The planning fallacy
People underestimate how long their own projects will take, even when they know how similar projects went. Knowing about the bias doesn’t remove it. Software estimates are among the clearest cases.
The IKEA effect
People value things they built themselves more highly than equivalent things they didn’t. Which means the team assessing whether the custom system was worth it is the least reliable group to ask.
Three reasons that survive both
It’s your actual differentiator and the standard version would blunt it. The platform genuinely can’t do it. Or the workaround costs more over two years than building once — calculated with a real multiplier on your estimate, because of the above.
Everything else is solved
Payments, auth, email, scheduling, storage, analytics. Building these means maintaining them forever, against companies whose entire business is doing it better.
The test
If this worked exactly as well as your competitor’s, would you lose anything? If no, buy it. If yes, build that part — and it is almost always smaller than the system around it.