BizSpark is Not Good Enough to Depose LAMP

by Krishna on March 16, 2009

Microsoft has a program for startups called “BizSpark” that offers Microsoft development and production licenses with the following terms: (via StackOverflow blog)

  • No fees to join
  • Access to as many production Microsoft product licenses as you need, for free, for three years
  • MSDN Professional full access subscription to download the software

At the end of three years

  1. you have to pay $100
  2. you have to license the software you’re actually using

The startup must

  • Be in the business of software development
  • Be privately held
  • Have been in business for less than 3 years
  • Have less than US $1 million in annual revenue.

While this is a forward-looking step, I don’t think BizSpark goes far enough. People are fond of arguing about the relative strengths of a particular technology, but in the case of adoption, entry costs play a disproportionate role. While $300 seems very inexpensive, it is not when compared to close to $0 for a LAMP stack which gives you mySQL/postgreSQL, PHP, Perl and Python.

Now, I know all the arguments for justifying spending the $300 for a Microsoft development stack. For example, you can argue that using .NET will increase productivity that will generate savings many times greater than the expenses. The software cost is only a fraction of development costs, especially if the startup is very successful.

Practically, though, these rationales only make sense for startups that are true business concerns. For every application produced by an “official” startup, there are hundreds of small applications created by programmers working alone or in small groups (1–3). These applications may be hobby projects created without much hope of striking it big, but some do. These programmers like to keep their costs minimal because they don’t think they will benefit. Because LAMP is easily available at low cost, that is the default choice for many such developers.

(While $1 in hard cash may be the same conceptually as $1 in programmer effort, they are not the same in practice if you don’t have the $1 or you are not willing to spend it. In real life, some people, to save a dollar, are willing to spend several times the equivalent human effort, because the human effort cannot be employed to earn the same dollar. Coupon cutting is one such example. This is actually more complicated than I make it out to be, probably will discuss this in a future post.)

The second problem is the cost after the three year period. Some startups will be enormously successful and for them, this doesn’t matter. But what if the application is bringing in revenues just keeping pace with salaries of the programmers? That would be a big hit. Now, three years is a reasonable time for a company to succeed, but things can go wrong like, for example, the economy may take a turn for the worse and affect the business, as it is doing now. With a LAMP model, the costs remain same, and there is no sudden jump of expenses at the end of the third year. A better model in this situation would be a revenue or profit sharing arrangement.

I may be ignorant about the details about how this works and would love to be educated otherwise. But if I am correct in my understanding, BizSpark is good, but not good enough. Microsoft is losing a lot of programmers who are building small applications and, while I have enormous respect for the .NET platform, better technology is not going to attract them. You need an even lower barrier to entry. Make it as cheap as LAMP for entry and then have a business arrangement that minimizes financial risk for the startup.

Comments on this entry are closed.

{ 1 trackback }

Previous post:

Next post: