Questions · Published
Should I build my app idea?
TL;DR
Building something rough to learn from is worth doing early, as long as you put it in front of the people you'd sell to. The bigger decision is when to invest in a proper first version: once you know who it's for, what the problem costs them, how they reacted to trying it, and ideally someone has committed to pay.
With AI tools, you can have something working in an afternoon. So the more useful question is what kind of building you're doing, and whether it's teaching you anything about the people you want to sell to.
Two kinds of building
Building to learn. A rough version, a clickable prototype or a one-page website with a price. It takes hours, and its job is to give customers something real to react to. This is worth doing early and often. Watching someone try your idea on their own work shows you things a conversation alone won't.
Investing in a first real version. Something customers can rely on, that you'll support, charge for and keep improving. This takes real time or money, often hundreds or thousands of dollars with an AI builder, so it's worth doing once you know enough to aim it well.
Most of the risk is in mixing these up: treating every rough version as the start of the product, and adding feature after feature without showing it to anyone.
When it's worth investing in a first real version
You're in a good position when you can say yes to most of these:
- Can you name people who have the problem? Real people who fit your customer. Five is a good start.
- Do you know what they do about it today, and what it costs them? Their current workaround is your real competition, and the cost tells you what the problem is worth.
- Have some of them tried something you've built, on their real work? What they did with it matters more than what they said about it.
- Has anyone paid, or committed to pay? A payment, a pre-order or a signed letter of intent. It shows someone has accepted your offer. Whether they get the outcome and keep paying is what you'll learn next.
- Can you say what the first version must do? The one outcome your first customers need, and what can wait.
If the first two are missing, talk to people. If the third is missing, build something rough and sit with someone while they try it. If the fourth is missing, put a price in front of them. There's more on that in how to know if people will pay.
Why it's easy to keep building anyway
Building feels like progress. Every prompt produces something you can see, and there's always another feature to add. Talking to customers, and especially asking them for money, feels much scarier. So it's easy to keep building and tell yourself you'll find customers once it's ready.
I call this the Product Illusion: mistaking a growing product for a growing business. Building early is fine. The trap is building on while learning little about who will choose it, how you'll reach them and what they'll pay.
A simple check: what do you know about your customers now that you didn't know last week? If the answer is "not much", put down the tools for a few days and go and talk to people.
"Not yet" is a useful answer
If you're not ready to invest in a first real version, that's useful to know. You've avoided putting months into something that might not sell, and you know what to find out next. Often it's a handful of conversations, a rough version someone tries on real work, and one honest ask for money. The guide to starting a software business walks through how to do that.