Hydrasight research confirms that business process management suites (BPMS) continue to be seen as an important component of functional process improvement in many organisations and, to a lesser extent, for cross-functional processes. However, we also observe that many organisations continue to be challenged by their own immaturity with implementing both functional and cross-functional process improvement. This is due to the difficulty associated in codifying processes that often lack sufficient detail or have complex exception handling requirements.

Hydrasight believes that the use of BPMS technology, on both functional and cross-functional processes, can be an appropriate application area for BPMS products. Most BPMS products now offer capabilities to support complex process flows, systems integration, authoring of business rules, automation, operations visibility and modeling standards needed to successfully implement these business processes. However, we caution IT organisations (ITO) and business professionals to recognise that BPMS products provide little support in respect to defining accurate process models where such definitions do not already exist or where processes may be unsuitable for codification (e.g., discretionary decision points).

Hydrasight observes that leading BPMS vendors offer ‘business friendly' tools to aid users in functional and cross-functional process definition (e.g., Lombardi BluePrint, Metastorm Provision). However, we note that process definition for complex processes, and/or processes unique to the organisation, remains a challenge with all BPMS products. Our research confirms that developing a process definition that accurately represents the way work is performed and optimized to achieve desired process outcomes is the single most important factor in determining results gained - whether BPMS automated execution is utilised or not. Hydrasight further notes this is symptomatic of a larger issue where improvement practices such as Six Sigma and Lean also experience significant difficulty in producing desired results on processes not already clearly documented and understood. From our research, we believe this a key factor in limiting the potential of BPMS products to become the foundation of the integrated process enterprise and of process improvement initiatives in general.

Hydrasight research shows that organisations successfully adopting BPMS often focus on processes that are already well-documented or where best practice process models are readily available, diminishing the value of the dynamic optimisation capabilities of the automated suites. This is often due to greater availability of best practice models and the disconnect between ad-hoc business processes and those that are already codified/encapsulated in software applications (e.g., human resources management as one example).

Hydrasight research shows that having access to best practice and/or existing process definition does not guarantee success.

We further observe that even when extensive ‘as is' research is performed before producing a process model, results may still not be consistent with how the process ‘really works' due to the subjective nature of the people involved (e.g., both interviewers and interviewees).

We typically observe that cross-functional processes are rarely well documented or clearly understood in the majority of organisations. This is further complicated by multiple stakeholders with different functional perspectives. The rise of interest in business process management has prompted many organisations to begin looking at these processes with the intent to support them in a more programmatic way. While we recognise the value of this approach, and the efforts to bring cross-functional processes under programmatic control, Hydrasight cautions organisations that the challenges are great-and that outcomes often fall far short of business expectations.

Hydrasight foresees that the subjective elements of BPM, such as the interpretation of process definitions and/or execution decision points, will remain one of the greatest challenges for organisations to face in leveraging the potential of BPMS products. We caution ITOs and business professionals to ensure they pay close attention to the importance of designing process models that accurately reflect how work ‘really gets done'-or how it will be done if the process includes moving to a ‘to be' state of improvement. Moreover, we advise organisations to thoughtfully consider the role of the tool versus the general benefit of process improvement, whether automated or not. In general, we recommend a proof-of-process approach as opposed to a proof-of-concept approach. This assists in avoiding project ‘failure' being directly linked to the tool, rather than the way the tool has been implemented by the business (refer "What's wrong with a proof of concept?"). In either case, Hydrasight observes that codifying an inaccurate process model is likely to lead to failure that extends beyond the intended process goal into issues of diminished confidence with BPMS products, other processes in the organisation and the IT organisation in general-whether they are directly involved or not.

Hydrasight's full set of published trends for 2009, including CIO, virtualisation and SOA, can be found at hydrasight.com (registration required, no charge).