VC in DC : Selling Free Software: Thoughts on Open Source Economics

Sun	Mon	Tue	Wed	Thu	Fri	Sat
			1	2	3	4
5	6	7	8	9	10	11
12	13	14	15	16	17	18
19	20	21	22	23	24	25
26	27	28	29	30	31

Selling Free Software: Thoughts on Open Source Economics

A recent column by CNet's Matt Asay reminded me of the power of one the fundamental rules of software sales, the '1 in 20 rule'. Almost 20 years ago now, I was running an email software company with a few other guys. Our challenge was that our advertising and marketing efforts would prompt users to call us via a 1-800 number to request a free 30 day software trial. We would respond by shipping diskettes and inevitably CD's to the caller. The caller would presumably try the software and we would do our level best to manage the caller's software trial to adoption and sale.

That process brought us to complete awareness of the '1 in 20' rule. We found that of those that requested the trial software, only 1 in 5 would install it. Of those that installed the software, only 1 in 4 would buy it. The '1 in 20' rule says only 1 of the 20 people that request a software trial or nowadays, download a software trial product, will buy it eventually. In working to drive sales, we discovered that of the two operative ratios, 1/5 install and 1/4 of installers buy, we could only effectively influence the former. We were never able to change the ratio of buyers amongst those that installed the software. We could, however, change the number of people who installed the software after they received it.

We started overnight shipping the software to strike while the iron was still hot, we offered free T-shirts to successful installers, and we called within hours of the software delivery to offer help with the installation process. We even began shipping the trial software in a triangular shaped box to prevent it from being buried in a desktop stack. Ultimately, we were able to increase installations to 1 in 4 and sometimes even a little better.

Fast forward almost 20 years and we see the ratios playing out in open source software

Image via CrunchBase

with a twist. For most of the prevalent open source software economic models, the premise is the core software is free. Perhaps, they may follow the "open core" model followed by Zenoss where advanced functionality is available as an upgrade but the fully functional core is free. But in any case, the download is free and the model is based upon some set of heavy users buying support and/or additional functionality. The data says that the ultimate conversion rate in open source software is 1% will buy something. That twist downward makes sense to me as only the heaviest users will buy. For plenty of folks, the core functionality is fine and support isn't required. For the company, the savings associated with not selling to people who will never buy are material.

So, the '1 in 20' rule is now '1 in 100' for open source software. Which sounds bad until one considers the raw download numbers of open source projects from a popular site like Sourceforge which maintains open source projects and provides downloads. Sourceforge's 2007 downloads totaled 800 million. Expectations for 2009 are in excess of 1 Billion downloads. The law of large numbers?!?

The aforementioned article is linked here. The author asks the question about why so many Firefox downloads are uninstalled. I suspect the answer lies in the time tested ratios above.