I'm in the process of working out Support for a small Product. I'm using GetSatisfaction, and although I'm concerned about it's speed of use, for this very small Product it's more about learning the methods and wotnot involved in Supporting.
Did you build your own Support System software? If YES, what did you use to build it?
If you are talking about a support product (bug tracking/knowledge base/etc) I would recommend buying one. The amount of time it would take to build one would be better spent on creating a product you can make money on. There are plenty of free/cheap products out there to do this for you.
I'd highly recommend against building your own support system software, otherwise you'll end up having to support two systems. If customers know you've built you'll find they start asking you to make changes to the system. Definitely look for an off the shelf solution and customise it to fit your needs.
We built our own. It started in Access forms, then ASP w/Access, then ASP with SQL Server, then ASP.NET with SQL Server. It handled all issues, all departments, and had significant history, dashboards and tracking. It had significant custom needs for our line of business. We had a high volume helpdesk with high level of service which required callbacks within a certain period. The system was tied into our web timeclock so that notifications of high priority support issues would go out to people who were in the office, not just everyone in the department. The system was tied into the general phone operators who could see open tickets and route calls to appropriate departments and would also tie into the customer's business performance data.
If you can buy something off the shelf, go for it. Like anything, the more demanding you are and less able to change to fit the software, the less you will find off the shelf software meeting your business requirements.
We built our own, and frankly, it lacks in many of the basic features that the design team didn't foresee. I'm currently in the process of trying to fix those issues, but it's making the project far too bloated. Sadly, I wasn't involved in the original, but if I was... I'd have pushed hard to buy or even use opensource alternatives.
Don't write something you don't have to. There are solid support systems out there, and even some opensource examples that can easily be built upon. At the very least, use one of those for the base and simply expand upon it as the need arises.
Thats my 2 cents. But I'm in a very small programming department, so my perception might be skewed into requiring a ROI on about anything I load into my IDE.