Two years ago, I complained that BPM had lost its appeal as a technology, acknowledging a dearth of public discussions on the topic. Well, I should have been careful about what I wished for, for I definitely got it, in droves. The partisan debate has been fueled again, ignited with this post, relayed by InfoQ with proper Intalio bashing, then followed on this Google Group, with lots of bogus pseudo-scientific statements to make for a fascinating read. To be clear: I’m loving every part of it. BPM is a topic I am passionate about, and a field I dedicated my (short) career to. But let’s make no mistake: this silly discussion about BPEL and Pi-Calculus is more than a marketing spin on things most of us will never fully understand (myself included). It has been brought forth for a very important reason: true distributed process execution is of the utmost importance for BPM’s killer application: Business Process Outsourcing (BPO). Let me explain why.

The enterprise software industry is currently going through the most profound transformation of its relatively short 40 to 50 years old history. Cloud computing is fundamentally reshaping the architecture of enterprise systems, and the combined forces of Open Source and the Software as a Service model are making once-profitable business models utterly obsolete. The industry is maturing, going through an irreversible industrialization process. At the top of the food chain, alpha consolidators like IBM, Oracle, and SAP are building massive stacks distributed by vast armies of sales representatives and resellers that smaller vendors will never be able to compete against. At the bottom, component suppliers are leveraging commercial open source business models a la COSMO to undercut their competitors. In the middle, casualties will pile on, for most players won’t find any decent exit in tomorrow’s economic environment.

If you stop reading the story now, you get a pretty bleak picture for the enterprise software industry at large, and might feel like the game is over. The industry matured, and as a result opportunities went away. This would be a mistake though, for a new category of players will emerge, and create a whole slew of new opportunities. So carry on reading, for I’m not done with the story.

The new wave of opportunities will not be found in software packages or services. There will be no next Oracle or SAP, and Salesforce.com’s core business model remains to be proven (I think it will, but it’s beside the point). Instead, it will come in the form of next-generation Business Process Outsoucing (I hesitate to call it BPO 2.0, having failed to the 2.0 tempation too many times myself), powered by BPM 2.0 technologies, and fueled by unprecedented transformations in the Financial Services Industry.

The conventional wisdom is that FSI companies are early adopters of new technologies. While valid in many respects, this viewpoint overlooks another characteristic of the industry: its antiquated vertical integration. Unlike other industries like automotive or computer manufacturing, most financial services institutions remain vertically integrated, fulfilling most functions internally, in a rather inefficient manner. In today’s economy (and even more so in tomorrow’s), a retail bank processing less than 5 million credit card transactions a day should outsource this entire process to a third party. And if it has less than 1 trillion dollars of assets under management, it should not be in the securities business. The same is true for most other processes found in banks today, be it mortgage processing or loan servicing.

Moving forward, most financial institutions will come to the same realization, and will focus on the only process they can master: Customer Relationship Management. Every other process will be outsourced to a handful of vertically-integrated mega banks, or a new breed of dedicated service providers. This industrialization process of the financial services industry will be triggered by the current financial crisis, will be forced by the reinstatement of a Glass-Steagall type act, and will take 10 to 20 years to unfold. Through this process, banks will have to decide which processes they keep internally, which processes they outsource to third-party service providers, and which processes they in-source on behalf of third-party financial institutions. To support such a radical transformation of the financial services industry, they will need to leverage a technical infrastructure that supports the definition, deployment, and management of highly distributed and extremely agile business processes: BPM 2.0.

This transformation will in turn come to define the aforementioned next generation of Business Process Outsourcing. It will show how companies can collaboratively outsource business processes (as opposed to the badge-flipping IT outsourcing of the 90’s), developing the technologies and best practices that will make such outsourcing manageable, scaleable, and cost effective. But for it to work, it will require the use of BPM technologies that support the definition and execution of end-to-end processes which sub-processes (represented as separate swimlanes in BPMN) can be deployed on separate servers, managed by different organizations, and migrated back and forth at will. And this is precisely where the rubber meets the road: whether legacy workflow vendors like it or not, BPMN is the only notation that supports the definition of such collaborative processes, and BPEL is the only execution language that supports their distributed execution on top of a Service Oriented Architecture. This is precisely why BPMN+BPEL matters.

Once financial services institutions will have successfully completed their renaissance through such an industrialization process enabled by Business Process Outsourcing, other vertical industries will notice and follow suit, thereby giving birth to a new breed of BPO service providers that leverage BPM 2.0 technologies, Software as a Service business models, and relatively cheap(er) offshore labor, in a highly leveraged fashion.

Anticipating such an evolution, many IT outsourcing firms (especially the top 5 in India) and BPO outfits will consider acquiring BPM pure plays looking for an exit in today’s unforgiving market environment. I have a contrarian advice for those: resist this temptation, and license a standards-based offering supported by the COSMO business model instead. It will be far cheaper, much better quality, and far easier to maintain over the long run. Your competitive advantage won’t come from the ownership of a proprietary piece of legacy workflow software. It will come from your understanding of how best-in-class processes can be automated by best-of-breed software, serviced by highly-trained staff, and integrated with your customers’ own business processes, in a very agile and radically distributed way. And Intalio is here to help you.

End of story.