Last week Derek Singleton of Software Advice published a post on his blog regarding how open source ERP applications have in many ways failed to be as successful as their proprietary counterparts. Although in many ways this is a hackneyed debate, I think several of Derek’s comments merit further discussion. I work for an Open Source BPM and Open Source Workflow Company and we are certainly face the same challenges as Open Source ERP companies, or Open Source DMS or loads of other Open Source Enterprise Application providers with regards to how best to produce, market, and sell our products. So let me comment on two of Derek’s points in particular: 1) the idea that Enterprise Apps are Sold not Bought, and 2) SugarCRM is Fake Open.
Enterprise Applications are Sold not Bought
Without a doubt there is much truth here. Big business is all about relationships between people. Simply put, people do business with people they like and trust, and once they like and trust someone, they will tend to do business with that person again and again. Historically, ERPs have been big ticket, very high touch purchases that define companies for many years to come. As such, this is not the type of application that managers will tend to download and buy. Rather, as Derek points out, companies expect to be wined and dined or at least receive a few visits from sales teams and pre-sales engineers before buying a product. This is all part of the trust building that must happen for a sale to be possible. A buyer needs to get a warm and fuzzy feeling in order to issue a purchase order, and the buyer needs to know that the vendor’s support contract will take care of him after he makes the purchase. The higher priced the item, the more true this is.
Of course, it is true that the Internet has changed much of this dynamic. And it is also true that people are buying big ticket items over the Internet all the time without ever meeting someone face to face. However, software, especially enterprise software, involves support and consulting. And support and consulting involve people. And, it is also worth pointing out that in most cases, these “people” tend to be local (hence the need for a strong partner channel). Deploying ERP Software and BPM Software usually involves a lot of hand holding at the organizational level and most companies don’t want to do this over the phone and via online meetings.
So, what does this discussion of whether software is sold or bought have to do with open source, i.e. software that provides access to the source code and is distributed under an open source license? Well, absolutely nothing. However, it has a lot to do with “traditional open source business models” of driving a high volume of leads through downloads and then helping these leads self select themselves as potential customers. In other words, Derek jumps straight to the equation of
OPEN SOURCE SOFTWARE = OPEN SOURCE BUSINESS MODEL
So, let’s just be clear that the subject really is the traditional “commercial open source” business model and not so much open source software itself. Considering this, I agree that the common characteristics of the open source business model does not lend itself very well to Enterprise Software Sales, or at least not as well as having suited salespeople knocking on doors and attending to local territories.
For this reason, even the open source ERP companies have structures and strategies that actually start to resemble proprietary ERP software companies quite a bit. The best presentation I have ever seen that compares the traditional software model to the open source model was a presentation given by Larry Augustin (ironically, now the CEO of SugarCRM) in 2007 at OSBC entitled: A New Breed of P&L: Open Source Business Financial Model. After eloquently describing all the unique points about Open Source and how Open Source Software is so special, Larry shows how by year 6 (the time it takes for a business to scale to >$50 million – the beginning of a size interesting to VCs), the financials of an open source company (yes, probably venture backed) tend to look identical to that of a traditional proprietary software company (see slide 26!). What?! Amazing, but it’s true. The industry that wants to think of itself as so different, is apparently not so different in the end (quite literally the beginning is different but the end is not). The same laws of physics (or atleast business) apply.
SugarCRM has Succeeded but it is Fake Open
It is temptingly easy to start making this argument. But this argument assumes as I mention above that the historically used open source business model is the definition of open source. It is not. This also goes to Derek’s misconception that open source companies should avoid VCs. Again, if an open source company wants to grow in a certain way, i.e. slowly and organically, then yes, the company should absolutely avoid VCs. VCs will insist on growing fast and big often changing in radical ways to do so. In the case of SugarCRM, this meant jumping into the cloud with both feet first and greatly deemphasizing the original ideas of their open source business model. Others are making similar decisions (look at Knowledgetree today vs. 1 year ago).
So, are ideas like moving to the cloud, deploying a significant sales channel, or increasing marketing expenditures concepts that go against open source? No. Are they concepts not usually associated with early stage (non-VC financed open source companies)? Yes. Does this equate to “Fake Open”?
I think that Yossi Ben Haroosh sums it up best in his comment at the bottom of Derek’s blog by saying, “you raise money, no matter what software you are making, the rules always change.”