Build first what hurts most when it breaks

When you build a product with little time, you don't decide what to do.

You decide what goes first.

And that matters almost more.

Here's how I think about it.

For years we built our products in the gaps left by client work. We didn't have a product team. We had spare hours. And with spare hours you can't do everything at once: two half-finished features are worth less than one finished one.

So you need a rule for ordering. The one I use most is simple.

Build first the part that hurts most when it breaks, not the part that looks best in a demo.

What sounds important and what is

Take a booking engine for a holiday rental. It has two big pieces: the calendar, which says which days are free on every platform where the place is sold, and card payments.

Payments sound more important. It's what brings in the money. It's the first thing anyone asks in a meeting: "Can people pay online?"

But think about what happens when each one fails.

If payments fail, there are workarounds. You ask for a bank transfer or charge on arrival. It's a nuisance, but nobody loses their holiday.

If the calendar fails, you have two families with their suitcases outside the same room. There's no workaround for that. It's a phone call nobody wants to make.

So the calendar goes first. Even if it's less impressive.

Customers ask for what they can imagine

Here's what makes it hard: the customer will almost never ask for it.

Customers ask for what they can picture themselves using. Paying online, they can picture. A double booking, they can't, until it happens to them.

That's why setting priorities can't be delegated to the request list. It's your job. Your job is to imagine the problems before they do.

It's also your job to say it clearly to whoever is building with you. If someone has two tasks for the same weekend and nobody has told them which comes first, they'll do both halfway. Not for lack of effort. For lack of order.

Three questions for ordering

When I need to decide what comes first, I ask three questions:

• What happens if this fails? Not whether it's useful, but how much damage it does broken.
• Is there a manual workaround? Whatever can be done by hand for a season can wait. Whatever can't, can't.
• How many people does it affect? A failure that bothers one person isn't the same as one that ruins everyone's day.

With those three answers, the order usually sorts itself out. And it almost never matches the order you'd have put in a presentation.

Why it matters more without investors

A company with plenty of money can build several things in parallel and see what works.

We can't. Services pay for the product, and that only works if every hour of product goes to what matters most. When time is scarce, order is strategy.

What brings in money has to work, of course.

But a business taking payments by hand for a while is a nuisance.

A business with a broken heart is a disaster.

P.S. If you don't know where to start, ask yourself what would make you pick up the phone on a Sunday. Start there.