A hosted online shop is like a tent. You pitch it in an hour, you need nothing else for it, and for a first season it is exactly the right answer. The trouble starts when you want to live in it all year, add a workshop and turn the heating on.
Most articles on this subject are written either for somebody about to buy a shop, or by a supplier who wants to sell a migration. This one is for somebody who already has the hosted shop, finds something about it wrong, and does not know whether that is a reason to leave or something to get used to.
None of the five points below is on its own a reason to leave. What decides is how many apply at once.
Short answer
- Five thresholds: fees, plugins, manual work, speed and data. One on its own means nothing.
- The most common real reason for leaving is neither price nor speed but that something you need simply cannot be done, and nobody can change that.
- Platform and plugin fees are judged on an annual total, not monthly. Monthly, everything looks cheap.
- Migrating is not about technology but about data and addresses. The most expensive mistake is losing search positions, not misconfiguring a theme.
- If your problem is solved by one plugin for a few hundred a month, stay. Leaving is months of work.
Five thresholds
Go through them in order and count how many apply to you. The reading is under the table.
| Threshold | When it is crossed | Why it matters |
|---|---|---|
| Fees | Plan and plugins exceed 5,000 CZK a month | Over three years that is 180,000 CZK, the price of a custom build |
| Plugins | You have more than ten and two of them overlap | Each is third party code that slows the shop and can break |
| Manual work | Somebody retypes data between systems daily | An hour a day is 250 hours a year, six working weeks |
| Speed | Mobile score under 40 and nothing can be done | You have no access to whatever is causing it |
| Data | You need an integration the platform does not support | The only threshold money cannot get around |
One threshold crossed means stay. It gets solved with a plugin, a higher plan, or by getting used to it.
Two or three mean do the arithmetic. Add up annual fees and hours of manual work and compare that with the price of a custom build including running it. For plenty of companies in this band, the migration pays for itself within two years.
Four or more mean you are already paying for not leaving. Companies sit in this state for two years and keep postponing, because they are afraid. The fear is justified but aimed at the wrong thing: moving the shop is not the danger, losing search positions is.
The threshold not worth working around
Four of the five can be lived with for a while. The fifth cannot, and deserves attention of its own.
A hosted platform lets you do exactly what is in its catalogue. When you need an integration with the stock system you actually run, a configurator following your rules, or prices by customer type, there are two answers: either a plugin exists or it cannot be done. There is no third option, and neither a higher plan nor a better supplier changes it.
This is the most common genuine reason for leaving that we have seen. Not price, not speed. The company grows to the point where the platform stops it doing what it does. And because it happens gradually, most people read it not as a platform problem but as their own bad luck.
We retype orders into the stock system by hand because there is no integration. We are used to it and we have somebody part time for it. It takes two hours a day and nobody thought to cost it.
Two hours a day is 500 hours a year. At 300 CZK an hour that is 150,000 CZK, more than the whole migration. We worked it out and decided in an afternoon.
When to stay instead
Most people who find this article should stay. It is fair to say that up front rather than at the end.
- When you have under a hundred orders a month. At that volume a custom build does not pay for itself and the manual work is manageable.
- When your problem is one missing feature that a plugin covers. A few hundred a month is negligible against a migration.
- When there is nobody to run the shop. A custom build means owning updates, backups and outages. A hosted platform does that for you and it is real value, not marketing.
- When the problem is sales, not the platform. A shop with no visitors is not saved by migrating. You will move and have just as few orders, only more expensively.
The last point is the most common. A company feels the platform is behind weak sales, because it is the one thing that can be swapped. Before costing a migration, check how many people visit the shop at all and how many reach the basket.
What a migration actually involves
Technology is the smallest part of it and the cheapest. Ordered by effort, not by what people fear.
- 1Export the list of every address and its traffic from the existing shop. This comes first, not just before launch. Without it there is no redirect map and you lose search positions.
- 2Clean the catalogue. A migration is the only moment when going through products, merging duplicates and dropping what has not sold in years actually pays. On the new platform there will never be time again.
