Home Page  Research  Teaching  Economics of IT  For Startups  Calculus

If someone advising you on business planning tells you that "financials don't matter for an Internet start-up", start looking for new advice. Granted, five-year projections are excessive, but a financial plan is crucial. Any nascent company should have a simple, well-articulated idea of its revenue and profitability potential, and of its near-term cash needs. The numbers probably won't pan out exactly, but the modeling process gets the company thinking about the economic viability of its concept.

Creating a Simple Revenue Model
Start with a few parameters, for each period you are projecting revenues for:
  • One or more relevant and reliable industry numbers, that will lead to your final demand figure. For instance, the projected total number of online shoppers, or the projected sales of computers online, or the projected total advertising spending on non-banner Internet advertising. A couple of useful sources for such projections are the Industry Standard Metrics page, and Forrester Research (if you dig into the site, there are a lot of free projections available).
  • A set of simple conversion factors that translate these industry number(s) into demand figures for your products. Typically, these takes the form of penetration rates, clickthru rates, and/or conversion rates. It is customary for successful Internet businesses to see penetration rates grow non-linearly. These are definitely not easy numbers to estimate. Try drawing parallels with existing offline businesses or analogous online businesses. You could also try some inexpensive market research online, or survey a small set of target customers.
  • A simple pricing scheme and profit margin. It is usually wise to start with competitive pricing and no margins, and increase the projected prices and margins as your projected demand grows. To benchmark your pricing, make sure you either peg your prices to comparable existing products, or talk to a few experts in your industry about potential willingness-to-pay.
Make sure your have the same parameters for each period. Combine these numbers to generate your demand estimate for each product line, and your total revenue estimates. Test the sensitivity of your model using a few scenarios (low, medium, high) for different combinations of the parameters, and figure out which of them affect your revenues the most.

For instance, if you plan to sell research to online traders (as many of my students' business plans do), you could:
  • start with the total number of online traders (industry number), estimate the percentage of traders who would buy your research each year (conversion factor), and what subscription fee you would charge for your research annually (price).
  • start with the total number of online trades (industry number), estimate the percentage of trades that are likely to be preceded by the need for a research report (conversion factor), the percentage of trades you would have access to (conversion factor), and the price per report (price).
  • start with the total projected online brokerage fees (industry number), estimate research spending as a percentage of total brokerage fees (conversion factor), and estimate what percentage of this you could capture (conversion factor).
This gets you started. Don't get much more complicated than this. Its better to have a few well-researched and validated assumptions in a simple model, rather than a sophisticated model that makes lots of assumptions that are difficult to substantiate. Try and isolate the most pertinent factors and numbers, depending on what your business model is, what you feel really drives demand in your market, and how easy it is to estimate the relevant conversion factors.

Copyright © 1999, 2000 Arun Sundararajan. All rights reserved.