I’ve been ramping up my startup research over the past 6 months and have noticed a startling trend.

Well, a few startling trends.

With the advances in AI, it’s never been easier to launch a startup.

Barriers to launching a startup, like building a website or writing ad copy that used to take weeks can now be done in minutes.

For people who like to experiment and test startup ideas, the delivery of tools like chatGPT have been a godsend.

Despite these tools, many startup entrepreneurs I speak with are missing the boat completely on how to make money.

Most of the ideas I hear are some variation of - I want to take business X and automate it with chatGPT. Founders tell me that once they automate business X, people will line up out the door to pay for it.

Some founders even tell me that if they had the money to pay an engineer to build it, they’d be able to prove it immediately.

These ideas couldn’t be further from the truth…

The Actual Truth

Let's start with what they get right.

Yes, there are a lot of opportunities to use ChatGPT to automate tasks that used to take time. And yes, someone might actually pay for it.

However, there are at least two major mistakes with this line of thinking.

Mistake #1: Founders believe they need money to launch their startup.

Although this used to be the case, AI and no-code tools like Bubble have completely changed the math.

Just last week I used ChatGPT to write code and build a bot that scrapes data every day from a government database.

Two years ago, this would have been impossible to do without a software developer. Now, anyone with an internet connection, time, and focus can build this in a couple of hours.

Hiring a developer has become one of the least important things for new startups today.

Mistake #2: Founders believe that once the product is built, there will immediately be buyers.

This is where I see a large majority of startup founders get tripped up.

They are caught in what I call the ‘Field of Dreams’ trap… and think “If you build it, they will come.”

While this might have been true once upon a time, it's actually very far from the case today.

Too often, first-time early-stage founders will invest loads of time and money into building a product, believing that once it goes live it’ll be an instant success.

Unfortunately, I’ve seen lots of these founders pour their lives and savings into building a product only to launch and have nothing but crickets.

Because it has become so easy to build a product, there’s never been more competition for customers.

Although building the technology used to be the hardest part of launching a startup, this is no longer the case.

In reality, the hardest part of launching a startup today is being able to get customers cheaply.

Cheap is key here because anyone can spend a small fortune and get customers.

The best companies today are able to beat the competition by paying less for their customers.

Think about companies like Tesla or Apple. These are two examples of big companies that would be successful without paying a cent for marketing.

This allows them to run circles around their competitors that have to pay handsomely for every single new customer.

The money they save on advertising can be spent on other important things, like paying top talent to help them build the best products - or in Tesla’s case - offer a great product more affordably than any of the competition.

Putting it All Together

So what’s the takeaway for aspiring startup founders?

Ultimately, startup success does not require a team of developers or loads of cash (although that certainly helps).

The most important skill for today’s startup founders is to figure out how to attract and sell to customers.

In some ways, tools like AI make this easier than ever.

The scraping tool I described earlier is part of a test I’m running to perform cold outreach to potential customers at the lowest cost possible.

Even before I build the product, I am starting by testing whether I can get customers cheaply enough to make a profit.

With ChatGPT and a drag-and-drop tool to build websites, I was able to punch out a homepage in an hour. I’m planning to test the website by building a fake checkout button and seeing how many customers try to make a purchase of a product that does not yet exist.

If the test fails, I’m going to move on to the next experiment.

I’ll save a fortune by testing if the product will sell before building it, and maybe I’ll learn something in the process.

This experiment would have been far more difficult and time-consuming in the years before ChatGPT.

Ultimately, I believe the people who will make the most money off of AI will not necessarily be people building nifty new AI tools that save time. These folks might do well, but like all hot new trends, many will fail.

The better alternative for startup founders today is to focus on figuring out how to use AI to build customer lists cheaper than anyone else.

In today’s crowded market, the ability to get customers at a low cost has become the most essential skill for building a successful business.

Founders that focus on using AI to gain an advantage in sales and marketing will run circles around their competitors.

*Note: If you’re trying to build a startup doing something super high-tech or complex, ChatGPT is probably not going to be able to build the application you need. However, most startups aren’t going to be building these types of products anyway.