The short answer
Buy software for the things every business does the same way, such as accounting, payroll, email and basic CRM. Build a custom web app only for the process that sets you apart or that no product fits without painful workarounds. Most companies should buy far more than they build, and the custom piece, when there is one, should be small and sharply defined.
If the term is new to you, our explainer on what a web application is covers the basics.
When buying is the right call
Buying is right when a mature product already does at least most of what you need and the gaps are things you can live with. You get a working tool this week, a vendor who handles security and updates, and features you would never have paid to build yourself.
- The process is common across your industry.
- You can change how your team works to fit the software without losing anything customers value.
- You need it soon.
- The number of users is small, so per-seat pricing stays modest.
- You do not have anyone to own a software product internally.
A useful discipline is to try adapting your process before you reject a product. "It does not work the way we do" is sometimes a reason to build, and sometimes a sign that the way you work grew by accident.
When building is the right call
Building is right when the software would encode something your competitors cannot easily copy, or when you have stretched existing tools to the point where the workarounds cost more than the fix.
- The process is your advantage. A quoting method, a matching rule, a scheduling approach that wins you work.
- Staff re-key data between systems every day, and errors are reaching customers.
- Customers need to log in and see their own data, orders or documents in a way no product offers under your brand. See customer portal development for that case.
- Per-seat pricing has become painful because you have many light users.
- You want to sell the software itself, as a product or as part of your service.
- You have requirements about data location or handling that available products cannot meet.
Decision table
Score each row honestly. If most of your answers fall in one column, that is your direction. A split usually points to the middle path described below.
| Question | Points to buying | Points to building |
|---|---|---|
| How standard is the process? | Every firm like yours does it the same way | It is specific to you and customers notice |
| How well do existing products fit? | Most needs met, gaps are minor | You would use a fraction of it and work around the rest |
| How soon do you need it? | Weeks | You can wait months for a first version |
| How many users, and how often? | A few heavy users | Many occasional users, or external customers |
| How much does it need to connect? | Standard integrations exist | It must sit between several systems with your rules |
| Who will own it? | Nobody has time | A named person will own the roadmap |
| What if the vendor changes price or closes? | You could switch with modest pain | It would stop your business |
The real costs on each side
Both routes have costs that do not appear on the first invoice, and comparing sticker prices is how companies choose wrongly.
Costs of buying that people miss:
- Subscription fees that rise with every user and every plan tier you grow into.
- Paid add-ons for features you assumed were included.
- Set-up, data migration and staff training.
- Integration work to make it talk to your other tools.
- The ongoing cost of workarounds: exports, manual steps, a second tool to fill a gap.
- The cost of leaving, if your data is hard to get out.
Costs of building that people miss:
- Your own time to define requirements, test and make decisions. This is significant.
- Hosting, monitoring, backups and security updates every year.
- Maintenance as browsers, frameworks and connected services change.
- New features, because a useful tool always attracts requests.
- Everything a product gives you for free: permissions, audit logs, exports, password resets. See the hidden basics in how to scope a web app MVP.
Compare the two over three to five years, not on day one. Building has a high cost at the start and a lower cost per year. Buying is the reverse. Where the lines cross depends mostly on your user count and how much the subscription grows. For what shapes the build side, see our breakdown of web app development cost.
The middle path: buy the core, build the edge
The best answer is often both: keep proven products as the system of record and build a small custom web app that sits on top of them through their APIs. Your accounting, CRM and payments stay with their vendors. The custom piece handles the one workflow that is yours.
Consider a made-up example. A specialty distributor keeps its accounting package and its inventory system. It builds a small ordering portal where wholesale customers sign in, see their own pricing, reorder past orders and track deliveries. The portal reads stock from the inventory system and writes orders and invoices back. The business gets the custom experience where customers see it and takes on very little software to maintain.
This works only when the products you buy have good APIs, so check that before you subscribe to anything. Our plain-language guide to what an API is explains what to look for.
A five-question test
If you are still unsure, answer these five questions in writing. They settle most cases.
- Have we run a real trial of the two best existing products with real data, not just watched a demo?
- Can we describe, in one sentence, the thing the products cannot do, and does a customer care about it?
- What does the workaround cost us each month in staff time and errors?
- Who inside the company will own a custom tool for the next five years?
- Could a small custom piece on top of a bought product solve it?
If you cannot answer question 1 with a yes, do that first. It is cheap, and it either ends the discussion or gives you a precise list of gaps to build for.
Next step
Bring us your answers and the products you have tried. We will tell you plainly if we think you should keep buying. If a custom build or a custom layer makes sense, we will help you cut it to a first version and give you a fixed-scope quote. Read more about our web application development service, or start a conversation on the contact page.