I'm going to be honest with you. I have written about 50 full-featured websites, and when I mean full-featured, I mean some of them have terabytes of data, full-featured control panels, and back-tested portfolio strategies. I put insane amounts of effort into some of my sites.
And none of that is the reason that the two sites that worked made it.
One of my sites, I just bought a domain name that was for sale and had a lot of backlinks. And then when I built the new site, I made sure it did the exact same thing that those people who were visiting the site would expect. Apparently, this is pretty important because Google will just drop the site on the floor if it doesn't do roughly the same thing. It's about matching search intent.
And it immediately started generating $200 a day without even me trying very hard.
I then proceeded to build a whole bunch of useless things that nobody used, even though they worked very well for me and my friends. Because I hadn't learned my lesson.
I spent years getting better at building websites when what I actually needed to get better at was getting people to visit them.
The lesson is that distribution is absolutely everything. And grabbing that shortcut to distribution changes the economics of your startup so quickly.
My second site did well because I joined 150 Facebook groups that focused on real estate. And I used that to promote one of my data-gathering sites. I had been building that thing for probably 6 months now, and I had two paid users. It was an absolutely insane ratio of effort to reward. Now I have title companies calling me and asking me how to integrate because I spent 2 hours every night for the last 6 weeks posting on Facebook groups.
The first time, I bought distribution. The second time, I built it by hand. In both cases, the website itself was the easy part.
The only thing that matters is figuring out how to get your site out there in front of people without spending enormous amounts of money and without getting blocked for spam on all of the different social platforms.
If you can figure that out, then it almost doesn't matter what you build. Think about it. People make money filming themselves talking to cell phones.
More than ever, distribution beats everything else.