The conversation almost always starts the same way: "we have traffic but it doesn't sell; I think we need a redesign." And a redesign is the most expensive, slowest answer to a problem that's almost never there. Design explains a small part of why someone doesn't buy. These five things explain the large part.
1. Mobile isn't "the small version"
Most traffic arrives on a phone and most shops are reviewed on a 15-inch laptop. Things you only see from a real handset, not from the browser's simulator:
- The buy button sits below the fold and you have to scroll to find it.
- The photo gallery fills two full screens before the price appears.
- Form fields open the wrong keyboard, and typing a postcode on an alphabetic keyboard is enough to make someone leave.
- A cookie bar covers the primary button and its close button is twenty pixels wide.
None of those four is fixed by redesigning. They're fixed by looking.
2. Shipping cost shows up too late
It's the number one cause of cart abandonment, and it isn't a mystery of human behaviour: it's that you changed the price after the person decided. Someone who talked themselves into €39 and finds €46.90 at the last step isn't comparing prices, they're reacting to the goalposts moving.
The fix isn't free shipping. It's saying the cost earlier: on the product page, with the amount or with the threshold above which it's free.
Expensive shipping said in time converts better than cheap shipping said late.
In the Canary Islands this compounds, because on top of freight there can be import handling. Full story here.
3. Checkout asks for things it doesn't need
Every field is a chance to abandon. Questions that usually don't belong: creating an account before buying, a second surname, a company name when 95% are consumers, a phone number nobody will call, confirming the email by typing it twice.
And one that does belong but is badly placed: the coupon. A large, visible discount field tells someone there's a better price than theirs, and off they go to another tab to find it. Many don't come back.
4. The product page doesn't answer what's blocking the purchase
A product page isn't a catalogue entry: it's a conversation with someone who has three doubts and no way to ask them. The three are almost always: will it fit or work, when does it arrive, and what happens if I get it wrong.
When the size equivalence, the measurement in centimetres, the lead time in days or the return policy in one understandable sentence are missing, the person doesn't email support: they close the tab. The best source for what's missing is the email you already get.
If three people asked the same thing this month, that goes on the page today.
5. You're buying the wrong traffic
Sometimes the shop is fine and what's wrong is who's arriving. Campaigns optimised for clicks instead of purchases, audiences that are too broad, or an ad promising something the product page doesn't deliver. It converts badly not because the page fails, but because the person never wanted that product.
You find this by splitting conversion by source. If direct and brand-search traffic convert at 3% and one particular campaign converts at 0.4%, the problem isn't the shop: it's that campaign. And fixing it is a different matter.
The order I look at it in
- Open the shop on a real phone and buy something, paying, start to finish.
- Find where people drop between product page, cart and checkout.
- Split conversion by traffic source.
- Read the last thirty support emails.
- Only then, talk about design.
The first four are free, take an afternoon, and resolve most cases. The redesign that's sometimes needed afterwards then happens knowing what to fix — which is the difference between a new site and a new site that also sells.
If you have traffic and no sales, send me the shop's URL and where you think it's leaking. Fixing the shop before raising the budget is what consulting and growth is for.