Why is it that Open Source Software hasn’t had as big an impact in the area of business applications as it has for infrastructure software? It’s a question that regularly circulates around Gartner.
[note: by “business application” I’m referring to the acronym-laden category of software that encompasses an organization’s business processes – things like ERP, SCM, HRCM, CRM, ETC., ETC., ETC.)
Now, don’t go overreacting to that first sentence, please! I am not denying the existence of many good Open Source business applications (ERP5 is still one of my favourites in terms of the elegance of its approach). Nor am I denying that there are organizations benefiting from the use of such projects.
But let’s be honest here. Open Source business applications have just not had the same disruptive impact that Open Source operating system, database, middleware and collaboration projects have. The fundamental question is whether this is simply a matter of timing or whether there is a structural impediment to this ever happening.
Personally, I think it’s a combination of both.
In order to explain this I need to come back to the concept of “Collective Competency” (Gartner clients can also read the associated research note – “Collective Competency: A New Business Pattern“). Collective Competency represents the strategic decision frameworks and ongoing tactical efforts needed to create collectively-owned assets. These types of assets – distinct from “entity-specific assets” – are emerging as the only real way for an organization to deal with distortions in a value chain system created by a dominant supplier.
When you look at Open Source as a manifestation of Collective Competency what starts to become clear is that it has been the software vendors’ efforts to alleviate themselves of value chain system dependencies which have fuelled so much Open Source development. Oracle doesn’t want to be dependent on Microsoft for an operating system so they support Linux. SAP doesn’t want to be dependent on Oracle for the DBMS so they try to drive MaxDB. Those concerned about Oracle, Microsoft and SAP dominion over middleware get stuck into things like JBoss.
But the story changes when it comes to business processes. The fact is that there is no real value chain dependency between vendors when it comes to the business process layer of the software stack (that will probably change in the future as SOA matures – but that’s a separate discussion). Generally they have been happy to build their own discrete process stacks and customers have been happy to buy them.
This is the structural impediment. Without a value chain system dependency somewhere it is doubtful that Open Source will have a serious impact on the business applications landscape. But that brings me to the timing issue, because I think there is a value chain system dependency which has the very real potential of emerging.
It is a mistake to see enterprises as the “user” of a business application. The processes represented within these products are ultimately used to deliver their products and services to market. From that perspective, business application providers are part of these organizations’ own value chain system. So the question here is whether or not business application vendors start distorting the value chain systems of their collective customer base.
I think this is absolutely a possibility, if it isn’t already happening in some areas. Increasingly business applications are cementing organizations into high dependency relationships with their suppliers, bound by steep exit costs and vendor copyright control of the IP. Those relationships are becoming increasingly expensive as maintenance costs continue to defy gravity (with the potential of taking on a whole new dimension as some business application vendors continue to explore “value-based” pricing models). All of this for what is effectively a partially used collection of non-differentiating process capabilities.
So, is there a point where enterprise IT organizations will start questioning the long term value they’re receiving from these business applications? If that starts to happen what they’ll rapidly realize is that the only sustainable long-term solution is to participate in the creation of collectively-owned business application assets.
In other words, it would be the Collective Competency efforts of the user community, specifically targeted at business application vendors, that will be the most likely scenario that will drive Open Source business applications.
There are some examples of this happening already. The SAKAI project in higher education is one. Collaborative Software Initiative is a company which is building a business model targeted at creating exactly these types of solutions.
But in general I’m not holding my breath. The fact of the matter is that most enterprise IT organizations do not look at business applications through the lens of value chain system dependency. They look at it through the lens of mitigated custom development cost. I’m not suggesting that’s not valid – but if that’s the only measure its a dangerous mistake on the part of the CIO and his or her IT organization.
I think the will be a growing realization by users on the nature and extent of the value chain system dependency they have on their business application providers. And if business application vendors don’t manage this carefully then watch out! Those users will do unto them the way software vendors have been doing unto each other for the last decade.
So says the law of Collective Competency!