What is A/B testing?
A/B testing splits traffic between two versions of a page and compares the results — half see version A, half see version B, and after enough visits the difference becomes distinguishable from chance. A/B testing is a genuinely rigorous method that needs more traffic than most local service businesses have, which makes it the wrong tool for most of them.
The short definition
A/B testing splits traffic between two versions of a page and compares the results.
Half see version A, half see version B, and after enough visits the difference — if there is one — becomes distinguishable from chance.
The traffic problem
That last clause is where it falls down for small businesses.
Detecting a real difference requires enough conversions in each version to rule out coincidence. With a handful of enquiries a month, a test would need to run for many months — during which your season changes, your competitors change and your pricing changes, so whatever you measured is no longer about the page.
Running a test anyway and acting on an early result is worse than not testing, because it produces a confident conclusion drawn from noise.
What to do instead at low volume
Change one obvious thing and watch for a month. Not rigorous, and honest about being a judgement call rather than a proof.
Fix the things that are known to matter. Phone number visible on mobile, page speed, message match with ads, fewer form fields, proof near the decision. These are not close calls needing evidence — they are established and worth doing directly.
Ask customers. Five phone calls asking what nearly stopped them booking produces better insight than a badly powered test.
Watch the whole funnel. Enquiries and booked jobs by month, by source. Slower and more meaningful.
When A/B testing is genuinely right
When a page receives enough traffic and enough conversions for a result to be trustworthy, and when the decision is genuinely uncertain.
For a Houston service business that is usually a paid landing page with sustained spend, not the homepage.
Testing mistakes to avoid
Stopping early. Results swing wildly at the start; the first version ahead is frequently not the winner.
Testing several things at once. Then you know something changed and not what.
Testing trivia. Button colours rarely matter. What the page says, and whether it matches what the visitor was promised, matters enormously.
Ignoring seasonality. Comparing a June week to a July week in Houston compares weather, not pages.
The honest summary
A/B testing is a good method applied to the wrong businesses more often than not. If you have low traffic, spend the effort on fixing known problems rather than on measuring small ones.
Where to start
List what you already suspect is wrong with your busiest page. Fix those first. Test later, if volume ever justifies it.
Frequently asked questions
Can a small business run A/B tests?
Usually not usefully. Detecting a real difference needs enough conversions in each version to rule out chance, which at a handful of enquiries a month takes so long that the season and the market change first.
What should I do instead?
Fix the things already known to matter — phone number visible on mobile, page speed, message match with ads, shorter forms, proof near the decision. Then ask five customers what nearly stopped them booking.
When is A/B testing worth doing?
When a page has enough traffic and conversions for a trustworthy result and the decision is genuinely uncertain. For a service business that is usually a paid landing page with sustained spend.
What are the common testing mistakes?
Stopping early when results are still swinging, changing several things at once, testing trivia like button colours, and comparing periods with different seasonality.