My thanks go out to Paul Buhler, Chief Scientist at Modus21, LLC (www.modus21.com ) who alerted me to blowback from a comment I made at our recent London BPM summit (http://www.gartner.com/it/page.jsp?id=928017) . In my presentation on best practices for selecting a BPMS , I remarked that the right answer is often more than one BPMS, matching the BPMS’ capabilities to the needs of your process improvement project. Sandy Kemsley heard of this remark second hand (http://www.column2.com/2010/03/but-customers-dont-want-three-bpmss/ ) and erroneously concludes that Gartner is “…bowing to pressure from platform vendors that have multiple fragmented BPM offerings (e.g., IBM), and that it’s not a good thing for customers.”
Since Sandy has picked up the quote second hand, it is no surprise that the quote is taken out of context and that Sandy is misinterpreting it. Therefore, I thought I’d take the opportunity to clarify here.
In fact, our view is quite the contrary to Sandy’s assumption! My comment was made in the context of my discussion of the weaknesses in conventional and recent approaches to categorizing processes, (including Tom Davenport’s in Thinking for a Living). Most categorization approaches tend to oversimplify the endless variety of interactions between human, system and information resources that happen in any process. Processes range from the routine — which tend to be those that are highly structured and can often be automated with applications — to processes that progress in unpredictable ways and often require the interpretation and judgment of experienced individuals in order to be successfully completed. My research finds that different styles of processes still need different technological implementation approaches because of the inherent complexity in the work interaction patterns themselves. Even when the buyer’s needs are appropriate to a BPMS implementation approach – in other words, he/she requires visibility into the work in the process pipeline, adaptability of both work items and the process design and accountability for state changes to the work (all key indicators for a BPMS approach) – few BPMSs on the market are proven yet for handling multiple process styles.
Consider the broad range of unstructured processes. There is no BPMS that is well-proven at handling the full range of unstructured styles – including case management, network optimization, dynamic task management and content collaboration – equally well. For example, Singularity, Pallas Athena, Global360, Appian and EMC handle case management especially well, whereas Metastorm, K2, Agilepoint, Handysoft, and Fujitsu handle dynamic task mgt especially well and Adobe handles content collaboration especially well! Furthermore, of the over 70 BPMS vendors we follow at Gartner, Pegasystems is the most-proven BPMS for handling a broad range of process styles spanning structured and unstructured processes, with Lombardi and Savvion not too far behind.
In a market that is still in the “early majority” stages (reference G. Moore’s Crossing the Chasm), it is no surprise that mainstream buyers will not be happy with my recommendation that “3 BPMSs may be the best choice”. This reluctance underscores the work still to be done by BPMS vendors to deliver the “whole product” that will appeal to a mainstream buyer and drive this market across this last chasm. Yet even as BPMS products move to the mainstream, I predict that no vendor will deliver the “be all, end all” platform solution for every process challenge, especially as process challenges are addressed through dynamic composition rather than new application development to automate prescribed behavior.
Readers should note that I do not mention the largest software infrastructure vendors at all in this position……….Happy BPMS shopping!