Off-the-shelf system or custom software?

"It depends" is a correct answer to this question and a useless one. Below are the criteria that actually matter. For most companies the right answer is not custom software — and we don't mind saying so.

Start here: how many people will use it?

If three people use a tool, the cost of adapting to an off-the-shelf product is low. If thirty people use it, ten minutes a day lost to a poor fit adds up to a serious annual figure.

Roughly: the more users, the more the equation tilts towards custom.

An off-the-shelf system is enough if...

Your process runs close to the industry norm. Accounting, invoicing, basic CRM — these are mature categories. Rewriting them is almost never sensible.

Adapting to the product doesn't hurt. Sometimes "the software wants it this way" turns out to mean switching to a better method.

You have budget or time constraints. A packaged product works tomorrow; custom software works months from now at best.

This isn't where you compete. Nobody wins customers because of their leave-tracking process. Where you don't differentiate, off-the-shelf is fine.

Custom software is needed if...

The process itself is your edge. If what separates you from competitors is a process, forcing it into a standard product erases the advantage.

You're working around the system. If you export to a spreadsheet, edit and re-import; if approvals happen over chat; if the same data goes into two systems — the packaged product isn't actually working.

You need several systems to talk to each other. What's usually needed here isn't a new product but an integration layer, which can cost far less than a full custom build.

Licence cost scales per user. As the team grows, per-seat licensing can overtake the cost of a custom build within a few years. Do the arithmetic.

A third option: both

This is what we most often recommend in practice. Leave accounting in the packaged product, move sales and operations onto a system built for you, and connect the two.

That way you use mature, cheap software where it's mature and cheap, and build your own only where you're genuinely different.

One thing to do before deciding

For one week, note every moment you work around the current system. Every export, every manual copy, every question asked over chat.

That list decides for you. A short list means the packaged product is working. A long list means you're already imitating custom software with human effort.

---

If you've made that list and still can't decide, send it to us. If we think custom software isn't warranted, we'll say so.