by Jon «maddog» Hall, president of non-profit organisation Linux International and global GNU/Linux pioneer. The article is reproduced with consent of mr Hall as well as LPI director of communications, Scott Lamberton.
I have a friend who I love as a brother. The other day they wrote me in email telling me that he was starting to do home brew beer, and he and his brother were going to invest in this “as a business”.
While I too love home brewing, I recognized his syndrome, and gave him the best advice I could:
Enjoy it as a hobby, but before you invest in it as a business, DO THE MATH.
I knew that he already had a successful company and several hobbies as well as a wife and family that make demands of his time, so I wrote back asking him exactly what he had in mind with the beer, as I have known many, many “home brew” enterprises that have failed as a business.
When I say “DO THE MATH”, I mean to decide what your goals are for starting the business, how you are going to start the business and grow it, and what you feel would be “success” in the business, which may or may not include money to pay you for your time spent.
What my friend is experiencing now is the joy of taking raw materials and on a very small basis creating new formulas for beer, designing and applying the labels onto the bottles and holding them up for inspection and appreciation by others. Thus he may joyfully spend ten or fifteen hours of work to make a batch of fifty bottles of beer. Run this way, his business does not “scale” very well, and by the time he made enough money by doing this to support his wife, two children and his other hobbies, he would either have to find time in the fifth dimension or change drastically the way he was producing his beer, which might reduce the pleasure he got from doing it.
The same is true of Free Software. So many times I have seen people who want to make money “doing Free Software”. They like working on Free Software, doing their individual “hacks” and contributions, and feel that they want to do this “full time”. Some take a few odd jobs, get paid to do some work for “friends and family” or local small businesses, and want to take the plunge of leaving their full-time programming job to join the world of full time Free Software support. For these people I have my best advice (which is the same that I gave my friend for brewing beer), DO THE MATH.
What they like about Free Software is the artistic creation. They are typically not as enamored with creation documentation, supplying long-term support, and all the other “business issues” of running a business.
I know that business plans are “boring” for a lot of technical people, but doing a business plan (and its close cousin, the marketing plan) forces you to think about how much business revenue you will need to sustain you in whatever endeavor you try using Free Software. It will force you to think about the fact that you typically can not work 366 days a year, 25 hours a day and that you need to think about how to attract new customers at the same time that you are doing work for old customers, and (as an example) you may want time to read, absorb and try out new technologies. Build that time (nominally called “training”) into your business plan.
A friend of mine in Germany once complained that he had so much work to do that he did not have time any more to “investigate neat stuff”. He was the President of his company and had about five consultants working for him. “How much do you charge per day for your time, and how much do you charge for your consultants?”, I asked. He told me that he charged 1000 US dollars a day for his consultants and for him. “This is terrible business”, I said, “You should charge 5000 US dollars a day for your time, and keep your consultants at 1000 dollars a day.” “But I would get a lot less work at $5000 a day”, he said. He might get a lot less work at 5000 dollars a day, but (in effect) that was what he wanted, and at 5000 dollars a day he would only have to work one day a week to earn as much as he earned at 1000 a day, leaving four days for “investigating neat stuff”. Plus, the lower-priced work could still be funneled off to competent people under him. In reality he might get a lot more work offered to him at 5000 dollars a day, since people might appreciate the value of what he was bringing to the table. And if he lost too much work, or a good customer, he could always discount his time, but going in with too low a bid and trying to raise prices afterwards is hard.
DO THE MATH. Technical people typically tend to underestimate the cost of marketing and sales, particularly in medium to large business proposals. While enough business for a one person company may come through “word of mouth”, you may find that the costs of growing the business past “one or two employees” needs an order of magnitude greater revenue to support it. If you wish to maintain a “local” business, does your local market have enough business to support you?
A friend of mine in Thailand pointed out that one of the great reasons various “projects” do not work in rural areas is that the cost of the maintenance of the project is greater than the total amount of revenue that the entire village makes in a year. No village can afford to spend 100% of their revenue keeping their computer systems alive, unless you can guarantee that the computers generate more money for the village then the computers cost to run.
In many places a small “service” company has one set of infrastructure and legal issues to overcome, and as they grow they find more and more complexity and cost in meeting these same issues.
This is why I encourage technical people to read about “entrepreneurship”, and to learn some of the issues of starting and growing a real company instead of thinking of it as “just charging for giving support to Free Software.” Without “DOING THE MATH” what might start off with you really enjoying the act of creating Free Software and dreaming of earning a living writing Free Software could turn into a nightmare of trying to sustain your business over time.