Flux Workflow Surviving Despite Open Source

“How does one sell software if every other software in the same space is open source?”

Here’s an interesting question, “How does one sell software if every other software in the same space is open source?” One would think that it would be an impossibility for a small company. However there’s a bright ray of hope, Flux Corp a developer of a Java based workflow engine is actually showing us how to beat the odds.

In the Java world, open source workflow solutions are a dime a dozen. To find a company actually making money in this space would be surprising. It so happens that Flux Corp does make a decent living and is in fact growing. How does it do it? Fortunately, David Sims the founder of Flux Corp gave me a call provided me a glimpse at the solution.

For starters David keeps costs extremely low. Flux Corp’s headquarters is in Billings, Montana (population under 100,000). A typical 4 bedroom home (2,500 sq ft) in Billings costs around $135,000. By comparison, where I live (i.e. the greater D.C. area) the same home costs at least half a million dollars. Which makes one ponder, how in the world can one create a profitable startup in an expensive metropolitan area? Well, I guess the answer is that you let all those wealthy VCs in the same area to foot the bill! The point however if you want to bootstrap, you move. Just ask Mike Giles who moved to western Massachusetts to build Furl (acquired by Looksmart) with his own money.
David avoids using print adverts in print magazines because he hasn’t seen the payoff. He now essentially uses Google Adwords to do his marketing.

The second tip is that to keep in business you have to listen to customers. The primary kind of customer and until recently the only kind of customer had been Java developers. That is, customers who coded to the APIs and used the Flux GUI tool to design and monitor workflows, although it’s true that the IT operations staff will use the Flux web interface to monitor and control workflows. Apparently, there had been little interest standards like workflow languages like BPML or XPDL. Furthermore, BPEL had never seemed to be appropriate. What keeps Flux alive is that it continues to add features that developers are truly concerned about.

Flux started out as one of the first commercial java based job schedulers in the market. Over time commercial competitors have come and gone (i.e. Kronos, Tempo). However, the biggest competitor of course would have been an open source scheduler (i.e. Quartz). When Quartz came along there was a noticeable financial impact. It essentially took away the customers with basic requirements. Flux stays alive by providing features (i.e. priorities, split-joins, concurrency throttles, “job dependencies”, “file dependencies”, and timeouts) that goes beyond its open source competitors.

From the original base of providing the leading java based scheduler in the market, the company has branched into an even more treacherous place, that is the Workflow market. One of the difficulties with the word “workflow” is that too many people have their own idea as to what it is. Anyway, from David Sims’ perspective the workflow product supports orchestrated software tasks like coordination, branching, looping, splits and joins. The BPM product supports humans in the loop allowing for decision points that support vetoes and consensus.

Flux Corp’s main competitors appear to be OakGrove Reactor and Dralasoft (acquired by Verity). In an interesting story, OakGrove Reactor once used the Flux Job Scheduler for its scheduling. It’s biggest open source competitor according to David is OSWorkflow. Which is interesting since I didn’t think that OSWorkflow was one of the better open source workflow products out there. However, it appears that considering Flux’s developer audience, these customers seem to be choosing simplicity over features.

The most interesting take in all this is the peculiar effect of open source competition. It is obviously detrimental to a software business, however it also an interesting side-effect. It raises the barrier of entry for other businesses. For example, say I wanted to get into the Job Scheduling business. I wouldn’t make any money out of it unless my product was better than the open source alternatives (i.e. Quartz). However, would that product be better than an existing product like Flux that has been incrementally improved over several years? In essence, because of the presence of open source in your market, your business will not have the benefit of feeding off the low end.

David Sims’ business survives because of a very interesting observation in economics. That is Christensen’s Law, “the conservation of attractive profits”. Christiansen theory is about the migration of value over time, High margins move up and down the value chain over time. It doesn’t move always in the direction of going up the food chain, it can go the other way too. That is, despite scheduling or workflow becoming commodities, there’s profit to be reaped from being the expert in your little niche.

Republished with permission from the blog of CE Perez.