What this guide covers
This guide gives a decision framework for choosing an ecommerce platform, based on catalogue structure, checkout requirements and integration load. It does not rank platforms against each other in the abstract, because the right answer depends entirely on what you sell and how.
Assumed knowledge
- A rough idea of your product range and how it is structured
- Knowledge of which other systems the shop must talk to
Most platform comparisons are written by people with something to sell. This one is organised around a single question: how far do your catalogue and your checkout rules sit from the platform's default model?
The further away you are, the more you will pay to bend the platform, and at some point building it properly costs less than the bending.
Start with the catalogue, not the design
Platform choice is decided by product structure far more than by appearance. Work through these before looking at any demo.
- How many products, and how many variants per product
- Whether products are configurable, made to order, or bundled
- Whether pricing changes by customer, quantity or contract
- Whether stock is held in one place or several
- Whether anything is sold by weight, length or area
A catalogue that answers all of these in the simple way will be well served by a template platform. A catalogue that answers two or three of them unusually is where costs start to diverge.
The three realistic options
| Option | Strongest when | Weakest when |
|---|---|---|
| Hosted platform, for example Shopify | You want to run it yourself, the catalogue is conventional, and you value not managing infrastructure | Checkout rules are unusual, or platform fees scale painfully with volume |
| Self-hosted, for example WooCommerce | You want full control of the data and the admin, and the catalogue is moderate | Nobody owns updates, hosting and security after launch |
| Bespoke build | Catalogue or checkout does not fit the default model, or an integration is the real product | The catalogue is conventional, in which case it costs more and does less |
Hosted platforms
A hosted platform handles infrastructure, security patching and payment compliance, which is genuine value and often underrated.
The constraint is the checkout. On most hosted platforms, the checkout is the part you can least change. If your business needs part payment, quote to order, purchase order approval or delivery rules that depend on the product rather than the basket, check exactly what is possible before committing, not after.
The other consideration is cost shape. Platform fees that look small at your current volume are worth modelling at three times your current volume.
Self-hosted platforms
WooCommerce and similar give you the data, the admin and the freedom to change anything. For a business that wants to manage its own catalogue and content without a developer, the admin experience is often the deciding factor.
The honest caveat is that self-hosted means somebody has to own updates, backups, security and performance. When that responsibility is unassigned, the site degrades quietly for a year and then has an expensive problem. Decide who owns it before you choose it.
Bespoke
A bespoke build is justified when the model is the problem.
- The catalogue cannot be expressed properly in the platform's product model
- The checkout has to behave in a way the platform resists
- An integration with stock, ERP or manufacturing is the actual work
- Performance at your catalogue size is a commercial constraint
If you cannot point at one of those four, bespoke will cost more and deliver less than a well configured template build.
Questions that change the answer
Ask these of any platform before deciding:
- What happens to my URLs if I move to this, and who writes the redirect map
- Can I export my full catalogue, customers and orders in a usable format
- Which parts of the checkout can I change, and which are fixed
- What does this cost at three times my current order volume
- Who is responsible for updates and security after launch
A note on replatforming
If you already sell online, the cost of moving is not just the build. It is the migration: URLs, redirects, product data, customer accounts, order history and reviews. That work is frequently underestimated and is where replatforming projects lose money.
If the current platform mostly works and the complaint is that it is slow or awkward, fixing it is often cheaper and lower risk than moving. Moving is justified when the model is wrong, not when the site is tired.
Key takeaways
- Decide on the catalogue and checkout model before looking at any demo
- Hosted platforms handle infrastructure well but constrain the checkout most
- Self-hosted gives control, but somebody must own updates, backups and security
- Bespoke is justified by an unusual model or a heavy integration, not by preference
- Model platform fees at three times your current order volume, not today
- If the current platform mostly works, fixing it usually beats moving
Frequently asked
Questions readers ask most
Is Shopify or WooCommerce better for an ecommerce site?
Neither in the abstract. Shopify suits a conventional catalogue where you want infrastructure handled for you. WooCommerce suits a business that wants full control of data and admin and has someone to own maintenance. The deciding factor is usually how unusual your checkout rules are.
When is a bespoke ecommerce build worth the cost?
When the catalogue cannot be expressed in the platform product model, the checkout needs behaviour the platform resists, an integration with stock or ERP is the real work, or performance at your catalogue size is a commercial constraint. If none of those apply, bespoke costs more and does less.
Can I move platform without losing my search traffic?
Yes, if the URL migration is planned properly: a signed redirect map, tested before launch, with every changed address pointing at a genuine equivalent. The migration is the risky part of replatforming, not the build.



