The MBET Experience

The University of Waterloo – Master of Business, Entrepreneurship and Technology Program

The MBET Experience header image 2

Examining the Economics of Software Distribution over the Internet

March 5th, 2007 · No Comments

A couple weeks ago, we were assigned the following reading as part of our marketing course: http://www.firstmonday.org/issues/issue6_12/ilan/index.html. This article examines the assumption that production of software involves low marginal costs.

As someone who has started a software company that is in the process of developing its first product, I found this article fascinating. The traditional assumption in the software business is that once the money has been spent to develop a product, the cost of replicating the product is essentially zero. This implies that there are no variable costs associated with a software product.

The article referenced above challenges this assumption, by arguing that software distributed over the Internet has not been produced until is has been downloaded. For example – if there are no downloads in a given day, production is zero. If there are 100 downloads, 100 units of software have been produced. Further, the author argues that distribution of software is the process of driving traffic to the website where it can be downloaded.

By adopting these views, it is now possible to assign a Cost-of-Goods-Sold (COGS) to a software product. I think that the Internet Distribution Chain is especially useful in calculating the true cost of production and distribution of software. Simplifying the equations used, we end up with:

Cost per Acquisition = Total Cost / Number of Customers

Basically, you can compute the cost based on the total amount spent on driving traffic to a webserver divided by the number of customers acquired through these efforts. For example, suppose I spend $1000 on Google AdWords for 1,000,000 impressions. I get a 2% click through rate, or 20,000 hits. With a download rate of 5%, my product ends up being downloaded by 1,000 potential customers.

Of these 1000 downloads, 5% become customers, for a total of 50 customers. My total cost is $1000, and I got 50 customers. Therefore, I spent $20 to acquire each customer. This $20 is the variable cost of producing and distributing my software! If my product only sells for $19.99, then I might as well close up shop and go get a real job (or work for the government)!!

The key takeaways that I took from this article are:

  1. Sometimes the underlying assumptions of an industry need to be re-evaluated and challenged. This opened my eyes to a new way of looking at delivering web-based software.
  2. When determining how to price software delivered over the Internet, using the Internet Value Chain to estimate costs is extremely helpful. This will ensure that products are priced profitably.

As a final thought, in the calculations used above, if you can figure out a way to drive traffic to your site for zero cost, then the variable costs for distributing your product also go to zero!

Tags: Uncategorized