Earlier quoted context omitted.
He could have kept his day job and spent a few hundred dollars or less to have someone make some fake graphs. Hell, he could have made some fake graphs over a few weekends. Then he could have done exactly the same market research.
My one question there would be was he certain he could figure out the statistics to get to what he wanted? I feel like you need to get a rough idea of that much at least before you start pitching. Obviously you don't need every bell and whistle, all the data pipelines and so on. But basic proof would still be good to make sure you can deliver something vaguely in tune with what you're promising. But dropping all that…
I disagree. The latter is a huge investment. You're not burning any bridges by asking if something would be useful even if you don't know if you can do it yet.
Of course even this is a wrong framing. Ideally he'd actually have sold the idea, by which I mean have actually taken money for the idea, before building it.
Then, if he couldn't build it, he could give the money back.