A/B Testing for SMEs: Decisions Backed by Data, Not Gut Feeling

"I think the red button works better." I hear sentences like this often, and usually they are followed by a change to the website that no one ever checks. The problem: gut feeling deceives. What you like doesn't have to appeal to your customers. This is exactly where A/B testing helps. You show two versions of your page to different visitors and let the numbers decide which one goes down better. In this article I explain, without jargon, how it works and what you as an SME should watch out for.

What is A/B testing, really?

With A/B testing you create two versions of a page or an element. Variant A is usually the original, variant B the change, for example a different button text, a new headline or a shorter form. Half of your visitors see A, the other half B. Neither knows about it. In the end you compare which version brought more enquiries, clicks or sales.

The big advantage: you test with real visitors in the real situation, not in a survey. No one tells you what they would theoretically do, you see what people actually do.

Why this is worthwhile precisely for small companies

Many think A/B testing is only for large online shops with millions of visitors. That isn't quite true. An SME benefits too, because even small improvements make a noticeable difference. If your website brings 30 enquiries a month and a better version turns that into 36, that is 20 percent more, without a single franc of extra advertising.

And it isn't only about money. A/B testing ends endless discussions. Instead of arguing within the team about which headline is better, you run both and look at the result after two weeks. That saves nerves and produces decisions everyone can stand behind.

What lends itself well to testing

You don't have to rebuild the whole page. Often small, clearly delimited changes bring the best insights. Good candidates for a first test are:

• The text on the most important button, "Send enquiry" versus "Get free advice".
• The main headline of the home page, does it name the benefit or just the service?
• The length of the contact form, do too many fields put people off?
• The first image, a photo of you versus a neutral stock image.
• The placement of the contact button, right at the top or only after the explanation?

My advice: always test only one thing at a time. If you change button, image and text at the same time, you won't know in the end what made the difference.

How to evaluate a test properly

A test is only as good as its evaluation. This is where most mistakes happen. Three points are decisive:

• Wait for enough visitors: with ten clicks, a lead is pure chance. Only from a few hundred visitors per variant does the result become reliable. With little traffic a test simply takes longer, that is normal.
• Let it run long enough: run a test for at least one to two full weeks. Behaviour on Monday differs from Sunday, and a single good day distorts the picture.
• Look at the right goal: more clicks on the button are nice, but in the end what counts are the enquiries. Measure what genuinely benefits your business.

If one variant is clearly ahead, make it the new standard version. If both are neck and neck, the change simply wasn't important, and that too is a valuable insight, because it spares you further discussions about it.

Where you can start

You don't need expensive software. For simple tests, Google Analytics combined with a bit of patience is often enough: you run variant A for two weeks, then variant B for two weeks, and compare the enquiries. It isn't perfect, but for a start it is entirely sufficient. Anyone who wants more precision uses a specialised tool that serves both variants at the same time.

Important in Switzerland: as soon as you measure visitor behaviour, you generally need their consent via a cookie banner. A/B testing and clean data protection go together, not one without the other.

The most common pitfalls

So that your first test succeeds, here are the mistakes I see most often:

• Stopping too early: after three days A seems ahead, so the test is ended. That is gambling, not testing.
• Changing too much at once: then the result is there, but not explainable.
• Keeping the loser anyway: anyone who looks at the test and in the end still follows their gut feeling can spare themselves the effort.

A/B testing is no black magic, it is an attitude: instead of believing what works, you check. Even a single clean test per quarter takes you further than any assumption. If you'd like to know which version of your website brings more customers, I'm happy to help you set it up and evaluate it. Write to me via the contact form and we'll look at your page together. You'll find more practical tips for SMEs in my other articles.