[Also published on Linkedin Oct 5, 2015 – https://www.linkedin.com/pulse/flows-goals-processes-deconstruction-km-mukku]
- Just as in a legal action discovery process is used to dig out facts hidden deliberately or otherwise, a process is also discovered in an organisation. The legal action discovery process seeks to unearth information that is being kept from the judicial process. Process discovery in a similar fashion attempts to unambiguously define the process for the purposes of, say, automation or optimisation. Process discovery involves analysis, interpretation, identification of the activities, and an understanding of the actions taken by actors in executing the activities. Just as surprises pop-up in legal action discovery, process discovery often uncovers activities that are performed by actors to overcome what they see as either bureaucratic or needless regulations. A retrospective review of (what turns out to be a badly) an implemented process will also highlight such process blind spots.
- The difficulty in discovering a process is at times comparable to the difficult in a legal action discovery. The process discovery will at the minimum contain the following steps, “Process Investigation Preparedness”, “Document Search, Interviews, Walk-Throughs,” “Analysis and Review, Simulate and Validate,” and “Publish”.
- The first step, “Process Investigation Preparedness” is the most critical; proceeding without a thoroughly vetted completion of this step will have a domino effect of failures. You should have a “go, no-go” meeting at the end of the first step, to save a lot of grief and money.
- Walk-throughs involve shadowing the actors as they perform their work, and thus get a feel for its true nature.
- Complicated processes will mandate a simulation exercise to validate the understanding gained in the preceding analysis. But, unless there is a tool at hand it is unlikely to be cost effective. In most cases the tool has been purchased prior to such an exercise, so perhaps it may turn out o be not such an issue in most cases.
- