A physics PhD and startup founder explains how entrepreneurs can get the most basic business advice all wrong

• Happy Numbers is an artificial intelligence-enabled math education platform.
• CEO Evgeny Milyutin said the biggest lesson he's learned in his time in startups is the importance of knowing your customer.
• This is pretty common business advice - but he's probably not the only founder who initially went about following it all wrong.
• It took him a while to realize that meant visiting the classrooms where students and teachers would be using his product, talking to them, and watching them use it.

When the iPad debuted in 2010, Evgeny Milyutin was a 26-year-old physics PhD student in France.

On the side, he and his longtime friend Ivan Kolomoets had been tutoring their friends' kids in math. It occurred to them then that there was a prime opportunity to improve the quality of math education with emerging digital technologies.

At the time, this was a truly novel idea: It was still seven years before Netflix CEO Reed Hastings would invest in a math education startup called DreamBox Learning.

Today, Milyutin and Kolomoets are the founders of Happy Numbers, an artificial intelligence-enabled math education platform. The goal is to help teachers personalize education: Milyutin described the program as a "virtual teaching assistant."

Students are set up with iPads or laptops and plug away at interactive math exercises; then the program delivers feedback to the teacher based on the students' performance.

Milyutin - who confesses that he struggled with math in primary school - cited a 1984 review by the late Benjamin Bloom, which reports that students who received one-on-one tutoring performed better than 98% of students taught in a conventional classroom.

"It would be great to have one teacher for every student, but it's not always realistic," Milyutin said. "So this is where I feel technology can come into the game."

Individual schools or school districts can purchase subscriptions to Happy Numbers (though Milyutin said he's also sold a few subscriptions directly to consumers). In the last year, Milyutin said, there have been 17 million exercises solved on Happy Numbers.

Milyutin and Kolomoets skipped a crucial step before they started prototyping

The years between 2010, when the idea for Happy Numbers first struck the founders, and 2014, when the company officially launched, were hardly glamorous. Milyutin said they created one prototype after another, but "didn't get much luck."

For startup founders, it's almost inevitable that their product will have to go through multiple iterations before it's ready to market. And this wasn't Milyutin's first entrepreneurial rodeo: He'd already developed a biomedical sensor for identifying drugs in blood samples and impurities in drinking water, as Business Insider previously reported. (He subsequently sold the patent to a German sensor company.)

But looking back, Milyutin can pinpoint one reason why he and Kolomoets struggled, at least in the very beginning: They were building prototypes before they fully knew the students and teachers who made up their core user base.

"Everyone talks about 'you need to understand your customer' or 'you need to understand your user,'" Milyutin said. But it took him a while to realize that meant visiting the classrooms where students and teachers would be using his product, talking to them, and watching them use it.

"Going to the actual classroom gave me so much inspiration and ideas of where the product should go." Specifically, they heard from teachers that their goal was to individualize instruction.

Milyutin advises other entrepreneurs to spend more time than they think they need to on understanding their customer

Milyutin has some insight into why he and Kolomoets skipped this step at first. "You kind of feel you're bothering people," he said, referring to his requests to sit in the back of classrooms and ask the teachers questions afterward. Eventually, he learned this wasn't as much of an inconvenience as he'd thought - and that it would be impossible to start his company without making these requests.

What's more, Milyutin said, he'd sometimes get overeager, assuming that after he'd spoken to a teacher over the phone for two hours, he knew everything about their experience. "Talking to more people" - as many as you can - "is always better," he said.

Those people included experienced education researchers. Milyutin said he would read blogs and books on education - and if he liked what he'd read, he'd get in touch with the author to learn more.

Milyutin's single best piece of advice for other small-business owners is to invest time (probably more time than you think you need to) in your customer.

In an email to Business Insider, he added: "Understand exactly what their day looks like and what keeps them up at night. A true understanding of all their pain points, even if they are not directly related to the problem your product is solving, will help you develop a solution that customers will trust."