Like many others I have been following the flurry of posts and comments that followed Jorg Janke’s blog post about Compiere. Without a doubt Jorg’s post is a great contribution to those of us involved in the day to day business of running open source software companies.
However, most of the subsequent posts and comments focused on a hackneyed ideological debate surrounding open core and open source software in general. Is it just me or is this discussion really only interesting to an insular herd of open source VCs plagued recently by less than stellar returns and the cadre of lawyers that service this sector?
For those of us that spend our days and nights running profitable, non-VC-run open source software companies, there is often little time for such banal ideological chatter. My shareholders are more impressed by numbers (the ones on the bottom line of the income statement), and our company is focused on creating world class software, engaging with a vibrant community to do so, and delivering value day in and day out to customers. With this in mind, who has time for an “open core” debate which completely misses the subtleties of running most successful companies?
Since I seem to have appreciated something different in Jorg’s blog, here are a few of the points I feel are good take aways:
1. Training. I guess this should be obvious, but for some reason it is easy to forget. I immediately cut and pasted some of Jorg’s thoughts and sent them directly to our head of sales. “… people tend to like the application they had most exposure to (and understood best).” This just drives home the point that we are already seeing. Those that take our BPM Software training course tend to continue on and purchase other services related to our BPM and Workflow Software. Jorg’s 80% up-sell number is interesting, especially for us since we are operating in such a similar space (BPM Software – Enterprise Software). Enterprise software tends to have a longer learning curve and so you have to figure out a way to get users over the “suck factor” to use the words of another open source exec/friend. I know we know this, but it never hurts to hear it again and go back to our strategy and see if we are really executing on it. Are our sales people taking the leads from trainings and performing the appropriate follow up? Are they being placed in the appropriate lead nurture campaigns? Are we focusing enough resources on training?
2. Partners. You either love ‘em or you hate ‘em. Or in the case of Open Source Software – both! If there was ever a difficult subject in the world of open source, it is the subject of partners. Why do no open source companies admit this? I’ve been drilling open source execs on this point for years – how much of your revenue comes through partners and how much comes directly? Why is it I always get the same answer – “well, we actually don’t get that much revenue from our partners, but they are a really important part of our strategy,” most will say. I spoke with a partner of one open source ERP vendor the other day about partnering with us, and the partner asked me if we were going to force them to sell our Enterprise Version. He told me that his Open Source ERP vendor forces him to do this, but they usually don’t – instead they usually install the Open Source version, charge an installation, and charge the client their own support. No secret there, right? Wrong – this company was listed as a Gold partner for this particular vendor! So do all Open Source companies just suffer from SAP and MS channel envy? Is it really worth the uphill battle?
3. Enterprise functionality and the balance of power – This is by far the most important point for me, and I believe for almost any open source company. And it is the point that Simon Phipps seems to over simplify while trying to draw ideological lines. The fact is not whether Open Core is good or bad. This type of ideological fanaticism serves no useful purpose. The truth is that “some” enterprise versions are bad and others are not. That’s right – no simple rule. I wish there was one, but there isn’t. The fact is that the Open Core needs to deliver enough value to make sense to enough people. If an open source company “cripples” its open source core, the community will soon head for greener pastures. And the fact is that the successful companies using open core are managing not to cripple their community editions.
As Mark Radcliffe points out how could open core be doing so much damage to the community version if only 10% of an open source company’s users are using the Enterprise version and the other 90% are using the open source? Good point (although I think this number is actually much less in almost all cases).
So here is an idea. Instead of worrying about approved licenses and what makes a company truly open or not, let’s just look at the numbers. If more than 10% of a “supposed” open source company’s users are actually paying money and using the software under a commercial license, then that company should no longer be called an Open Source Company. That’s right – they will be banished from this exclusive community. If less than 10% of your users are paying you money, then no worries – you will still have the right to be a card carrying member of the open source software company community.
So there you have it, if you are an open source software company enjoying revenues in excess of $80 million USD per year – you better watch out because the OSI police are coming to get you because you must be cheating!
