A store doesn’t break all at once.
It degrades, slowly, and you get used to it. What was fast a year ago is now “a bit slower”, then “I have to wait for it”, and one day you realise you’re avoiding opening the admin at all.
Here are five signals worth taking seriously, and the order of fixes that pays back fastest.
Sign 1: the admin takes eight seconds to load, and you’re the only user
Customers don’t complain about speed, because they’re looking at product pages that are cached. You’re looking at the order list, which isn’t cached and is recalculated from scratch every time.
Eight seconds each time you open the list, twenty times a day, is a quarter of an hour lost every day. But that isn’t the real problem.
The real problem is that the same slowness hits the processes you don’t see: processing an order, updating stock, sending a confirmation. Once those slow down, they start to fail.
The most common causes: an oversized meta data table, plugins writing their own history into that same table, and hosting that was chosen on price back when you had fifty products.
Sign 2: you have forty or more plugins and you keep putting off updates
This one is the easiest to check. Open the plugin list and count.
Over forty on a store means the functionality was assembled, not built. Every plugin is someone’s code, with its own development cycle, updated independently of the rest. The chance of two clashing grows with each new one.
The symptom that comes with it: you put off updating for months because you’re afraid to. That’s the worst possible outcome, because now you have both the risk of a clash and the risk of a security hole.
What to do: an audit. On most stores, somewhere between ten and fifteen per cent of plugins are either unused, or doing something two others already do, or replaceable with twenty lines of code.
Sign 3: checkout works, but you don’t know where people drop off
A question to test yourself: how many people put something in the cart yesterday, how many reached the address field, how many reached payment, how many paid?
If you can’t answer within a minute, you aren’t measuring.
This isn’t a technical shortcoming, it’s the most expensive hole in the whole setup. Without those numbers, every change to the checkout is guesswork, and the checkout is where you lose the most money per square metre.
What to do: set up events for each step, and only then touch anything. The order matters. Anyone who changes things first and measures afterwards will never know what helped.
Sign 4: every price or promotion change needs a developer
If you have to call someone to change a promotion, add a coupon or edit the text on a product page, you have two problems.
The first is cost. The second is faster and worse: you don’t run the promotions you would have run, because it isn’t worth the phone call.
That’s lost revenue that shows up nowhere, because it never happens.
What to do: whatever you change often has to be in your hands. That’s a question of how the store is set up, not a question of your technical knowledge.
Sign 5: turnover is growing, conversion rate is falling
This one is the most counterintuitive, and the most serious.
Sales are going up because you’re investing in ads and traffic is growing. But the percentage of visitors who buy is falling month after month.
That means you’re paying more and more for each customer, and the growth you can see is hiding the problem instead of revealing it. When the ad budget stops, everything stops.
The most common causes: slowness under heavier load, a checkout that doesn’t work well on some phones, or traffic you’re bringing in that was never going to buy.
The first two are solved in the store. The third is solved in the ads, and that’s usually where it’s hiding.
“Outgrown” doesn’t automatically mean migration
This is the part nobody tells you, because a migration is a bigger job and a bigger invoice.
In practice it’s often enough to:
Audit the plugins. Clean up the database. Move to hosting built for stores. Replace two or three heavy plugins with something lighter. Set up measurement.
That’s a few days of work that puts the store back into a state where it can run for years to come.
Migration makes sense when the business logic has outgrown the platform: when you have complex per-customer pricing, integrations with a warehouse and accounting, a large number of variants, or several markets with different terms.
If someone proposes a migration before they’ve done an audit, ask why.
The order of fixes, by return
1. Measurement. Without this, everything else is guesswork. A few hours of work.
2. Speed. Images, caching, a plugin audit, the database. The biggest effect per hour invested.
3. Checkout. Now that you’re measuring, you know which step to touch.
4. Product page. Photos, description, availability, delivery, returns.
5. The overall design. Last. Almost always last.
Most people work through this list backwards, start at number five, and then wonder why the nice new site didn’t bring more sales.
Send us your store’s address and you’ll get a read on these five points, specific to your case, plus an estimate of which of them pays back fastest.