Buying software that already exists is almost always cheaper and faster than building it. So the real question is not “should we build?” but “is there anything that already does most of this?” These five questions usually settle it.
1. Does something already do 80% of it?
If so, start there. Many organisations commission custom software for a need that a good off-the-shelf product covers. Try the product with real work before deciding it isn’t enough.
2. Is the way you work a strength?
If your process is what makes you good, forcing it into someone else’s workflow can flatten it. If your process is the same as everyone else’s, you gain nothing from a custom one.
3. How many workarounds would you live with?
Every off-the-shelf product needs some. A few is normal. If people would be copying data between systems or keeping side spreadsheets every day, that is a cost to put against the price of custom work.
4. What will it cost to own?
Custom software needs hosting, updates and support for as long as you use it. Compare the cost over five years, not the price on day one, and include the subscription fees you would otherwise pay.
5. Who will look after it?
A product has a supplier whose job is to keep it working. Custom software needs the same, whether that is the people who built it, your own team or someone else. Decide this before you start.
The middle route
Often the best answer is neither extreme: a product that already exists, adapted to your work, or one that connects to the systems you already use. That keeps the cost of custom work for the parts that truly need it.
Trunkfield Software, Lichfield