Clunky websites have not faded, and Dutch online buyers report more of them now than in 2021

A clumsy order flow is easy to file as a teething problem that time and better software will solve. The figures of the past four years do not support that comfort. Each year Eurostat runs a survey on internet use among people aged 16 to 74 in every EU country. In the rounds that cover buying online, it asks those who bought something in the previous three months which problems they ran into. In 2025, with most fieldwork done in the second quarter, 23.7% of Dutch online buyers said a website or app had been too difficult to use or had worked unsatisfactorily. In 2021 the same answer came from 19.0%, and in 2023 from 20.8%. Across the EU the share rose from 9.3% to 11.5% over the same years.

So the share who report it has grown, not shrunk. Four years of work on webshops and checkouts have not brought this answer down.

Read the trend, not only the gap with Europe

The Dutch figure is roughly twice the EU average, and it is tempting to conclude that Dutch webshops are worse. The survey does not support that conclusion. Dutch respondents report more problems of every kind: 57.9% named at least one problem in 2025, against 35.4% across the EU, and the Dutch share was already 56.2% in 2021. Slower delivery than indicated was mentioned by 35.3% of Dutch buyers and 19.9% in the EU. Whether Dutch shoppers buy more often, expect more or simply answer more critically, the survey cannot say.

The useful reading lies within the Netherlands. The overall share of Dutch buyers with any problem barely moved between 2021 and 2025, while the share pointing at the website itself rose by almost five points. That is one answer gaining weight among Dutch complaints, among shoppers who keep buying online all the same. Treat it as a direction rather than a precise measurement: these are answers people gave in a survey, and the published table gives no margin of error for the change from one round to the next.

What a shopper means by a site that worked unsatisfactorily

The question covers the whole website or app, not only the checkout, and it says nothing about whether the shopper then abandoned the order. A site with a confusing menu can earn this answer as easily as a broken payment page. What the survey measures is a memory: a moment bad enough that the shopper still recalled it when asked.

In my reading, those memories come disproportionately from the steps the shopper cannot skip. A product page that loads slowly is an irritation. An address form that clears every field after one error, a delivery choice that cannot be selected, a payment that returns from the banking app to an error page, or a basket that is empty when the shopper picks the order up again on a laptop, these block the purchase at the moment the shopper has decided to buy. They are also the faults a merchant often does not see. A conversion report counts the order that did not happen. It does not record that the shopper typed the postcode three times first.

The survey answers next to this one point the same way. Among Dutch online buyers, 5.9% said the final costs were higher than indicated, a mismatch the shopper may first meet in the last steps before payment. And 18.3% had difficulty finding information about guarantees and other legal rights, the kind of information a shopper may look for, such as how to return an item, before committing to an order.

Look for faults the shopper notices and your dashboard does not

A funnel shows where shoppers leave. It rarely shows why the ones who stayed were annoyed, and those shoppers are the ones who answer surveys like this one. Three sources inside your own business come closer to what they experienced.

First, errors in the checkout itself. Log every validation message the checkout shows, per field, and every script error on the checkout pages. A field that triggers an error for a large share of shoppers is a field whose label, format or keyboard is wrong. A script error that appears only on one browser or one payment method is a fault that a desktop test from the office never shows.

Second, the moments between systems. Count how often a payment attempt ends without the shopper returning to a confirmation page, and how often a returning shopper starts with an empty basket after building one on another device. Both look like abandonment in a report. To the shopper they look like a site that does not work.

Third, your customer service. Tag every contact that starts with a problem placing an order, such as a payment that seemed to fail, a discount that did not apply or an order confirmation that never arrived. These are the same complaints the survey records, written in the shopper’s own words and tied to a date, a device and a payment method.

Start this week with the third source, because it costs nothing. Pull last quarter’s contacts about placing an order, sort them by checkout step, and take the step with the most contacts. Then go through that step yourself on a phone, as a guest, making one deliberate mistake and switching device halfway through. Fix what goes wrong there first, and compare the number of contacts for that step a quarter later.

Sources: ec.europa.eu, ec.europa.eu

Related reading