On the Business of Software forum, someone recently asked: how can software vendors compete with open source software?  Fog Creek is no stranger to that question.  FogBugz started as a bug tracker, and there are plenty of open source and proprietary competitors in that field.  And Fog Creek Copilot, our remote desktop service, incorporates software from the open source VNC project, and competes with various flavors of VNC and other open source and proprietary alternatives.

Our products compete against lots of other products - free and not, proprietary and open source.  To compete against all of them, we focus on the true, underlying interests of prospective customers.  (Yes, we management trainees just discussed "Getting To Yes" at work.)  I'm going to talk about how FogBugz and Fog Creek Copilot compete.

People consider using open source products for a few reasons. One possible customer interest is "saving money so I can look good to the boss."  In that case, we might talk about total cost of use and ownership.  We can compete there, since setup and support are so quick and easy.  And if the customer wants to see the source code and modify the look-and-feel, for control and reassurance, FogBugz has that too, since you can view and change the source code and the database schema.

But if a core desire for the customer is "using open source for ideological reasons" or "using in-house resources to massively customize all applications," then FogBugz is basically not going to meet that interest.  It's not for them.

You've probably also heard about other ways to compete with open source software: better design and user interface, brand reputation, customer service, features, ease of setup, and so on.  And we absolutely do compete on those factors, for example against Bugzilla, and also against quite a few commercial bug trackers.  But they aren't ends in themselves.  They all serve underlying customer interests.  A few examples:

• Avoiding hassle: Using Bugzilla is like interacting directly with a database using a web form; the user has to learn to behave like a computer.  FogBugz sets up in minutes and treats users as humans, so it takes practically no time to learn.   Everything just works.  That way, customers can do what they're good at, instead of wasting time configuring a tool that's supposed to help.  The design, and our quick, smart support, reassures customers and saves them time.
• Painless migration: People have fled from Bugzilla to FogBugz so many times that we actually wrote a special importer for Bugzilla databases.  We get emails from users literally begging us to help them convert their Bugzilla database to a FogBugz database.  Now every instance of FogBugz ships with this script.  So that's another way we compete: we make it very easy to switch. This saves customers time and worry.
• Vendor responsiveness to customer requests: We have a built-in screenshot tool so testers can submit screen captures as bugs in just a few clicks.  We did this because customers asked for an easy way to attach screenshots, and we took that one step further to make the experience seamless and intuitive.  Customers get better bug reports, improving their software quality.  This works right out of the box and Bugzilla doesn't have anything like it.

As a service, Fog Creek Copilot has a different kind of competitive advantage.  It works through firewalls because we host a reflector.  We manage the server and pay for the bandwidth.  So it doesn't matter that we provide the client applications' source code under the GNU Public License.  Any copycat service would also have to provide a reflector (which costs money) or give up the core customer need of ease of use (because it wouldn't work through firewalls).  The partially hosted architecture provides some shielding against open source competition, but only because it fulfills a key customer interest.

Don't forget that competition intensifies over time.  Your competitors will watch your software improve, and copy it, and there will come a time when you can't make money off it anymore.  Open source developers are great at cloning.   Good software takes ten years but then it's done, and you will have copycats every step of the way -- open source and closed source.  In the long term, we hope that customer service, brand reputation, and other stuff outside the actual codebase will give us an advantage others can't duplicate.