Figuring out someone else’s model

Here's what I *generally* do as a 2nd year VP (who still gets tasked with building models occasionally...).

I tend to only give the assumptions / input tabs a quick glance to familiarize myself with the different assumptions, but I don't really start there. I think that starting with the assumptions when I don't really know what drives the actual business / P&L is a waste of time and has no context. So I'll start at revenue. What drives revenue? Then I work backwards to figure out the operating build, the key drivers of revenue / unit economics, and what increases / decreases revenue. This way I now have some sort of foundation knowledge to work from.

I'll then go through costs in a similar fashion. While I am doing this, I create my own tab with a high level "output" where I build my own P&L to make sure what's actually happening in the model is correct as I flex different assumptions and toggles, and also making sure everything is summing and totaling up correctly. I've found creating my own scratch work tab extremely helpful.

Once I've gotten a good sense of the P&L and the build to EBITDA, I'll generally focus on capex and the capital structure, depending on the business. Another high priority is making sure the returns capture the correct cash inflows/outflows. I've seen plenty of times where the initial equity investment number is incorrect / earn-outs aren't captured / debt not flowing properly / terminal value calculation is wrong / etc.