop short of contacting you, and fixing that. It is almost always cheaper than buying more visitors and it is almost always done second.
What is conversion rate optimization and where do you start?
It is the practice of increasing the share of visitors who take the action you want, usually calling or submitting a form, without increasing traffic. You start by measuring where people currently stop, which requires working tracking, then by fixing the largest drop first.
For most local business sites the largest drops are in three places: the first screen, where a visitor decides whether they are in the right place; the form, where every unnecessary field costs submissions; and the mobile experience, where load time and tap targets do more damage than anything on desktop.
Most advice on this subject assumes traffic volumes local businesses do not have. With a few hundred visits a month, A/B testing cannot produce a valid result, so the work becomes evidence based judgement rather than statistics, and anyone claiming otherwise is selling certainty they do not have.
Why this is the first thing to fix
If your site converts two visitors in a hundred and you double the traffic, you have doubled your cost to get the same proportion. If you take it to three in a hundred, every channel you run becomes fifty percent more productive at no additional media cost.
That compounding is why conversion work should come before a budget increase rather than after one. It improves search traffic, paid traffic, social traffic and direct traffic simultaneously.
It is done second because it is unglamorous and because it requires admitting that the problem might be your website rather than your marketing. In our experience that admission is worth money.
The other reason it gets skipped is that it needs measurement. You cannot fix a leak you cannot locate, which is why tracking is a prerequisite rather than an extra.
Where people actually stop
Every visit that does not become an enquiry stopped somewhere specific. The job is finding where.
For local businesses the losses cluster early. A large share of visitors leave from the first screen, which means the page failed to establish that it was the right place quickly enough.
The second cluster is at the point of action: a form too long, a phone number that is not tappable, a booking system that asks for an account, or a chat widget covering the button.
Identifying which of those is happening takes an afternoon with analytics and session recordings, and it usually surprises the business owner.
How the work actually runs
The discipline is resisting the urge to start changing things before knowing where the loss is.
Make sure the measurement is real
Calls tracked, forms tracked on successful submission rather than on click, chats counted, and all of it reconciling with the enquiries you actually received. Optimising against numbers that are wrong is worse than optimising against no numbers.
Find the biggest drop, by device
Conversion rate by page and by device. Mobile and desktop frequently differ enough that a combined figure hides the entire problem, and on most local business sites the mobile figure is the one that needs work.
Watch real sessions
Session recordings of people who did not convert are the most useful qualitative evidence available, and they are cheap. You see the person scroll past the thing you thought was obvious, tap something that is not a link, and leave.
- Recordings filtered to non-converting mobile sessions
- Form analytics showing which field people abandon at
- Search terms used in site search, which reveal what people cannot find
See the remaining steps: How the work actually runsHide the remaining steps: How the work actually runs
Ask the people who answer the phone
The most underused source of conversion evidence in any business. Whoever answers enquiries knows exactly which question everybody asks, which means they know what the website failed to answer.
Change one substantial thing
Headline and offer, form length, the first screen, the mobile call to action. Substantial changes, one at a time, so that when something moves you know what moved it.
Measure over a real period
A month against a comparable month, with similar traffic and similar seasonality. Not a fortnight, and not against last week.
The phone conversation in step four costs nothing and regularly produces the single most valuable change on the list.
The honest truth about A/B testing at your scale
Most published conversion advice comes from businesses with tens of thousands of visitors a month. At that volume, split testing produces reliable answers in a fortnight. At four hundred visits a month it does not, and running the test anyway produces a confident number that is noise.
The arithmetic is unforgiving. Detecting a realistic improvement, say from three percent to four, requires a few thousand visitors per variation before the result means anything. A site with four hundred monthly visitors would need well over a year, by which time seasonality and market changes have invalidated the comparison.
Read the full breakdown: The honest truth about A/B testing at your scaleHide the full breakdown: The honest truth about A/B testing at your scale
So we say which basis a recommendation rests on. Some changes are supported by established usability evidence that does not need retesting on your site, such as making a phone number tappable or removing unnecessary form fields. Some are supported by your own qualitative evidence, such as session recordings or what your team hears on the phone. And some are judgement, which we label as judgement.
Where a site does have the volume, we test properly: one substantial change, a predetermined sample size, and a result read at the end rather than peeked at daily.
