What I check first on a Shopify store

Most stores that come to me ask for a redesign. Almost none of them need one.

A redesign changes everything at once, which means you learn nothing from it. If revenue goes up you do not know why, and if it goes down you cannot roll back the one thing that did it. So before I touch anything, I go through the same five checks. They take about an hour and they usually find the problem.

1. What the ad promised

I open the ad, then I open the page it points to. If the ad says one thing and the page says another, nothing further down matters. The visitor has already decided they are in the wrong place.

This sounds obvious. It is still the single most common thing I find, because the person writing ads and the person who built the page are usually not the same person, and nobody is checking the join.

2. The first screen on a real phone

Not the desktop preview. An actual phone, on mobile data, cold cache.

What I want to know is what a buyer can see and do without scrolling. If the product name, the price, and the add to cart are not all in that first screen, every scroll after it is a tax you are charging the visitor for no reason.

3. Where the page gets slow

I care about when the page becomes usable, not the score out of 100.

A store can pass a speed test and still feel broken, because the number people optimise for is not the number people experience. So I load it on a throttled connection and watch. The moment where I would have given up is the number that matters.

4. What the reviews say the product does

Not what the page says it does. The two are different more often than you would expect, and the reviews are usually more persuasive than the copy.

Customers describe the product in the words other customers search for. That language is free, it is already yours, and most stores never move it onto the page.

5. The checkout, with my own money

I buy something. All the way through, on my own card.

Nothing else surfaces the small stuff: a required field that should not be, a shipping cost that only appears at the last step, a discount box that invites people to leave and go hunting for a code. You cannot see any of that from analytics, and you will not believe how bad it is until you have felt it.

Then, and only then

After that hour I have a short list, usually three or four things, ordered by how much traffic they touch. That is the plan. It is rarely exciting and it almost never involves a redesign.

The unglamorous version of this work is that most conversion problems are not design problems. They are a promise that did not match, a page that took too long, or a checkout that asked for one thing too many.