Planning Will Save a Software Startup Money

One of the best ways to save money is to plan ahead: as simply as possible while being explicit about assumptions, defining what constitutes minimum acceptable forward progress, and identifying what results would justify external investment. For an example see Guy Kawasaki’s blog for

• Oct-1-2007: Financial Models for Underachievers: Two Years of the Real Numbers of a Startup
• Oct-14-2007 Glenn Kelman’s Financial Model
• Generic Model derived from Kelman’s

Also bear in mind Kawasaki’s reason for the post

Glenn’s point isn’t to make a statement one way or the other about Redfin’s business or to even give you a crystal ball for seeing whether you’ll succeed. A model, after all, doesn’t drive demand or serve customers; it only helps you count up the beans if you do. We’re posting this model because its basic structure might help other entrepreneurs who don’t know where to start.

My net net is that it’s vital when you are bootstrapping to have a simple written plan, that way you know what you are changing to either take advantage of opportunities or update your assumptions based on operating experience. A couple of other observations and suggestions.

1. Headcount is the significant cost driver in a software startup
2. Run three sets of numbers:
   • minimum that you expect to achieve; if you don’t hit these then think about folding your tent.
   • nominal case, such that you are as likely to underperform as outperform (break-even cash flow would be a good target here).
   • results that would merit additional investment to exploit the opportunity that you’ve uncovered.
3. The structure of the model is as important as the particular numbers, document your assumptions about values and the key relationships between unknowns.
4. Planning for your team’s time and the estimating the cost in man hours of key transactions or accomplishments (delivering a new version of software, closing a sale) is as important for a small startup as tracking the dollars: the time spent precedes the product and the revenue.
5. Kelman has a few items that are driven by the fact that he has itinerant sales people, most software startups will delay a large sales force build out until they have a repeatable scalable sales model (VC’s can drive you to build a sales force sooner than your knowledge of the market would warrant).

It may be a stretch to think of planning as a cost saving measure, but if you’ve ever hired someone and didn’t have everything else aligned to make them productive, or leased space that you soon wanted out of, it gets easier. At least for me the “measure twice, cut once” approach is mandatory for employees/contractors and office space.