What we will not do is show you a two week test with thirty conversions and describe the winner as proven.
The fixes that usually come first
In the order they typically produce results on a local business site.
- Test that your forms actually deliver. Submit one from a phone on mobile data. Silently broken forms are the most common and most expensive failure we find.
- Make the phone number a tappable link in the header on mobile, not an image and not only in the footer.
- Remove every form field that is not needed before a conversation can start.
- Rewrite the first screen so it says what you do, where, and what to do next without scrolling.
- Answer the question people actually ask on the phone, on the page, in plain language. Usually price, availability, insurance or what the process involves.
See the full checklist: The fixes that usually come firstHide the full checklist: The fixes that usually come first
- Fix mobile load time, because a page that takes six seconds to become usable loses people before any of the above matters.
- Add real evidence that real people work there: names, faces, credentials, photographs of the premises.
- Make sure nothing covers the primary action on a small screen, including chat widgets and cookie banners.
The first item is free and takes two minutes, and it is worth doing today regardless of whether you do anything else on this list.
Speed is a conversion lever
Google commissioned Deloitte and 55 to study mobile load time across 37 brand sites and more than 30 million sessions, published in 2020.
SourceMilliseconds Make Millions, Google with Deloitte and 55, 2020, 37 sites and 30M+ sessions
Which is why speed work and conversion work are frequently the same project rather than two. On a slow site, load time is often the largest single conversion problem.
The conversion problem that is not on the website
A meaningful share of what gets diagnosed as a conversion problem happens after the enquiry arrives.
An enquiry that waits four hours for a reply has usually contacted two competitors in the meantime. A missed call with no callback is a lead that reached the point of dialling, which makes it the most qualified lead in the pipeline, discarded. A quote sent and never followed up is the largest pool of recoverable revenue in most local businesses.
This is worth saying plainly because it changes where the money should go. Improving a website from three percent to four percent conversion is valuable. Answering the phone reliably and following up twice instead of once frequently produces more, and costs nothing in media.
We will tell you when the leak is operational rather than digital. It is a shorter engagement and it is the honest answer.
How this connects to the rest
Conversion work sits inside web design and development and depends on conversion tracking being real, which is why the two are usually scoped together.
It overlaps with landing page optimisation, which is the same discipline applied specifically to pages receiving paid traffic, and with website speed optimisation, which is frequently the largest single conversion problem on a slow site.
It improves the return on every other service simultaneously, which is the argument for doing it before increasing a media budget rather than after.
Where the leak turns out to be after the enquiry rather than before it, the work moves to CRM implementation and marketing automation instead.
Start by submitting your own form.
From a phone, on mobile data. If nothing arrives, you have found your conversion problem and it costs nothing to fix. If it does, bring us the numbers and we will find the next one.
Straight answers.
What is a good conversion rate?
There is no universal number and anyone quoting you one without knowing your traffic mix is guessing. A site receiving branded search from people who already chose you will convert far better than one receiving cold discovery traffic, and neither figure means anything about the other.
The useful comparison is your site against itself over time, split by device and by channel.
Can you A/B test our site?
If you have the traffic. Most local business sites do not have enough conversions for a valid test in a sensible period, and running one anyway produces confident conclusions from noise.
Where volume is low we work from usability evidence, session recordings and what your team hears on the phone, and we say which basis each recommendation rests on.
How long before we see results?
Structural fixes such as a broken form or an untappable phone number show immediately. Changes to messaging and layout need a month against a comparable month to read fairly.
Seasonality matters more than people expect, which is why we compare like periods rather than consecutive weeks.
Should we do this before or after more traffic?
Before, in almost every case. Improving conversion makes every existing channel more productive at no extra media cost, and buying more traffic into a leaking site multiplies the loss along with the gain.
The exception is a site with so little traffic that there is nothing to measure, where getting visitors has to come first.
Do we need a heat mapping tool?
Session recordings are more useful than heat maps for local business sites, because the sample is small and watching ten real non-converting sessions tells you more than an aggregate colour overlay.
Both carry a performance cost and, on sites handling health information, a privacy question worth answering before installing one.
Where this comes from.
Primary documentation and published research behind the guidance on this page.
Talk to the team
A short call, a look at how the business currently shows up, and a straight answer on what we would do first.