Diligence, done honestly, is mostly the act of killing deals you were excited about. For every opportunity that reaches my members, roughly twenty are reviewed and declined. Here is what the surviving five percent went through.
First, the numbers — not the pitch-deck numbers, the raw ones. Cohort retention, contribution margin by channel, burn multiple, revenue concentration. If a founder will not share these under NDA, the process ends there, and that alone eliminates a third of deals.
Second, the people. Reference calls that go beyond the founder's provided list: former employees, customers who churned, earlier investors who did not follow on. The question is always the same — would you back this person again, with your own money?
Third, the terms. A good company at a bad price is a bad investment. I model what has to be true for the entry valuation to produce a real return at realistic exit multiples, and I assume compression, because assuming expansion is how the last cycle's mistakes were made.
Finally, my own cheque. I do not bring a deal to the circle that I am not investing in myself. This is not a marketing line; it is the mechanism that keeps my incentives honest. A curator with no skin in the game is just a broker with better vocabulary.