Mike Cannon-Brookes has written about the high price == stable company fallacy, answering the calls to make his products more expensive.
There are a lot of good points there.
I have always been amazed at the psychology of clients. For example, take the J2EE application server space.
We pretty quickly got to the point where choices were:
- “Free”: JBoss / open source [note: open source certainly isn't free as in beer...]
- Expensive / High end: BEA / IBM / Oracle / …
- Medium: Resin, Lutris, JRun, and the like
I saw a lot of clients dismissing the medium camp as they saw things as black and white:
“We are doing large enterprise stuff, so surely we need BEA/IBM”
“We can ‘get away’ with the free JBoss/Tomcat”
So, I always felt bad for the middle tier which had compelling products for a lower cost, but had to always justify this, and also SELL MORE.
April 12th, 2005 at 5:55 am
Dion,
There is a big difference between middleware and applications.
Working for a large company, I know how complicated rolling out any new middleware is.
It requires agreement from security, production, infrastructure and the development teams.
It requires integrating into the existing monitoring solution, training the production folks in supporting it etc.
This means that :
1. You want to be sure the company will be there tomorrow
2. The cost of the license is not the dominating factor
When it comes to end user apps, you can afford to take a risk, and bet on the smaller players as the cost of making a mistake is much lower, and the license cost dominates.
That is why Atlassian have a business model that is attractive to large organizations, where as mid-range application server vendors don’t.