Startups: build something that people will actually use will require data from those people who are in the problem space. It is often in a niche, which means you can only get it by doing customer interviews or observe their behaviors by yourself.
So now if you went to college I want you to forget everything your marketing professor has taught you. Unless you have a professor who actually did something in the real world, which is rare otherwise they won't be a professor.
When you start an idea, you need both first-hand research(customer interviews, observe users, etc.) and second-hand research(white papers, industry data, etc).
Here are the steps to do your research:
Too many people only rely on second-hand research, that is not going to work. If by searching up numbers can solve a problem, then it must be a problem that's already solved. Any problem that has not been solved is often in a niche. You have to collect that data on your own.
I gave a talk at an accelerator program at Tsinghua University in China, and this guy asked me if his product can offer the time to find the product-market fit. he was trying to do automation in the study abroad industry. One thing to notice, the traditional business philosophy is build first approach. He then mentioned he couldn't sign any deals before he has the product. So I replied I don't know your industry enough to have a say about it but most likely yes. He said no, his product is too complicated to "start slow"(referring to the process of finding product market fit). So I asked him how does he know if people need his product? Because as an international student, I have been in contact with those agencies when I was in high school. I don't like automation, I want personal services, long story short I already hate the idea myself as a person in that problem space. But I thought that's ok, I cannot represent everyone right. He then pulled out some industry data that basically says there are more and more people every year choose to study abroad. He was very convinced that was enough evidence his idea is legit. I then asked did you talk to anyone? He then said yes, hundreds of them! I was like we are getting somewhere! I asked him what did the data say? He jammed.
It ended badly, I dropped out of the conversation because I couldn't take it. It was like looking at the 17 years old me but a stubborn one.
This is not rare, so many people are defending their build-first approach and rely only on second-hand researches. Good luck with that! I don't know how they are in 2020 but got a mind from 1920.
Your problem is your problem, no one should understand it better than you do. DO THE WORK, GO GETHER THE DATA BY YOURSELF. DON'T BE A LAZY ASS!