A/B Testing

Split testing, A/B testing, Split testing, AB testing, Split-run testing, Variance testing
A/B testing is a method where you compare two variants of a web page, ad or e-mail to measure which performs better. You use this to increase conversion.

What is A/B Testing?

A/B testing is een methode waarbij je twee versies van een webpagina, advertentie of e-mail tegelijk test bij vergelijkbare doelgroepen. Je meet welke variant beter scoort op een vooraf gekozen doel, zoals conversie, klikken of aanmeldingen. Zo baseer je keuzes op gedrag van bezoekers in plaats van op aannames.

How A/B testing works in practice

You start with an existing page or campaign, the so-called control version. You then create one version with one change: a different button text, a different image, a modified headline or a shorter text. Both variants are randomly shown to visitors. An analytics tool like Google Analytics or a specialised testing platform like Optimizely or VWO tracks which variant performs best. After a statistically reliable period, you choose the winner. That one becomes the new standard, after which you can launch another test. Importantly, you only change one element per test, otherwise you won't know which tweak made the difference.

Why A/B testing came into being and why it still matters today

A/B testing originates from the medical science and direct marketing world, where control groups and test variants have been standard for decades. With the rise of web analytics and digital marketing, it became possible to apply this principle real-time and on a large scale. In practice, we often see with SME clients that assumptions about customer behaviour are wrong. A green button performs better than a red one, or not at all. A shorter text sometimes converts worse than a longer one. A/B testing reveals what works, regardless of opinion or experience. At a time when every marketing euro must be measurable, it is a powerful tool.

What A/B testing delivers for SME companies

For an online shop with 5,000 visitors per month, a 2% improvement in conversion can immediately yield 100 extra orders per month. For a service provider, a better lead form text can make the difference between 10 and 15 enquiries per week. A/B testing helps you improve your website' s conversion rate step by step, without investing heavily in a full redesign. You test what works, implement the winner and test further. This also makes A/B testing accessible for smaller companies: you can start small with one element on one page. In addition, you build up insight into how your target audience reacts, which also helps with conversion design and future campaigns.

Applications of A/B Testing

You can use A/B testing at any point where visitors need to take an action. In practice, SMEs mainly see testing on product pages, landing pages, e-mail campaigns and advertisements. The power lies in systematically improving small parts that together make a big difference in turnover or leads.

Increase conversion on product pages and landing pages

An online shop with 200 products can start by testing the call-to-action button on the best-selling product page. Does "Order" perform better than "Add to cart"? Does an orange button work better than a blue one? By testing one element, you directly measure the effect on sales. A B2B service provider can test on a landing page whether a short or long text generates more leads, or whether a video works better than an image. With SME clients, we often see that small adjustments, such as moving a trust symbol or tweaking a header, yield 10-20% conversion difference. That's measurable ROI without an extra ad budget.

Optimising e-mail campaigns for higher open and click rates

A newsletter with 2,000 recipients is big enough for a reliable A/B test. You can compare subject lines: does a question work better than a statement? Does first-name personalisation perform better than a generic opening? You can also test within the email: a button at the top versus at the bottom, a short versus long intro, or an image versus text. Platforms like ActiveCampaign and Mailchimp offer built-in A/B testing features. Our trajectories show that SMEs with email marketing often achieve 15 to 30% higher click rates by systematically testing and implementing the winners.

Testing ads for lower cost per conversion

Google Ads and Meta Ads offer native A/B testing capabilities. You can test two ad texts, two images or two audiences against each other. For example, an installation company can test whether "Free quote within 24 hours" works better than "Craftsmanship since 1998". An online shop can compare two product images: lifestyle photo versus white background. By eliminating the loser and shifting budget to the winner, you reduce the cost per click and per conversion. Combined with a well-thought-out SEO strategy and ad optimisation, you build an efficient marketing mix.

When A/B testing is the right choice and when it is not

A/B testing works best when you have enough traffic to measure a reliable result within a reasonable amount of time. For a page with less than 100 visitors a week, testing often takes too long. Also, if you want to make radical changes, such as a complete redesign, A/B testing is less suitable: you are better off testing with user research or prototypes. A/B testing is ideal for incremental improvements on existing pages or campaigns with measurable traffic. In doubt about whether you have enough data? Then start with one test on your best visited page and see if you see a clear difference within two to four weeks.

Want to apply this in your company? Monkey Vision helps SME entrepreneurs with web design, SEO and smart digital solutions. Plan a no-obligation discovery call and discover what is possible for you.

Plan a discovery call

Frequently asked questions

No, A/B testing compares two complete variants with one difference, while multivariate testing tests multiple elements at once in different combinations. In A/B testing, for example, you only change the button text and measure which one works better. In multivariate testing, you simultaneously test button text, colour and position in all possible combinations. Multivariate testing gives more insight into interactions between elements, but requires a lot more traffic to get reliable results. For SMEs with limited traffic, A/B testing is therefore often the practical choice: you test one thing at a time and build improvements step by step.

It depends on your technical knowledge and the complexity of your website. For simple tests on landing pages or emails, you can use tools like Google Optimize, Optimizely or built-in features of your email platform to do it yourself. You don't need a developer to customise a button text or image. More complex tests, such as testing checkout flows in an online shop or setting up server-side testing, require specialist knowledge. With SME customers, we often see that the first tests are done ourselves to get a feel for them, after which a web developer or marketing specialist is called in for structural optimisation. It is important to interpret the test results properly and not draw conclusions too early.

The biggest mistake is stopping a test too early. Statistical reliability requires sufficient data, often at least 100 conversions per variant. Stopping after a few days because one variant is ahead leads to wrong conclusions. A second mistake is adjusting too many elements at once: then you don't know which part made the difference. A third pitfall is testing without a clear goal. "We are testing the homepage" is too vague; "we are testing whether a shorter headline generates more sign-ups" is concrete. Our projects also show that companies often don't implement winners: you test, see an improvement, but don't implement the winner structurally. A/B testing only works if you actually use the results.

The best approach depends on your current situation and goals. Do you have an online shop or website with sufficient traffic but don't know where to start? Schedule a free 30-minute conversion scan at Monkey Vision. We walk through your site live, identify the three pages or elements with the highest test potential and give an honest estimate of what A/B testing can bring to your company. You will immediately receive a concrete test plan that you can pick up this month, including tool advice and prioritisation. No sales pitch, just practical steps that match your SEO and conversion goals. This is how you build measurable growth step by step.

About the author

Monkey Vision

Monkey Vision is a full-service digital agency based in London, specialising in web design agency, SEO and AI automation for SMEs. The knowledge base is compiled by our team of online strategists and continuously updated based on current insights.

Publication date: 26-04-2026
Last update: 26-04-2026