E-commerce conversion optimisation.
Most abandoned carts are not a marketing problem. They are a technical one.
Conversion work is usually sold as persuasion — copy, urgency, trust badges. In my experience a large share of lost sales on established stores are mechanical: the checkout is slow, a payment method fails silently on mobile, the address form rejects a valid postcode, or shipping cost appears so late that people leave. Those are engineering problems with definite answers.
What’s included
- Checkout performance — the page that cannot be cached
- Payment failure diagnosis across methods and devices
- Mobile checkout testing on real devices
- Form friction: address lookup, validation, autofill
- Guest checkout and account creation flow
- Shipping and tax shown early rather than as a late surprise
- Accurate tracking so your data reflects reality
Start by finding out where they leave
Before changing anything, instrument the funnel properly so you can see which step loses people. Very often it is a single step, on a single device class, for a single payment method — and no amount of copy testing would have found it.
Payment failures hide in plain sight
A payment method that silently fails on one mobile browser can cost a meaningful share of revenue without generating a single support ticket, because customers simply leave. Testing every method on real devices and logging failures explicitly is unglamorous and frequently the highest-value work available.
Trust your numbers first
If tracking is misconfigured — double-counted purchases, missing mobile events, ad blockers — every decision made from that data is wrong. Verifying measurement comes before optimising against it.
Common questions
Is this the same as CRO?
Overlapping but not identical. Agencies typically test copy, layout and offers. I fix the technical causes — speed, failures, friction and bad data. They complement each other, and the technical layer usually pays back faster.
How much improvement can we expect?
It depends entirely on what is currently broken. I would rather measure first and tell you what is achievable than quote a number I cannot support.
Can you work with our marketing agency?
Yes, and that is often the best split — they handle persuasion and traffic, I make sure the machinery underneath works.
Do you do A/B testing?
I can implement it. For most stores under a certain traffic level, fixing what is measurably broken beats testing variants you cannot reach significance on.
Have a project in mind?
Tell me what you're building and roughly when you need it. I reply to every genuine enquiry within one working day.