Almost every software horror story — the project that doubled in price, shipped a year late, or arrived as an unmaintainable mess — was avoidable. Not with better luck, but with better questions, asked before anyone signed anything.

The hard part is that you're evaluating something you can't fully inspect. You can't read the code, and even if you could, you couldn't judge it. So you end up trusting a pitch. The fix is to stop judging the things you can't verify and start judging the things you can. Here's the exact checklist we'd want a client to run on us.

1. "Can I use something you've actually built?"

A portfolio of polished screenshots proves someone can take a screenshot. Ask to use something real — log into it, click around, or at minimum get a live walkthrough of working software, not slides.

If a studio can't show you a single thing running, that tells you something. If they can, you learn more in ten minutes of clicking than in any sales deck. (It's why we built a full, working ERP — so we can show, not tell. Ask us for a walkthrough.)

2. "Do I own the code when we're done?"

This is the fastest way to find the trapdoor. The answer should be an immediate, unqualified "yes — all of it, in accounts you control."

If there's any hesitation — "well, it runs on our platform," "you licence it from us," "we host it for you" — you've found lock-in dressed up as a service. You'll be paying rent on your own software forever, and leaving will be painful by design. Walk.

3. "Fixed price, or hourly?"

Hourly billing means every inefficiency, every rewrite, every over-run lands on your invoice. The studio's estimating risk becomes your bill. There's no incentive to be fast, because slow pays better.

A fixed price flips the risk onto the people who can actually control it. It forces tight scope and real trade-offs up front — which is exactly the conversation you want to have before the money is spent, not after.

The most useful thing a studio can do in the first meeting is tell you what to cut. If everything you ask for is met with an enthusiastic "yes, we can do that!" — be careful. You're talking to a sales team, not a build team.

4. "Who actually writes my code — and will I talk to them?"

The classic agency pattern: senior people charm you into signing, then hand the work to juniors you never meet, mediated by a project manager who translates. Requirements get lost in the relay, and by the time you see the result, it's expensive to fix.

Ask who does the actual work and whether you'll be in direct contact with them. The best answer for a project your size is simple: the person building it is the person you talk to.

5. "What happens after launch?"

Software isn't "done" at launch — that's when the real-world edge cases show up. "Support" should be a defined thing with a number and a timeframe, not a vague "we'll be around." Get it in writing. A studio that disappears the day after go-live was never really on the hook.

6. "How many clients are you juggling right now?"

Attention is finite. A studio running fifteen projects at once is running yours in the gaps. There's nothing wrong with a small shop — there's a lot right with it — as long as they're honest about capacity. Fewer clients usually means your project gets the focus it needs and your deadline is real.

The red flags, in one place

  • Vague scope. "We'll figure out the details as we go" is how a €20k project becomes €60k.
  • No price until deep into "discovery." If discovery is a paid phase that ends without a firm number, you're funding their sales process.
  • Nothing you can touch. Testimonials and logos, but no working software you're allowed to use.
  • Proprietary everything. Their platform, their hosting, their tools — a maze you can only navigate by paying them.
  • No demos until the end. If you don't see working software until "delivery," you can't catch a problem until it's expensive.

The green flags

  • They put the important stuff in writing — price, deadline, ownership, support — before you commit.
  • They show you real, running software without being pushed.
  • They'll talk you out of features you don't need, or out of hiring them at all if a cheaper tool fits.
  • You see working software regularly — weekly, ideally — not one big reveal at the end.

The short version

You can't grade the code, so grade everything around it: can you see real work, do you own the result, is the price fixed, do you talk to the builder, and is the after-launch plan written down. Those five answers predict how a project goes better than any pitch.

For the record, in case you want to hold us to our own checklist — our answers are: yes, you can use the software we built; yes, you own 100% of the code; fixed price on a fixed deadline; you talk to the founders who build it; and 30 days of support is included. We take two clients at a time, and we'll tell you if you'd be better off buying something off the shelf.