To begin with, among the worst financial plans are those that you get if you take a template from a business plan competition or a bank in Germany and don't customize it to your particular business. These templates are usually very detailed on the costs side, listing everything from magazine subscriptions to stationary and postage, but the revenue projection is just one line – a pure estimate that is coming out of nowhere. Parkinson's Law of Triviality comes to mind!
The best financial plans of early-stage Internet startups in my opinion:
- are relatively simple – just one Excel tab or a few at most (a later-stage company will often require a more complex plan but in the beginning you can keep it simple)
- are based on the key drivers of your business (your conversion funnel, your projected ARPU, churn etc.)
- make your assumptions transparent and easy to change
- contain very few hard-coded numbers which would make the plan hard to revise (an exception to this are historic numbers, of course)
- avoid Parkinson's Law of Triviality – spend more effort on what really matters and lump together stuff like tiny expense categories
- contain a few extra lines for sanity checks (anyone who will seriously review your plan will perform them anyway, so why not make their lives a little easier?)