- 3Move the data: products, variants, prices, customers and orders. Orders often need not move at all; archiving them saves a good deal of work.
- 4Connect the systems you have been bridging by hand because of the platform. This is the part you are leaving for, and it should be done immediately rather than after launch.
- 5Launch, then watch 404s and the positions of your key products for four weeks. A dip lasting two to six weeks is normal; one still there after two months is a fault.
Three questions to answer first
The answers decide whether a migration goes well far more than the choice of new platform does.
Who will run the shop afterwards? A custom build means somebody has to apply updates, watch backups and respond to outages. Either that is your supplier and you know their response time, or it is your own person. The third option ends in an out of date shop within a year.
What happens to the data if we part ways with the new supplier? You have just left a platform because of dependence. There is no sense leaving for a different, smaller dependence. The code, the database and the domain should be yours, and it should say so in the contract.
How much of today's manual work do we want automated straight away? A migration is the cheapest moment to do it, because everything is being touched anyway. Adding it a year later means paying twice for the same thing.
How we handle it
Shops and custom systems are our largest package, and it is fair to say they are not worth it for everybody.
- Before we offer anything, we work out whether it pays. Annual fees, hours of manual work and order volume. If the answer is that you should stay on the platform, we write that.
- The redirect map is part of delivery, not a line item, because without it a migration cannot be done responsibly.
- Integration with the systems you already run is the main reason migrations happen, so it gets done at the start rather than after launch.
- Code, database and domain are yours and it says so in the contract. There is no point swapping dependence for a smaller dependence.
And the honest other side: your own shop means owning its operation. A hosted platform handles updates, backups and availability for you, and that is real value. With us those are handled by us, but it is a service rather than a given, and it belongs in the budget.
Frequently asked questions
When is it worth leaving a hosted online shop?
When several of five thresholds are crossed at once: fees over five thousand a month, more than ten plugins, daily manual retyping of data, mobile speed under forty, and needing an integration the platform does not support. One threshold on its own means staying and solving it with a plugin or a higher plan.
What is the most common reason for leaving a platform?
Neither price nor speed, but that something you need simply cannot be done. A platform allows exactly what is in its catalogue; when you need an integration with your own stock system or prices by customer type and no plugin exists, there is no third option. The company grows to the point where the platform stops it doing what it does.
When should I stay on a hosted platform instead?
When you have under a hundred orders a month, when your problem is one missing feature with a plugin available, when there is nobody to run a custom build, or when the problem is traffic rather than the platform. The last case is the most common: a shop with no visitors is not saved by migrating, only made more expensive.
What is the biggest risk in migrating?
Losing addresses and with them search positions. Hosted platforms use their own format for product and category URLs and a new system builds them differently. Without a redirect map that usually means an organic traffic drop of tens of per cent. The map is the one document that genuinely matters.
Do I have to move old orders across?
Usually not. Archiving them, for instance as an export, is enough, and the new system starts from zero. That saves considerable work and keeps the data available for accounting and returns. Moving them makes sense only when something live depends on them, such as a loyalty scheme.
How long does a migration take?
For an ordinary shop, expect twice as long as a new website of the same size, so months rather than weeks. The bulk of the work is not technology but cleaning the catalogue, moving data and connecting the systems you have been bridging by hand.
Summary
- Five thresholds: fees, plugins, manual work, speed and data. One on its own means stay.
- The most common genuine reason for leaving is that something you need simply cannot be done.
- Judge fees on an annual total. Monthly, everything looks cheap.
- An hour of daily manual work is six working weeks a year. Cost it before deciding.
- The riskiest part of a migration is not technology but losing addresses and search positions.
- A shop with no visitors is not saved by migrating. Check the traffic first.
Not sure whether a migration pays
Tell us what the shop runs on, what you pay monthly and what cannot be done on it. We will add up the annual total and compare it with the price of a custom build. If the answer is that you should stay, we will write that rather than fight for the job.