We tested 1,000 Magento stores. Food & drink retailers keep a phone waiting 17.4 seconds.
Google's own PageSpeed test, run one store at a time on a random sample. For food & drink retailers that was the best average score of any sector we measured — and still a seventeen-second wait. And being on the newest Magento barely helped anyone: current-version stores averaged 51.9 against 51.1 for end-of-life ones. It isn't the engine.
Yours takes about a minute to check: pagespeed.web.dev — enter your homepage, pick Mobile.
Free. We load your products on first, then walk you through it live.
It isn't the version.
It's the shop front.
We expected old Magento to explain the slowness. It doesn't. Stores on the current 2.4 averaged 51.9 out of 100. Stores on versions Adobe stopped updating years ago averaged 51.1. Statistically identical. Upgrading fixes your security — it will not fix your speed.
| Sector | Stores | Avg score | Under 50 | Wait for content |
|---|---|---|---|---|
| Industrial & trade supply | 41 | 44.1 | 59% | 17.4s |
| Home & garden | 94 | 47.0 | 56% | 19.5s |
| Health & beauty | 28 | 50.9 | 50% | 19.7s |
| Auto parts | 29 | 51.5 | 41% | 13.1s |
| Sports & outdoor | 30 | 51.6 | 43% | 18.2s |
| Fashion & apparel | 30 | 52.5 | 50% | 13.0s |
| Jewellery & gifts | 21 | 54.5 | 38% | 16.7s |
| Electronics & tech | 20 | 55.0 | 50% | 20.9s |
| Toys & kids | 24 | 55.7 | 29% | 14.4s |
| Food & drink — you | 35 | 55.9 | 31% | 17.4s |
Food & drink retailers
Food and drink came out top on score, which surprised us. Our earlier, smaller study had it bottom. The random sample corrected us, so we corrected the page.
A good score does not mean a fast shop. Seventeen seconds is a long time to stay hungry, and it is the number your customer actually feels.
Sector sub-samples are small — 20 to 94 stores each — so treat them as indicative. We publish the counts so you can judge for yourself. Note the pattern: a good score and a twenty-second wait sit happily on the same store.
"Wait for content" is Largest Contentful Paint — when the biggest thing on the page finally renders. Google's threshold for good is 2.5 seconds. Measured on a throttled mid-range phone on 4G: a lab test, but roughly what a lot of people are holding.
Most stores aren't broken. They're average — which is the trap.
46% scored under 50, but only 9% scored under 30. The genuine wrecks are rare. Nearly everything we measured is a store somebody plainly looks after — patched, hosted properly, taken seriously. If that's yours, nothing here says you've done a bad job. It says you've done the same job as everyone else.
You're not broken. You're average. And average is a 16-second wait. Nobody rushes to fix a store that's no worse than its rivals — that's how a whole industry settles at 51.8. It also means the bar for standing out is lying on the floor.
It usually isn't your hosting.
And it isn't you.
The theme ships far too much
Magento's standard themes ship a pile of JavaScript to every page whether that page needs it or not. That's most of the delay, and a bigger server doesn't fix it.
A rebuild was always £20k+
Magento 1 to 2 isn't an upgrade, it's a rebuild — every extension re-bought, hosting triples, and UK agencies quote £10,000–£30,000 over 8 to 24 weeks. Saying no to that was a reasonable decision, every time.
Nobody's job was to keep it current
Nine in ten stores we measured were on a current, supported version — and they were just as slow. Somebody kept the engine patched. Nobody's job was the shop front.
Keep the Magento engine.
Replace the front of it.
Your products, orders, customers and URLs stay exactly where they are. We put a modern storefront in front of them — and that's where the speed comes from.
Google PageSpeed on a client's live store — over 1,000 products, taking orders today, every module running. Measured 20 July 2026, on the same mobile test as every store in the study:
the test the study uses
(study average: 16.0s)
same store, same day
Worth understanding what that 3.6s is: Lighthouse tests a cold first visit — empty cache, throttled 4G, mid-range phone — the hardest conditions a store ever faces, and the same conditions every store in the study faced. So 3.6s against the study's 16.0s average is the like-for-like comparison, and it puts this store in the top 6% of the 1,000 we measured — only one store in eight cleared 70 at all. Out in the field it's quicker still: Google's real-user data for this store passes all three Core Web Vitals, with typical visitors seeing the main content in 2.3 seconds — inside Google's 2.5s bar. We publish both numbers because most people quote whichever one flatters them.
Anyone can score 100 on an empty page
Plenty of themes advertise a perfect score. So does this page you're reading — it's a single static page with no catalogue behind it, and that's an easy thing to do. Across the 972 real stores we measured, just 5.5% cleared 90. The honest test is a working shop: a real catalogue, real product images, search, filters, a basket, and the modules a business actually runs. That's where most of them fall over, and it's the number above that we'd stand behind.
