Conversion Rate

An online shop is visited by 10,000 people in one day. 200 of them actually buy something. If you divide 200 by 10,000, you get 2 percent. This exact figure is the conversion rate: the proportion of visitors who do what the operator wants them to do. The word comes from the English “to convert” — a mere visitor becomes a customer. Which action counts is defined by each company itself: a purchase, a newsletter sign-up, a download, or filling out a form.

Why two percent decides millions

For a company on the internet, there are basically two ways to sell more. You bring more visitors to the site, or you turn more of the existing visitors into buyers. The first path costs money, because advertising has to be paid for. The second path costs mainly work on your own website. That’s why the conversion rate is considered a particularly valuable lever.

The leverage effect is enormous. If a shop’s rate rises from 2 to 2.4 percent, that is, relatively speaking, 20 percent more revenue — with exactly the same advertising costs. For a company with 100 million euros in annual revenue, that amounts to an additional 20 million euros. For this reason, conversion rates regularly appear in quarterly reports and analyst calls.

It’s important to distinguish this from the click-through rate. The click-through rate measures how many people click on an ad. The conversion rate measures what happens afterward. An ad can generate many clicks and still be worthless if nobody buys. Only both figures together show whether advertising actually pays off.

Calculation, definitional questions and typical pitfalls

The formula itself is simple: desired actions divided by visitors, times 100. The tricky question is what you put into the numerator and denominator. Do you count every page view or every person? What about someone who drops by three times a day and buys once? Depending on the definition, very different values result for the same shop. Comparisons between companies should therefore be treated with caution.

The rate is usually improved through systematic experimentation. The standard method is called A/B testing: half of the visitors see the old version of the page, the other half see a modified one. After a few days, you compare which group bought more often. This is how you find out whether a green button really works better than a blue one — instead of arguing about it.

A common misconception is that a high conversion rate is always good. Anyone who severely restricts their advertising and only targets people who wanted to buy anyway increases the rate — and ends up selling less. The metric is a signpost, not a goal in itself. It only makes sense in conjunction with absolute revenue and cost per customer.

From shop software to AI recommendations

Anyone who orders something online constantly passes through optimized conversion paths without noticing. The notice “Only 2 left in stock”, the pre-filled shopping cart, the checkout process without a customer account — all of these are measures with a single purpose. Amazon is considered an extreme case: one-click purchasing was invented because every additional step loses buyers.

In the tech industry, this metric is now closely linked to artificial intelligence. Recommendation systems predict which product a particular user is likely to buy and arrange the page accordingly. Language models write product descriptions in multiple variants, which are then tested against each other. The success of such systems is almost always measured by the conversion rate.

In the news, the term is usually encountered in connection with figures from publicly traded companies. When Shopify, Zalando or Booking report on their business, the conversion rate is often the number analysts look at first. It is also used outside of retail: streaming services measure how many free trial users become paying subscribers. The logic remains the same, only the desired action changes.