Paul Jacobson

There are ways to forecast for startups and ways not to do it.  Guy Kawasaki has a few really good and sensible suggestions:

1. Under-promise and over-deliver;



2. Forecast from the bottom up;



3. Don't go too far out (12 to 18 months should be the maximum);



4. Plan to re-forecast every 3 months (I like his statement that long term for a startup is one year);



5. Don't let costs get in front of revenue;



6. Collaborate with your investors;



7. Think in terms of per-unit profitability;



8. Plan for marketing costs;



9. Create a 1 page report and stick to it;



10. Never miss a cost projection; and



11. Think big.

The problem with new startups (especially new tech startups) is the tendency to adopt the same mindset that plagued the late 1990s.  One of the most important lessons to come out of the dot-com bust is that even cool Web-based companies must have a real business model that addresses real needs and doesn't depend on sans serif fonts to succeed.

If you are starting a really cool, new business on the Web, be pragmatic and be sensible when you plan your imminent success.