Custom management software or a subscription: how to really choose

A company I spoke with recently pays just over two hundred euros a month for its management software. It has done so for four years. It uses maybe thirty percent of the available features, and in the meantime it has rewritten three of its internal processes to fit what the software allowed.

This isn't a story about bad software. It's a story about a choice made without criteria.

This article isn't meant to convince you that you need custom software. It's meant to give you the criteria to figure out which of the two cases you fall into, including the case — a common one — where the right answer is to stay on your subscription.

If you'd like a broader overview of the two options first, I covered the general pros and cons in Custom Software or Off-the-Shelf: What's Really Worth It for Your Business. Here we go one step further: concrete signals, the cases where custom isn't worth it, and the numbers over five years.

It's worth being honest about the advantages, because they are real and there are plenty of them.

Subscription software gets you started right away: you sign up, log in, and you're up and running the same day. The upfront cost is close to zero, which means you don't have to tie up capital to find out whether the tool actually works for you. Updates arrive on their own. There's support to call when something goes wrong, and a community of other users who have already run into your exact problem.

There's also a less obvious advantage: subscription software was built by observing thousands of companies. If your process is standard, that software embeds best practices you wouldn't have thought of yourself.

For a huge number of businesses, this is the right answer. Full stop.

The problems don't show up in the first year. They show up later.

The fee never ends. It's the simplest point and the one companies underestimate the most. Two hundred euros a month is still two hundred euros a month eight years from now, when the software has long stopped being new and is simply something that's there.

The cost grows with you. Almost all subscription management tools are priced per user. That means every new hire comes with a software cost attached, and the moment your business is doing better is also the moment the expense goes up.

The features you need are in the next tier up. This is the commercial mechanism the entire industry is built on. The feature that solves your specific problem almost always exists, but it sits two plans above yours, alongside twenty features you'd never use.

The hidden cost: bending your business to the software. It's the line item that never appears on any quote, and it's almost always the most expensive one. Every time someone fills in a field creatively because the software doesn't account for your case, every time someone exports to Excel to do a step the system can't handle, you're paying. Not in euros — in hours.

Your data lives in someone else's house. In day-to-day work, this changes nothing. It changes the day you want to move elsewhere, or connect the system to another tool, and discover how hard it is to get out what you've put in.

Four signals. If you recognize at least two, the conversation is worth having.

You have a process no standard tool respects. Not "we do it a bit differently." I mean a process that is how your company creates value, and that no off-the-shelf software accounts for because you're the one who invented it.

You're paying for modules you never open. If you're covering the cost of a full suite to use a quarter of it, the math changes quite a bit.

Your competitive edge lies in how you work. If what sets you apart from a competitor is speed or the way you handle a particular stage, standardizing that stage inside the same software they use is an elegant way to cancel out your advantage.

It has to talk to something you already have. Machinery, a legacy system, a warehouse with its own logic, a supplier that sends data in its own format. Integrations are the most solid technical reason custom development pays for itself.

This is the most important part of the article, and I'm writing it knowing it works against me.

If you have fewer than five employees, almost certainly not. The complexity isn't yet high enough to justify the investment, and the flexibility to switch tools is worth more than customization.

If your processes are standard, no. Invoicing, accounting, attendance tracking: these are solved problems. Building them from scratch means paying to rewrite something that already exists and works better than anything you could build in the first year.

If nobody in the company can follow the project, no — and this is the condition that sinks more projects than all the others combined. Custom software requires someone on the inside to answer questions, test interim versions and say what isn't working. If that person doesn't exist or doesn't have time, the project will produce software nobody wanted.

If you're trying to save on the subscription fee, no. You choose custom software because it does something the alternative doesn't, not because it costs less. Sometimes it does cost less, but that's a consequence, not a reason.

Let's take a typical company: ten users, mid-range subscription software.

In my experience, the break-even point usually falls between the second and third year — and it moves further out every time the custom project is designed bigger than it needs to be.

Which brings us to the last part.

The choice isn't binary, and the most realistic answer is almost always a hybrid.

You build custom only the process that makes you different: the part where your company has its own way of working. For everything else — accounting, e-invoicing, payroll — you use standard subscription tools, which are cheap and already compliant.

The result is a small project, delivered in weeks rather than quarters, that solves the real problem and leaves out everything that doesn't make sense to reinvent. It's also the best way to keep mistakes small: if the project is small, a misjudgment costs you little.

Answer just one question: is there a process in your company that no off-the-shelf software respects, and that would save you hours every week if it were?

If the answer is no, stay where you are and stop thinking about it.

If the answer is yes, you don't need new management software. You need to build that one piece, and leave the rest as it is.