If You Build It, Why Will They Come?

A few weeks ago I caught myself three hours deep into a feature nobody asked for. It was working, too. The buttons clicked, the data saved, and I was feeling pretty damn smart.

Then the question showed up uninvited: if I can build this in an afternoon, why can’t everyone?

That’s the weird thing about where we are right now. It’s the age of the idea guy. Claude, Cursor, Lovable, take your pick. You can build almost anything you can describe. Which means building just stopped being the hard part.

Field of Dreams was a movie, folks

“If you build it, they will come” works great when a ghost is whispering it to Kevin Costner. In the real world, nobody is coming. Not to your vibe-coded invoice tracker, and not to mine either.

Think about it from the buyer’s side. Why would someone pay for your product, or even sign up for it, when they can open the same AI tool you used and build their own version by dinner? You’re not competing with other apps anymore. You’re competing with “eh, I’ll just make it myself.”

I’ve written before about whether AI can build your website and why Lovable is a lot less magic if you already know how to build. This is the other half of that argument. Say the AI builds the thing flawlessly. Now what? Who shows up?

Why do some people still get customers?

Here’s one I’ve been chewing on. Brian Casel, the guy behind Builder Methods (and earlier Audience Ops and Restaurant Engine), built his own email marketing tool called Mimeo. It’s kind of like what I’m doing with my own stuff, which is why I noticed.

Now, I have no idea what his numbers look like. Customers, revenue, none of it. That’s his business, and I’m not going to guess. But I’d bet money on why people pay attention when he launches something: he has a track record of building products, and he’s spent years teaching in public. People already know him. That’s a head start no amount of prompting buys you.

Which leads to the real question. If everyone can build, what stops everyone from succeeding? And what does it take to be the person who does?

We’re all too smart now

I’m including myself here. Anyone can ask Claude or ChatGPT how to launch a product and get a perfectly reasonable answer in four seconds. So we all feel like we’ve done the homework.

We haven’t. Reading the answer isn’t the same as having the experience, and most of us aren’t willing to do the slow, awkward work that builds it. So people do one of two things. They pour everything into building and leave marketing for “later.” Or they take shortcuts on marketing, skip the part where they put their own hands in it, and then act surprised when it doesn’t work.

The people who’ve launched a few things know what an ugly first week looks like. They know the first ten customers come from conversations, not from a landing page. The AI can tell you that sentence. It can’t make you feel it, and it definitely can’t do the conversations for you.

Experience is the one thing you can’t prompt. You earn it the slow way, by shipping, getting ignored, and trying again.

I’m building my own email and CRM tool right now, so yes, I’m talking to myself too. Right now it has exactly one user, and he’s typing this sentence. The cobbler’s kid needs new shoes.

What launching actually takes now

Building is cheap, so the work moved. Here’s the list I’m trying to hold myself to:

1. Use it yourself first. At a minimum, the product has to solve one of your pain points. It has to earn its slot in your own stack. If you won’t use it daily, why would a stranger?
2. Painstorm. Like brainstorming, but for problems. Write down every way the current situation hurts: what’s slow, what’s annoying, what people duct-tape together. Pain you can name is pain you can sell against.
3. Talk to real people. Actual conversations, not a survey, and definitely not asking your AI what customers “probably” want.
4. Test the pain before you build more. Go where your people already complain: forums, Reddit, Slack and Facebook groups, comment sections. Post the problem (not the product) and see if anyone says “ugh, yes, same.”
5. Count your distribution. Do you have an audience, a list, a track record, a community, anything? If the answer is no, that’s your actual first project. Build that, then the app.
6. Pick one ideal customer. Not “small businesses.” One specific kind of person with one specific problem you can picture by name.
7. Explain it in one sentence. If it takes a paragraph and a whiteboard, nobody is going to repeat it to a friend.
8. Write the positioning statement. More on that below.

The one-sentence test

Once you’ve built something you’re comfortable with, the questions get uncomfortable fast. Who’s the ideal customer? Can you say what the product does in a single sentence? Can you stand behind a laser-focused positioning statement?

There’s a common fill-in-the-blank template for this, popularized by Geoffrey Moore in Crossing the Chasm:

For [ideal customer] who [pain or need], [product] is a [category] that [key benefit]. Unlike [the alternative they use today], it [main difference].

It sounds easy. It isn’t. I wrote a whole post about why that blank is so damn hard to fill, and I still stare at it. But if you can’t fill it in, you’re not ready to market it, and no amount of Product Hunt confetti fixes that.

Nobody’s waiting for your launch

Here’s the part that stings. The launch post, the “we’re live!” tweet, the email to your three friends: none of it creates demand. It just announces supply, and there’s more supply than ever.

Demand already exists somewhere, in people who are hurting right now and complaining about it. Your job is to find them, listen, and be the person standing there when they say “I wish something did this.” That’s slow. It involves awkward DMs. You’ll hear “meh” more than once, and you’ll want to go back to building instead.

(I know, I know. Building is more fun. Coding is a dopamine slot machine and nobody’s ever rejected a pull request I wrote for myself.)

A quick PSA, folks: if you’ve talked to ten people and none of them care, that’s not a failure. That’s the cheapest data you’ll ever buy. Better to learn it before month six than after.

Build small, then go find your people

None of this means don’t build. Quite the opposite. I think little internal apps are one of the best uses of this whole moment, and I’ve seen what happens when you build a plugin in the age of AI. The difference is intent. A tool built to fix your own problem has a built-in customer on day one. A tool built because you could has to go find one.

If you’re sorting out how AI fits into your business, here’s how we think about it. We’d rather help you build the right small thing than a big thing nobody opens.

So here’s mine for you: if someone could rebuild your product by dinner tonight, what’s the reason they wouldn’t?