Someone who has put a product in the basket has already done the hard part: they found you, chose and made up their mind. If they leave the checkout page, most of the time they have not changed their mind; they have tripped over something. A checkout optimisation project looks for exactly these obstacles, using data, not opinions about the colour of the button.
We work in two stages: first an audit that shows where and why shoppers are lost, then the fixes, in the order in which they are worth making. We do not promise a particular percentage increase, because nobody can know it in advance.
Where shoppers are lost: the funnel and session recordings
The funnel is the sequence of steps between the basket and the placed order: viewing the basket, starting checkout, delivery details, choosing a payment method, confirmation. In Google Analytics 4 we build a report showing how many people move from one step to the next, split by mobile and desktop. The report needs correctly configured events; if they are not in place, we start with those.
The figures tell you where, not why. For the “why” we use session recordings: tools such as Microsoft Clarity or Hotjar replay a visitor’s movements on the page. We switch them on only for visitors who have given consent, and with masking on the fields where personal data is typed. A few dozen sessions show clear patterns: the field people keep going back to, the button pressed several times, the scrolling in search of a cost.
The third source is ourselves: we place test orders on several phones, with every payment and delivery method.
The usual obstacles in the basket and on the checkout page
In Romanian stores the list repeats itself often, whatever is being sold. Shoppers here are used to paying on delivery, to parcel lockers and to having the invoice made out to a company.
Some points are commercial decisions, not programming: the free delivery threshold or whether to accept cash on delivery are yours to decide. We show what can be seen in the data.
- Delivery cost revealed too late: we show it from the basket onwards, together with the free delivery threshold, if there is one.
- Compulsory account: it must be possible to order as a guest; the account can be offered afterwards.
- Needless fields: a second phone number, date of birth, titles. For companies, the details can be filled in starting from the tax identification code.
- Payment methods: no cash on delivery or quick mobile payment, or a payment page that looks different from the store.
- Delivery: no parcel lockers, or a list of towns and villages that is hard to use on a small screen.
- Discount code field too prominent: anyone without a code goes off to look for one and sometimes does not come back.
Mobile errors, the ones that never appear in reports
Problems turn up on a phone that you never see from your desk: the letter keyboard on the phone number field, address autofill that does not work because the right attributes are missing from the code, the order button hidden behind the chat window or the cookie banner.
Error messages are a subject in their own right. “Invalid field” shown at the top while the shopper is at the bottom, on the button, often means a lost order. The message should appear next to the field with the problem, say what is wrong and leave the rest of the data intact. We also check the rare cases: payment declined, session expired, product sold out in the meantime.
Speed matters too: a checkout page weighed down with marketing scripts responds slowly at exactly the moment the shopper has the least patience.
From audit to checkout optimisation: the order of the fixes
The audit ends with a document in which every problem has its evidence (the funnel figure, the recording, the screenshot), an estimate of the effort and a priority. Some are solved in the settings of WooCommerce or of the payment plugin; others need changes to the theme.
After the fixes we track the same funnel steps. With enough traffic, big changes can be verified with an A/B test. Otherwise we compare similar periods and say openly that a good season or a campaign may explain part of the difference. The stages look like this:
- Checking the measurement: the events for basket, delivery, payment and order fire correctly.
- Analysing the data: the funnel by device and by traffic source.
- Observation: session recordings and test orders on real phones.
- The report: proven problems, ranked by impact and effort.
- Fixes and follow-up: one at a time, with the same reports rerun once data has built up.
Frequently asked questions
Does the whole store have to be rebuilt for the checkout to work better?
Most of the time, no. The problems sit in a few pages and are fixed there: the form, how costs are displayed, error messages, payment settings. A rebuild only makes sense when the theme or the platform stands in the way of these fixes.
Are session recordings allowed?
They are tools that process data about visitor behaviour, so we switch them on only after consent has been given in the cookie banner, and with typed data masked. How they are described in the privacy policy and how long the recordings are kept is something you settle with your data protection officer.
How much does the conversion rate go up after an audit like this?
It cannot be said in advance, and we advise you to be wary of anyone who gives a figure. It depends on how many real obstacles there are and on how much of the loss comes down to price or the offer, things the checkout page cannot fix.
This is a typical project description: it shows how we usually approach this kind of work and does not present a project carried out for a particular client. Every real project starts from your company’s situation, and the stages, timescales and price are agreed after the initial discussion.