Lack of Data Pushes You Beyond Your Comfort Zone

[There is] immense difficulty getting confidence in predicting how things will develop. It is easy to plug 75% growth into a model, but that's far from it actually happening. With all this uncertainty, my perception is that one would want to be super conservative or have downside protection (ie. low valuations) to make up for this forecasting reality. However, good companies get funding and it can be competitive, so you're forced to push yourself beyond comfort. Balancing the uncertainty risk and a lack of sector expertise with competition on deals is the most difficult thing I have observed thisfar. Damned if you do (higher values than you're comfortable with), damned if you don't (nobody to take your money). This is much more of a problem with late stage investing I'd imagine.

[Related:] valuations are so subjective and small numeric tweaks can prove anything, so keeping that in check is key.

Surprising How Little Data Goes Into a Decision

The single most surprising thing has been recognizing how little data these guys have to go on. I had an inkling that would be the case, but especially coming from a PE background where we diligence the heck out of everything, have tons of historical info from the company, typically have industry studies, 3^rd party experts, etc, seed stage venture guys (which is decidedly where [my firm] sits) have next to nothing to go on. Their feel for the entrepreneur, how the idea sounds, most importantly, reference checks on the entrepreneur and their team… and that’s about it.