Year: 2007

I’m not sure that I get the purpose of these late afternoon registration-day sessions, before most of the conference attendees have even arrived — are they intended to be introductory tutorials for those new to the field? Or a venue for Gartner’s second-string BPM analysts to dip their toe in the water? I don’t have the impression that Kerremans is completely comfortable with his material, much of which I have seen at previous Gartner events.

The material is, after all, pretty standard stuff:

1. What is a business process?
2. How much BPM technology is enough?
3. How will BPM technologies evolve over the next 5-10 years?

He gave a standard, yet succinct definition of a business process: a set of activities and tasks performed by resources (human and machine) using a variety of information (documents, images, expertise) interacting in various way (sequential/predictable and ad hoc) guided by policies and principles (goals, business rules, principles).

He discussed how many business processes are currently embedded in application silos, each of which may contain UI, middleware, rules and data, but that there is little integration across silos. One “solution” to this lack of integration that has been popular in the past is to replace all of the applications with a monolithic ERP system, but that creates larger problems in the long run when you inevitably have to integrate that system with other applications that do what it doesn’t do. A better solution is to tear down the walls between the silos and use BPM technologies (a blanket term that he used to refer to all integration technologies) to integrate the applications directly.

He showed a current Gartner hype cycle for a range of BPM technologies (the hype cycle is a Gartner technique for plotting visibility against maturity for various technologies: from the technology trigger, it rises swiftly to a peak of inflated expectations, drops into a trough of disillusionment, then rises gradually over a slope of enlightenment to a plateau of productivity). There’s an interesting distinction here between more mature technology classes that have BPM as a feature — human workflow automation, document and imaging management, integration brokers, portal servers, application servers, and rule engines — and those where BPM is the actual product: pure-play BPM tools and BPMS tools. Gartner shows pure-play BPM tools, those darlings of last year, about to plunge into the trough of disillusionment, while BPM suites are still climbing up the initial peak of inflated expectations.

In the next section, he takes on the question of how much BPM technology is enough, but never really answers the question. We saw a number of standard Gartner charts showing the functionality of BPM as Gartner defined it in the pure-play period as well as today’s suites trend, and some mapping of business needs to the value that is potentially added by BPMS.

The most important issue in the entire presentation (in my mind) is buried in here, however: do you “roll your own” by assembling best of breed from a variety of vendors, or do you buy a BPMS from a single vendor? Every organization already has some investment in parts of what I consider to be non-core BPM technology, such as content management or business rules management, and may have corporate standards in place for these things. Should they consider replacing some of that functionality with what’s available in the BPMS? Or allowing the BPMS to have some special dispensation to ignore the corporate standard? Although I tend to favour using widespread corporate standards if you have them in place, there are some valid reasons to consider an all-in-one BPMS — such as ease of troubleshooting and any resultant vendor finger-pointing — but ultimately, I think that it comes down to the same arguments as you would have against a monolithic ERP system. If your primary integration technology platform can’t integrate with what you already have, what good is it?

The last point was on how Gartner expects BPM technologies to evolve over the next 5-10 years, and I think that we’ll be hearing more about this over the next three days.

Since analysts (like Gartner) have expanded the definition of a BPM suite to contain practically everything but the kitchen sink, there’s now a problem of too many vendors all claiming to have BPM functionality (of course, you can always hire an analyst to help you sort it all out…), so some convergence and shakeout will certainly occur as players merge and acquire. The big platform players (Oracle, IBM, SAP, Fujitsu, Microsoft) are just starting to enter the market, which will place further pressure on the smaller players to join forces or be consumed by the bigger guys.

At last year’s Gartner BPM summit, I heard Daryl Plummer speak about ISE (integrated service environment) for the first time, and they’re still playing it up as a major part of a further convergence of tool categories between BPMS, ISE and IDE: overlapping product classes that haven’t previously considered to be competitors. I’m still not so sure about this argument; I see some pretty fundamental differences between ISE and BPMS.

On the BPM technology innovation side, Gartner is seeing things primarily around business agility: goal-driven and self-adjusting processes, and complex event processing. They see a future of business process platforms that will completely disrupt the packaged application (e.g., CRM) market, although I’m not seeing the potential for disruption so much as a gradual shift of functionality from packaged applications to BPMS platforms.

On to the evening reception, then an early night as I try to get adjusted to west coast time.

I totally forgot that there were some late afternoon sessions today at the Gartner BPM summit until I showed up at registration around 3:30 expecting it to be deserted, and found a lineup of people trying to get registered before the 4pm session started.

This first session of the conference — missed by most attendees, who haven’t even arrived yet due to extremely inclement weather conditions across half the continent — was entitled “BPM: The Discipline”, presented by Bill Rosser of Gartner.

Rosser’s focus is on strategy, planning and modeling, and he recently shifted to BPM from enterprise architecture and other areas. In fact, last year when I saw him speak at the BPM summit, he was talking about EA, and I wasn’t all that impressed with his presentation then because it failed to make an adequate link between EA and BPM.

Rosser defines “discipline” on a number of levels, from “a field of study, knowledge and expertise” to “a system of rules of conduct or methods of practice” to “a method to ensure adherence to the rules”, then spends most of his presentation on the question of whether BPM is a discipline. First of all, to be blunt, who cares? Secondly, from a purely logical perspective, I don’t think that he proves his argument of whether BPM is or is not a discipline.

Although he had some nice slides on why to pursue BPM — performance improvement, greater agility [sic] to change, and stimulus for innovation — plus some on top-down versus bottom-up implementations, there was really nothing here specific to BPM. I had the sense that this was material recycled from other presentations about other topics, since it was really just a general discussion about discipline and implementing strategy: you could replace the words “business process” with “enterprise architecture”, for example, and hardly have to tweak the presentation at all. In fact, one chart was based on an HBS paper on enterprise architecture strategy, making me think that Rosser had just carried these over from his work in EA. Although some of these ideas here are as valid for BPM as they are for any other business/IT strategy area, I would have expected something that is much more specific to BPM.