When a Custom Store Is Worth It, and When It Is Not
A practical test for whether your requirements actually justify building rather than buying.
When is a custom e-commerce build justified?
A custom store is justified when specific requirements genuinely cannot be met by an established platform — usually unusual pricing rules, inventory logic across locations, or integration with a system that has no API. Preferences such as wanting full control are not requirements, and most stores are better served by a platform with targeted custom work around it.
Custom e-commerce builds are proposed more often than they are justified. The platforms have become capable enough that most stores are better served by one, and the businesses that genuinely need custom usually know why. Here is how to tell which you are.
Start by writing down what does not fit
Not what you would like, but what an established platform genuinely cannot do for your business. Be specific: 'pricing depends on customer contract tier and order volume, with different tiers per product category' is a requirement. 'We want full control' is a preference.
Most lists produced this way turn out to contain two or three genuine constraints and a dozen preferences. The genuine constraints are usually about pricing rules, inventory logic across locations, or integration with a system that has no API.
Check whether the platform actually cannot do it
This step gets skipped surprisingly often. Platform capabilities change, and requirements that needed custom development three years ago are frequently native features now — server-side discount logic, structured product metadata, B2B price lists.
It is worth an hour with the current documentation, or a conversation with someone who works with the platform regularly, before concluding it cannot be done.
Cost the whole thing, not the build
A custom store costs more to build and then continues costing. Hosting, monitoring, security patching, dependency updates, and a developer available when something breaks at a bad moment. A reasonable planning assumption is around a fifth of build cost per year.
Against that, a platform costs a monthly fee plus transaction percentages plus apps. At high order volume the commission becomes material and the comparison shifts. At low volume it rarely does.
Compare over three years, including the infrastructure and support you would be taking on. That is the comparison that reflects reality.
Ask who will maintain it
This is the question that decides most of these cases. A hosted platform means you can hire commercially-minded people to run the store. A custom store means you need a developer available for the life of the system.
If the answer is 'the agency that builds it', understand that you are entering a long-term dependency, and negotiate the terms accordingly — your repository, your documentation, your deployment access.
The middle path most businesses want
In practice, most stores that consider custom end up better served by a platform with targeted custom work around it: an integration service handling the unusual logic, a custom app for the specific gap, custom theme code where the storefront needs it.
That keeps the infrastructure burden with the platform while solving the genuine constraints. It is less satisfying than a clean custom build and it is usually the right answer.
Takeaways
- Separate genuine constraints from preferences before deciding.
- Verify the platform cannot do it before concluding it cannot.
- Compare three-year total cost including infrastructure and support.
- A platform plus targeted custom work is usually the right middle path.
Related services
Related solutions
Related resources
More from the blog
Have a project like this?
A short conversation is usually enough to tell whether we are the right fit, and what the work would realistically involve.