One of the most fun debates I have had in recent times was with a couple of execs from IBM and Cognos, and with a senior analyst from IDC, at last week's Information On Demand event in The Hague. The question was innocuous enough: "what do you think is the addressable market for Business Process Management?" "That's tricky," I replied, "as I don't believe that there exists a BPM market, as such."

There followed a great deal of debate, at the end of which I remained to be convinced that there was such a thing right now as a BPM market. That's not to say that BPM doesn't exist—far from it, it is an essential facet of many technical capabilities. However it is exactly this factor that makes it very difficult to define BPM as a market.

Where can we see BPM? The ability to capture business activities and use them as a template for service delivery exists in many technologies, as indeed it has done for some time. We have for example:

• Within content management and collaboration tools, BPM is becoming an accepted term for that element of the software that manages the flow of content across different business roles.
• Within enterprise application integration, there is a clear need to understand how applications and dataflows map onto business activities. Initial releases of Microsoft Biztalk, for example, came unstuck until they built in this element.
• Within software development, a logical extension of modelling business processes as part of requirements capture, is to use the models to support solution delivery.
• Enterprise applications such as SAP have long been delivered through customisable workflows, which require a level of management from a business process perspective.
• Also, BPM works hand in hand with IT and business service management from an operational perspective, such that service delivery can be monitored and supported appropriately post-deployment.

BPM is clearly a highly valued element of IT. But can it really be considered a market—and if so, how should it be defined? Should it consist of the superset of all of the above, or just the BPM element, if indeed it can be separated out in any useful way? Or should it consist only of the "pure play" BPM vendors, those which have a heritage in one of the areas above but are positioning themselves by leading with BPM?

My debating stance, which happens to still match with my opinion, is, "none of the above." We discussed the comparison with the transport industry—throwing a car, a tank and a plane into a (figurative) bucket just because they all require an engine does not mean we can define the "engine" market. So, as with BPM, there isn't a coherent enough boundary to frame a market space.

The debate did not result in any shared epiphanies between the participants (though its always nice when analysts from different firms agree). For me, the thought processes didn't stop there: a few nights' sleep allowed things to ferment, alongside all those other things I picked up at IoD, not least a slightly flummoxed acknowledgement about the value of Master Data Management. Absolutely nothing wrong with MDM per se, but I still find it quite surprising that it took the industry so long to work out it would be useful to have a single, shared, defining structure for structured information assets.

To whit: following other conversations it occurred to me that the real pain with BPM remained how views of business activities could be shared across tools. IBM claim export/import capabilities between tools such as InfoSphere and Lotus, for example. But what lacks is the knowledge of which is the "master" view—an issue exacerbated when we consider how such information is distributed (and worse, locked in) to applications and software tools across the organisation.

Perhaps what we need, like with MDM, is Master Process Management—tools that enable representations of business activities to be catalogued independently of any application, and then translated between one application and another. The sign-off of a form in the content management system may also signal the acceptance of a new customer in the CRM system, but such information is stored in the heads of those using the tools, and delivered as a point-to-point linkage. How useful it would (and I'm speaking from experience, having been involved in several such activities) to capture such information and relationships once, and use the "single view of business processes" to feed all of the above applications, if not more.

But let's be clear. Apart from a couple of obscure vendors (pipe up, if you're out there), such capabilities do not currently exist. There's no MPM Gartner Magic Quadrant or Forrester Wave, and even if there were, few if any vendors would appear at all. All the same, if we did have MPM tools, I have no doubt that plenty of end-user organisations out there would be much better off. And indeed, we would have an addressable market.