Marathons are full of distractions.

I recently met a founder. We’ll call them Zach.

A few months ago one of Zach’s professors told him they were looking for something that could take their interview recordings and turn them into text that they could use in their research — a simple transcription tool. They offered to pay Zach if he could build it.

Zach is a bonafide hacker. In just a few weeks he’s able to build and ship it by himself from his dorm room. An AI-enabled SaaS product that works beautifully.

The interface is elegantly simple, the set up is frictionless, the cost is competitively affordable. The professor creates an account and begins paying him to use it immediately.

Zach is on to something, and he’s excited. So he goes out and finds an agency to help him scale it. He pays them a few thousand bucks and they spin up some Google / Meta ads for him.

Pretty soon he’s doing 20,000 site visits per month. Thousands of people from all over the world are signing up — Brazil, India, The Philippines.

Then he realizes something.

Of the thousands of visitors each month, only 4-5 end up actually paying him money. That’s something like a 0.25% conversion rate from visitor to paid user.

He can’t understand why it isn’t making more money. He has a great product, he built it around a real problem from a real person who offered him money up front — in theory it’s the ideal origin story for a killer startup.

So what’s going wrong?

The answer: he stopped focusing on his ideal target customer.

When you launch a startup that gets some initial traction, it’s very easy to start thinking about how big it can be.

After that first paid customer you begin to imagine other types of people who could use your product for other things.

You may not even care why they want to use it in the first place as long as they are showing up — you just want to get it in front of as many people as possible. Generalize and expand.

And therein lies the problem.

Zach isn’t converting his ad dollars into paid users because he started focusing on EVERYBODY. It’s clear the 20,000 people who show up to his website each month aren’t his target customers because they aren’t paying him money to use his product.

Rather than continuing to focus on his original target customer of US professors with a budget for tools to help them with their research projects, he is letting anyone in the door.

That is a recipe for wasted time and money.

My advice?

Early, before you spend a dime on user acquisition, it’s better to simply take your first user and find another person who looks exactly like them — the best way to do this is to simply ask them if they have any peers who would want to try the product.

This 1) validates that you are building for a concrete target customer, and 2) builds in word-of-mouth growth that serves as an endorsement for the product (ex: if my friend says I might like this product and they’re already using it, I’m much more likely to use it myself).

Then once you find a few of them via word of mouth, think of other places that that specific user will be and go there.

This is where competitive advantage is made. It’s always more advantageous to build a product that a FEW people LOVE, than to build a product that THOUSANDS of people merely find interesting.

I think this is especially true with AI / SaaS technology. These days anyone with the right amount of ambition can spin up a general-purpose AI transcription tool.

But building an AI transcription tool that’s built for ONE SPECIFIC TARGET USER (ie. US college professors working on research that require user interviews) has the benefit of vertical development.

Once you solve transcription for them, there are likely some other adjacent pieces of the puzzle that you could solve for them.

How do professors get their interview recordings in the first place? Can you add a feature to make that part easier for them? How do they incorporate these recordings into their research? Can you make that part easier for them? How do they summarize their interviews? Could you make that part easier for them?

Sprinkle in a few of these features and pretty soon you have an extremely defensible product, because other general-purpose transcription tools won’t want to compete on niche features like that.

The most successful AI / SaaS products will all begin to look this way over the coming months and years. Extremely niche products built for one specific type of user. Ideally a user with 1) a burning need, that they are 2) willing to pay money for.

Bottom line is: find an ideal user who’s willing to pay you money, bear hug them, and forget about everyone else.

In the marathon race to the finish line of sustainable revenue, it’s easy to get distracted. The winners are the ones who can stay focused on their one ultimate customer

Peace,

Ramsey