ECOMMERCE REPLATFORMING
A shop that falls over when you email your own mailing list is losing money on a date it can predict. Here is how to tell whether you have outgrown your platform or merely outgrown its hosting, what to move to, and why most shops should not build anything.
Most ecommerce platforms are fine. They are fine for years, and then a business grows past the shape the platform assumed, and every month a bit more effort goes into keeping it upright rather than into selling.
The symptom people describe is usually speed, but speed is rarely the real complaint. The real complaint is that the business has started arranging itself around the software: sending the newsletter in batches so the site survives it, not running the promotion because last time it fell over, keeping a spreadsheet because the platform cannot express how you actually price things.
That is the point worth naming. It is not that the site is slow. It is that you are bending the business to fit the tool.
The test. In the last six months, has anybody said "we cannot do that, the site will not handle it"? If a commercial decision has been made to protect the software, the software has become the constraint, and that is a different problem from performance.
Six symptoms, and the last one is the one that actually decides it.
The one moment you have engineered demand is the one moment the site cannot take it. Newsletters end up staggered over hours to keep it alive.
Because it lives in two places and they reconcile overnight, or by hand, and the gap is where the oversells happen.
Extensions conflict, something breaks, and it takes a developer and a weekend. So upgrades get postponed, which makes the next one worse.
Trade accounts, quantity breaks, configured products, per-customer pricing. The platform can nearly do it, and the gap is filled by a spreadsheet.
Orders retyped into the accounts system, stock updated in two places, invoices raised by hand. The cost is real and nobody counts it.
The promotion nobody ran, the channel nobody added. This is the expensive one, because it never appears on any invoice.
Roughly in order of cost and disruption. Most shops should stop at the second.
A surprising share of "we have outgrown it" turns out to be an underpowered server, no caching and a database nobody has tuned. Worth ruling out before anybody proposes a rebuild.
Verdict: check this first. It is the cheapest possible answer and sometimes it is the right one.
Somebody else runs the servers, handles the traffic spikes, keeps it patched and takes the payments. The app ecosystem covers most of what a growing shop needs.
Verdict: this is the right answer for most shops, and we will tell you so. If Shopify fits, take it. You will spend less and sleep better.
If one part is the problem, replacing everything to solve it is expensive. Integrations, caching, a checkout rewrite or a stock sync can often be done in place.
Verdict: worth costing properly before assuming a rebuild.
Built around how your business actually works, integrated with what you already run, and scaled for the days you care about.
Verdict: right when the section below describes you, and hard to justify when it does not.
We build bespoke ecommerce and we would still rather say this plainly. If your products are reasonably standard, your pricing is reasonably standard, and your fulfilment is reasonably standard, a hosted platform will serve you better than anything we would write, for a fraction of the money.
Building is worth it when the things that make your business yours are the things the platform cannot express. That is a real situation, and it is narrower than anybody selling a rebuild will tell you.
The platform is the ceiling on your growth. Not an irritation. An actual limit on turnover, where you can name the promotions you did not run and the emails you did not send.
Your pricing or fulfilment is genuinely unusual. Configured products, trade pricing, made to order, anything where the platform can nearly do it and the gap is filled by a person.
The integrations you need do not exist. When the back office is where the cost is, and no app connects your ecommerce to the system that actually runs your stock and your accounts.
Peak days are the business. If a handful of days a year carry a large share of the revenue, the architecture has to be built for those days rather than for the average.
The clearest example we have of a platform being the ceiling.
TMC had outgrown a Magento build that was constantly down and constantly being patched. It supported around forty concurrent users on a good day. With tens of thousands of people on their mailing list, they had to stagger newsletters over many hours for fear of taking the site down, which meant the one lever they had for demand was the one they could not pull.
We built a bespoke, load-balanced platform integrated with Stripe, PayPal, Apple Pay, Google Pay and Klarna, and linked it to Sage Brightpearl so orders, invoices, products, prices and stock sync in one direction of truth. They now send the email and get thousands of concurrent customers without a problem.
The result was a multiple-times increase in turnover, and a ten year partnership that is still running.
We price the job, not the day, and you see the price before anything starts. As a rough shape, using our current rate of £15 a token:
Whatever the size, the number is agreed in writing first and does not move afterwards.
Not if the URLs are preserved and redirected properly, which is a solved problem and a normal part of the work. Rankings get lost when that step is treated as an afterthought, not because a site was rebuilt.
Longer than anybody wants and it depends entirely on the integrations rather than the shop itself. Anybody giving you a number before reading what you have is guessing. The assessment gives you a real one.
Yes. Plenty of useful work is making a hosted platform fit a business properly: integrations, custom pricing logic, connecting the back office. Staying put is often the better call.
Then that is what the written assessment says, and you will not have spent much to find out. We would rather be the people who told you that.
Describe what you are working around today. You get a straight answer on whether this is hosting, a specific fix, or a real replatform, with a price for whichever it turns out to be.
Get a price