“Thorough” isn’t a personality trait. It’s the process.
Every developer says they’re thorough. It isn’t checkable and it isn’t a differentiator — it’s an adjective standing where a method should be.
So here’s mine, written down. Seven steps, each with the reason it exists. You can judge it, copy it, or use it to check whether whoever you hire is doing the same.
The method
Seven steps, in order, every time.
Why seven and not three. Most published processes are three steps because three is tidy. Each of these is a place the work goes wrong if it’s skipped, and I’d rather publish the real count than a memorable one.
Step 1
Understand.
Where the traffic comes from, what it was promised before the click, and what it actually meets after. A mismatch here produces symptoms everywhere downstream, which is why it’s first rather than assumed.
Step 2
Inspect.
Your product page, gone through the way a customer goes through it — not the way an owner does. Six specific questions (below). The reason this is manual is that the questions have answers a crawler can’t score.
Step 3
Measure.
What I saw, set against what your funnel data says. Observations without data are opinions; data without observations is a chart nobody can act on. Both or neither.
Step 4
Prioritise.
Everything found, ranked by what it’s worth against what it takes. This is the step that produces three things to do instead of forty, and it’s the one the $500 pays for.
Step 5
Implement.
The changes built — Shopify, mobile and desktop, on a duplicate theme. Same person as steps 1–4. The diagnosis is written against what can be built because the person writing it builds it.
Step 6
QA.
Analytics still firing, tracking intact, the full purchase path walked on a real phone and a real desktop before anything is published. The step that stops a conversion fix becoming an outage.
Step 7
Measure again.
Against the baseline from step 3. If it moved, you’ll see by how much. If it didn’t, that’s a finding and we use it.
The six questions I take to a product page.
A visitor who can’t answer these in the first few seconds leaves, and they almost never tell you which one was missing.
Run these on your own PDP before you talk to me. Most stores answer two or three. Which two is the useful part.
0 of 6 answered
The step most people skip, and why.
Measuring again is the only step that can make the person who did the work look wrong. That’s the whole reason it gets dropped.
It’s also the only one that tells you whether any of the other six mattered. A conversion project without a before-number is a project nobody can evaluate — including me.
I won’t guarantee you a figure. Conversion moves for a dozen reasons at once and anyone promising one has stopped looking at eleven of them. What I’ll do is record where you started and tell you honestly what happened.
Where to go next
If the process makes sense to you, the free read is where it starts. Send your store URL and roughly how many sessions a month.
Get the free read