Hacker News Reader

You probably meant top-down vs bottom-up.

An example of top-down analysis is to go to e.g. Gartner or similar research firms and read that they estimate the ridesharing market to be say $100B.

An example of bottom-up is to say, based on data from this and this article or data source, we estimate there are 1M cars doing ridesharing every year in the US. We will sell each of them a gadget for $10 so our market is $10M if we capture all of it.

Yes, that's exactly what I was trying to get at. Thanks for the clarification!

I'll second this. Wikipedia is a great resource for this stuff.

Also, check out some of the common frameworks like Porter's 5 Forces, SWAT analysis, BCG Growth-Share matrix, and the GE-McKinsey matrix. There are a lot of free educational materials on these. I'd take it slow and steady, and try to focus on orienting your thinking to the questions asked by these frameworks.

Also, consider reading the Form 10Ks for some companies whose products you like. They have a section where they explain their market and their role in it.

Good luck!

Seconding the 10K readings! You don't have to read the entire thing, usually just Part 1. Part 2 and Part 3 still have good stuff, you just have to read a lot more to find the gems.

> One small tip my cofounder learned in business school: I don't remember the exact terminology, but analyses that go from the inside outwards are more respected by investors than ones that go from the outside inwards. Don't take huge figures like "Number of Americans" and multiply by "Percent of Americans that do 'x'", but rather find localized, small figures, and extrapolate those up/outwards.

Not sure if it could be considered the same, but in physics there's a concept of "from first principles" where you obtain something starting from laws which you know are true (within the domain of applicability of a theory). "From first principles" is in contrast to making comparisons for example reasonings like "I know the answer in that other situation, so in my situation of interest I'm going to estimate the answer is X% more".

Calculations from first principles are tipically more robust but harder to do.

According to BayStartUp (they support startups in Germany), they advise you to do both top-down and bottom-up: This way you can validate the results from the extrapolation.