Which is why the demo uses your products rather than ours. Test it on your own catalogue and the argument settles itself.
Don't take our word for any of it — run Google's test yourself.
Built so one person can run it
Latest Magento 2.4.9, a front end you restyle yourself without a developer, and 40+ modules — most with AI built in, so the blog, the social posts and the product copy get written without you sitting down to write them.
Your URLs, your rankings
Same addresses, same titles, same content, redirects mapped one-for-one before anything goes live. Google shouldn't experience it as a new site at all.
Not really a shop? That's fine
Plenty of trade suppliers take the enquiry online and do the deal on the phone. We run one exactly like that — no basket anywhere, quote requests, proforma and BACS.
Live in days, not quarters
Agency rebuilds run 8 to 24 weeks. The expensive part — the front end — is already built and refined across every store we've done.
It's all on this page. Nothing to download.
Every sector broken down above, the three causes of the slowness, and a one-minute check you can run on your own site without paying anybody: pagespeed.web.dev, homepage, Mobile tab. No PDF, no email gate — bookmark the page, send it to a colleague, argue with it. The numbers don't change depending on who's asking.
The founding deal
The platform is £9,995 on our site and it stays there. A handful of founding places go for less than that, in exchange for a case study and a reference call.
£6,000build
50% on confirmation, 50% when we go live.
£600per month
Year two onward: £3,000 + £600/month.
Hosting is your choice — on your own account in your name (we set it up, we don't mark it up), or we manage it and bill you.
The £600 does one job: keeps it current.
- Magento version upgrades and security patches
- All 40+ modules kept up to date
- Front-end updates
- Priority support — you deal with Rick, not a queue
- Feature requests prioritised where feasible
If you stop paying, nothing switches off
After year one you can host it yourself and not renew. Everything carries on running exactly as it is — nothing phones home. You'd stop receiving updates, so keeping it secure becomes your job rather than ours. You'd own a working store either way.
We'll put your products on it before we speak.
Not a slideshow. Not somebody else's shop. Before the call we take a selection of your own products — your part numbers, your prices, your branding — and load them onto the platform. Then you give us half an hour on video, looking at your own catalogue running the way it should.
It's free, and there's no obligation on the other side of it. If it isn't obviously better than what you have now, you'll know inside five minutes and you can say so.
Slots from 9am to 11pm UK time, so we can cover any time zone. Prefer to talk? 0330 024 5225 or [email protected]
Straight answers
What happens to my Google rankings?
It's the right thing to be nervous about — it's the one asset a rebuild can genuinely damage. Your URLs stay your URLs: same addresses, same titles, same content, redirects mapped one-for-one before anything goes live. And page speed on mobile is one of Google's own signals, so this is fixing a signal you're already losing marks on, not putting one at risk.
How long is my shop down for?
It isn't. Your current site keeps trading the whole time we build alongside it. There's one switchover, done out of hours, and if it went wrong we'd point it straight back. You wouldn't lose a day's orders.
What about my existing orders and customer accounts?
They come with you — customers, order history, addresses. It's the same Magento database underneath; we're changing the front, not starting again. That's precisely why this is quicker than an agency rebuild.
Why is this so much cheaper than an agency?
Because the expensive part is already built. An agency builds the front end from scratch every time — a couple of hundred hours before anyone has touched your products, and you pay for all of it. We built ours once and have refined it across every store since. It's the difference between buying a house and having one built.
Which also tells you where we're not the right answer: if you want something designed from a blank sheet, bespoke to you, that's an agency job and worth their money.
Isn't Magento 2 slower than Magento 1?
Often, yes — and most people selling you a rebuild won't admit that. Magento 2 added a lot of abstraction, needs Elasticsearch, Redis and Varnish to perform, and the standard theme ships a mountain of JavaScript. That's exactly why we don't do a standard rebuild.
How do I know you'll still be here in two years?
You don't, and that's fair. So it's built so you don't need us: your hosting in your name, full source code, and if you stopped paying tomorrow the store keeps trading. We'd rather you stayed because it's worth it than because you're stuck.
What isn't included in the £6,000?
The £6,000 is the platform — the storefront and every module, installed on your Magento. What varies is your catalogue. A straight move of clean product data is usually most of it. Rebuilding product content, custom attributes like vehicle fitment, integrations with your accounts or ERP, and training are quoted separately once we've seen what's there. We'd rather tell you that now than surprise you later.
Run the test on your own site first.
It takes a minute, it's free, and it's Google's tool not ours: pagespeed.web.dev
If the number bothers you, we'll show you what your shop looks like without it.