In this post, I discuss why in my opinion process and data ‘go together’ as enablers for sustainable change. I leave the people and mobility parts for a later post.
Sustainable change is achievable by people, with data, and is applied to processes that are increasingly distributed. It stands to reason that these factors are all necessary parts of a successful, comprehensive business strategy to address problems and move forward consistently and without too many wasteful and unnecessary gyrations and backtracking.
And yet, it is all too common to find organizations with formal or de facto functional siloes, resulting in vastly reduced opportunities for collaboration and synergy, and incurring unnecessary costs through rework or under-utilization of available resources. Consider process improvement (PI.) How many times are PI initiatives apportioned across departments and functions such as Strategy, Quality, Risk Management, Operational Excellence, IT, Finance, a PMO, and of course stand-alone PI? This just about guarantees duplication of effort many times over, the unsettling perception by most that others may be cutting into one’s turf, and jostling to have one’s viewpoint prevail. Ill-conceived attempts at coordination of forces pulling in disparate directions ensue, all of which take the focus off the original goal of solving a problem for the customer, be they internal or external. This is the result of a failure in communication and in the efficient execution of plans meant to support poorly defined priorities.
Data and process: together or separate?
Data and process belong together. By this I mean they should not be viewed as the domain of different disciplines, and management certainly should not operate under the assumption that they belong in different corporate functions. It is self-defeating to state that data belong in IT, say, and process and quality belong somewhere else.
Why do I say this? My reasoning and experience tell me that any sensible attempt at PI requires data, and good data at that. If data are not trusted by end users and are in fact deficient to the purpose at hand by reason of incompleteness or inaccuracy, PI efforts at using them will fail due to squabbling by team members and managerial lack of support for the actions recommended. By the same token, data devotees need to understand PI, since it is only through the improvement of data-centric processes that data quality will get better. This is achievable by a balanced approach that incorporates both data governance and increased empowerment of end users on the data front.
A further reason why data and process belong together is given by the now established discipline of data mining, and the more recently formalized one of process mining. Simply put, data mining is not typically process-centric and yields little insight as to processes. Process mining, on the other hand, focuses on events and the tools to analyze event-related data. This is fundamental to improve our understanding of the dynamic behavior of systems and people interacting with them. With simulation at one end of the modeling spectrum having little to do with real data in a broad sense, and with data mining at the other end focused on data alone, process mining sits squarely in the middle and can help us develop models from event logs and study these models from the point of view of performance, allowed paths and transitions, bottlenecks, and compliance or conformance to desired behaviors. Many process modeling tools, such as Petri nets, have their roots in graph theory, and a vast number of properties of graph-based process models can be modeled and analyzed to predict future behavior. This is to say that data mining and process mining complement each other rather well.
The truth is that PI and data are all pervasive and need to become truly everybody’s business, if the mantra of ‘continuous improvement’ and the holy grail of robust, trusted data are to be realized. From an organizational standpoint, a good way to accomplish the desired goal is not to have players on different teams meet for lengthy and tedious monthly project reviews attended by dozens of reluctant team members, but to have small, agile, interdisciplinary teams of people who are able to deliver complete solutions working together on a daily basis. Physical co-location is better than emails, and far better than periodic meetings. Paying close attention to this will obviate the need for sporadic, incomplete, delayed, hard to audit, and eventually ineffectual communication among parties who are uninvolved participants at best and frequently on the defensive when questioned as to progress.
In the case of work having to do with data warehousing and analytics, say, it is important that a team is put together that is capable of handling both analytics and PI, as inevitable issues with deficient data in source systems will require an understanding of processes as well as skills in mapping these out and in communicating with end users concerning the current state and how best to go about changing it. PI staff need access to and an understanding of data, and data analysts need to grasp the essentials of PI if data are to be successfully vetted and their processes understood and properly documented for those who may follow in one’s footsteps.
Data are increasingly discussed in terms of their volume, velocity, and validity. The latter drfinitely requires an understanding of the data lifecycle and thus of data management processes. A further transformative dimension of change, not discussed here but just as important as the others mentioned, has to do with ubiquity of data and data acquisition, due to increased mobility of users and widespread availability of more powerful and fully networked hand-held — and, eventually, wearable — devices.
Towards integrative centers of excellence
While this may all seem humdrum and perhaps obvious to many, the truth is that, in practice, centers of excellence that have a dual focus on both process and data are not all that common. The straitjacket imposed by organizations on their people, whereby the structure comes first and then, within those boundaries, team members are tasked with solving new problems that inevitably cross borders and do not respect preconceived notions of authority over either staff or budgets, is a classic case of the cart before the horse.
A business exists to serve the customer. The customer has needs, and it follows that the business must adapt in order to deliver the best solution possible to address those needs. Much greater fluidity than is currently the norm is required in putting together teams that can quickly and expertly respond to emerging situations, changing requirements, and evolving financial constraints. This also speaks to the emerging role of well-rounded data and process scientists, versed not only in data but in processes as well, and able to interact with stakeholders effectively at a variety of levels. Making processes and data everybody’s concern and greater organizational agility will set one in good stead to better respond to change and add value to the business.