This question is usually framed wrongly from the start: "should we buy or build?" The right framing is: "which parts should we buy, and which should we build?" Very few businesses need everything custom, and very few fit a packaged product exactly.
The one-question test
For each process, ask: is the way we do this part of why customers choose us?
- No → use a finished product. Cash books, attendance, asset registers, basic payments — nobody ever won business by keeping a better cash book than a competitor.
- Yes → consider building. How you price for dealers, how you allocate cost per construction site, how you dispatch technicians — that is worth investing in.
Apply that test across your list of processes and the usual result is: 80% buy, 20% build. That is also the sensible split of the budget.
Compared on five dimensions
| Packaged | Custom | |
|---|---|---|
| Time to something usable | Days to a few weeks | 2–12 weeks depending on scope |
| Up-front cost | Low, paid monthly | Higher, paid by milestone |
| Long-run cost | Rises with users and branches | Mostly infrastructure and maintenance |
| Fit to your process | You adapt your way of working | The software follows how you work |
| Main risk | Vendor lock-in and pricing changes | Depends on the team’s quality; needs someone to maintain it |
Three cases where building is right despite the cost
- 1The process is a competitive advantage. Forcing it into a mass-market mould discards the thing that differentiates you.
- 2No product can cope with your scale or specifics. Tens of thousands of price rows by dealer tier and region, say, or batch traceability demanded by export customers.
- 3Subscription cost has passed the cost of owning. With many users, per-seat fees compounded over a few years can exceed a custom build — calculate over three years, not one.
Three cases where building is wrong
- Believing you are unusual without testing it. Trial a finished product for two to four weeks first; a great deal of "unusual" turns out to be habit.
- When nobody can describe the process clearly. Building custom software for a process nobody understands is the most expensive way to learn that process.
- When nobody will maintain it. Custom software needs an owner; without that plan it becomes a burden within two years.
Questions to answer before committing
- 1If I want to change vendor in two years, how do I get my data out and in what format?
- 2Does this product have an API to connect to the part I build, or only an Excel export?
- 3What is the three-year total for each option, infrastructure and support included?
- 4Who will maintain this software after the project ends?
SealCore sits on both sides of this question: we have a packaged product (SealSales) and we build custom software. So in a survey, "you do not need a custom build yet" is a genuinely possible conclusion — and saying it early is cheaper for both sides.
Want to talk specifics?
SealCore surveys at your premises and sends a fixed quote after the first session — including when the conclusion is that you do not need custom software.
