We're creating a new type of media and solving the issues consumers face when taking quality photos and video.
Instagram solved the problem of making photos captured on mobile devices attractive by creating filters.
We're solving the problem of failing to get the best angles or failing to capture the context you want to capture and resorting to selfie-sticks, drones, or just accepting that the photos you take are limited to the fixed frame of what is captured by the camera.
We are looking at all sorts of adjacent markets to define our market size, such as Volumetric Video, etc.
I'm trying to think of something like GoPro, and before Action Camera's were a thing, how would they have defined that market size to an investor?
Any suggestions?
All of that said, if you really want to try and do some kind of semi-meaningful estimation, I'd suggest reading How to Measure Anything by Douglas Hubbard, and then think about what data you do have access to, and what nth order effects you can use in a model, and what analogies you can construct.
For example... if you're creating something completely new, for which you really have no way to estimate demand, do something like this: Start with the population of the geographic region where you intend to sell your product/service... let's say it's the US, and the product is potentially of interest to all adults. That's around 327 million people or so.
Now, figure out what existing "thing" is closest conceptually to filling the same niche as your "thing". An example might be when TV was invented, they could have used radio as an analogy. This is not precise, but it doesn't have to be. Find the market penetration percentage for the analogous item, and assume approximately the same percentage of people would want/need what you're selling. Do the math, and call the output your Total Addressable Market[1].
If you can't find something that's a good analogy (or a bad analogy that's not so bad as to be completely unusable) then you could try other ways to get an approximation of the potential demand. It's old-school and potentially expensive, and probably not very accurate, but you could try a survey of the relevant population. You could also do the "landing page only" thing, run highly targeted ads, and see how many "conversions" (sign-up for your launch mailing list, etc.) you get and approximate from that. You could do pre-sells where you actually sell the thing before it's built. If you're confident you can ship, this could be good because it gives you $$$ to work with. But if you spend the $$$ and don't ship, giving out refunds could leave you fucked.
Anyway, there's a lot of stuff out there on this. Jump on Amazon and find a used copy of an old "marketing research" textbook (or get it off a pirate e-book site) and read through that. There is some semi-legitimate stuff you can do, but don't obsess over trying to get "real" numbers. Something reasonable, combined with bullshit, smoke and mirrors, wishful thinking, and bad statistics should do the trick.
[1]: https://en.wikipedia.org/wiki/Total_addressable_market