Sunday, May 18, 2008

Precisely wrong

At the Rice Alliance Startup Workshop this weekend I listened to Corey Blahuta talk about financial issues for startups. When talking about a company's initial projections, he pointed out that the first draft is surely "precisely wrong."

These projections usually take the form of a big Excel spreadsheet, although I've taken to doing them in Google Docs so that everyone can work off of the same version of the spreadsheet and I can feed in live data from my website.

For a B2C Internet business, the model starts with marketing and customer acquisition and then flows through website conversion, user adoption, user activity, advertising inventory, advertising sell-through, collections, fixed and variable costs, and finally revenue and profit.

For a B2B business, the model usually starts with salespeople, who make a certain number of phone calls, to get a certain number of meetings, that results in contracts being sent out, that results in closed deals, that results in collected revenue, that results in expenses, that results in profit.

The goal is to show how you put in money on one end (by hiring salespeople or spending it on marketing) and then more money comes out the other end (after expenses).

The initial projections are based on many assumptions that have not been validated, so of course they won't be exactly right. Often, the initial projections are orders of magnitude off of the actual performance.

That doesn't mean its not worth making projections. One of the audience members asked why you would even make the projections if you know they are not right, suggesting that it was a waste of time.

The point of your initial projections is not to try and estimate exactly how the company will perform over the next 3 years and what the costs will be. The point of the projections is to develop a mental model for thinking about the business and tools for discussing different scenarios and tradeoffs. By creating a model for the business you have something to poke at and talk about with your team and advisors.

Any good model is based on a list of assumptions, and at the start of your business most of those assumptions will feel like throwing darts with a blindfold on - not very accurate. But by listing them out, you can see the relative impact of one versus another. And as you learn more, you can update the assumptions to be more accurate.

Another use of the model is to compare your business to others that are similar. Say you are trying to figure out your customer acquisition cost. You should estimate what you think it will cost based on advertising rates and conversion rates. But then you should compare your customer acquisition cost to what you can find out about the customer acquisition cost of companies that are similar to yours. Chances are you will have a similar cost structure to them. If your number is way less than theirs, you might want to check if there is anything you have overlooked.

Investors and advisors who have seen other business know what its supposed to look like. They can help you identify hidden costs that you have missed or other revenue opportunities you are not considering.

Creating a model isn't just something that should be done when your planning a company or fundraising. You should be constantly updating your assumptions as you go. Within your model, you should identify a few key metrics that you can focus on to optimize your business.

Josh's latest bootstrap startup is

Reposted from Austinpreneur

No comments: