E-commerce
E-commerce conversion: find the step that loses your orders
Most stores lose orders at predictable places: delivery cost revealed too late, forced account creation, a checkout with more fields than it needs, and no visible answer to what happens if the item is wrong. Measure your funnel first, then fix the largest drop rather than the most interesting one.
Where do stores actually lose people?
Roughly seven in ten carts are abandoned across retail, and the reasons are consistently the same across surveys: unexpected extra cost at checkout, being forced to create an account, a checkout that is too long or too slow, and insufficient confidence about returns and security.
Notice what is not on that list — colour schemes, hero imagery, and the wording of headlines. Those matter, and they matter after the four above are handled.
How do I find my own worst step?
- Set up a funnel report by checkout step in your analytics, or read the built-in one if your platform has it.
- Identify the single largest percentage drop between consecutive steps.
- Complete that step yourself on a real phone over mobile data, slowly, noticing every moment of friction.
- Fix that one thing and measure for two weeks before changing anything else.
The discipline is changing one thing at a time. A redesign that changes twelve things and improves conversion tells you nothing about which of the twelve to do again.
Why does delivery cost matter more than product price?
Because it arrives as a surprise. A buyer who decided at £40 and reaches a total of £52 experiences a change in the deal, not a price they evaluated. The same £52 shown from the start converts far better than £40 that becomes £52.
The fix is a configuration decision rather than a design one: show delivery cost on the product page, or state the free-delivery threshold clearly enough that the total is predictable before anything is added to a basket.
What does a checkout that converts look like?
- Guest checkout available — forcing account creation to buy once is a large and unnecessary barrier
- Only the fields needed to fulfil the order; every optional field removed lifts completion
- A progress indicator, because people abandon when they cannot tell how much is left
- Address autocomplete and correct input types, so a phone shows a numeric keypad for a postcode
- The payment methods your market actually uses — Apple Pay and Google Pay everywhere, plus local options
- Returns policy, contact details and security signals visible at the payment step, not only in the footer
- Errors that say what to fix, next to the field, rather than a red banner at the top
How much does speed affect this?
More at checkout than anywhere else, because intent is highest and patience is lowest. A checkout taking four seconds to move between steps loses people who had already decided to buy — which is the most expensive category of loss there is.
Checkout performance is conversion work, not engineering vanity. Measure it separately from the home page, because the pages carrying the most JavaScript are usually the ones handling money.
What trust signals actually help?
| Signal | Effect | Caveat |
|---|---|---|
| Clear returns policy before checkout | High | It must be findable, not just present in the footer |
| Real contact details including an address | High | A form alone reads as a business that does not want to be reached |
| Genuine customer reviews | High | Only real ones; invented reviews are a policy violation and are detected |
| Delivery date rather than "fast dispatch" | High | Vagueness reads as uncertainty about your own fulfilment |
| Payment method logos | Moderate | Familiar and cheap, but no substitute for the above |
| Trust badges | Low | Widely ignored, and meaningless ones can reduce trust |
What should I test first if I can only change one thing?
Show the total cost including delivery before the checkout begins. It is the most cited abandonment reason, it is usually a settings change rather than a build, and it can be measured within a fortnight on most stores.
If that is already handled, the second is guest checkout. Both are cheap, both are unglamorous, and both consistently outperform the redesign that was being considered instead.
Questions people ask about this
- What is a good e-commerce conversion rate?
- It varies enormously by sector, price point and traffic source, so the useful comparison is against your own previous number rather than an industry figure. A store converting paid traffic at 1% and organic at 4% has a traffic quality question, not a conversion question — which is exactly the sort of thing an averaged benchmark hides.
- Should I run A/B tests on a small store?
- Usually not. Reaching statistical significance needs traffic most small stores do not have, and an underpowered test produces a confident wrong answer. Fix the known problems first — cost transparency, guest checkout, speed, trust signals — and save testing for when volume makes it meaningful.
- Does site speed really change orders?
- At checkout, clearly yes, and the effect compounds with intent: the further someone is through the funnel, the more expensive a delay becomes. The honest caveat is that speed will not rescue a store with unclear delivery costs and no returns policy — it amplifies whatever the rest of the experience is.
- How do I know whether the problem is traffic or conversion?
- Segment by source. If organic converts well and paid does not, the ads are attracting the wrong intent. If nothing converts, the problem is on the site. If everything converts and volume is low, the problem is neither — it is that not enough people are arriving, which is a different project.
Who wrote this
Anas Bin Masud builds e-commerce sites and does technical SEO for businesses in the UK, Canada and Pakistan — fifteen live client sites, six of them stores taking real payments. The examples in these guides come from those builds and from the audit that rebuilt this site, not from a content brief. More about how I work, or read the case studies.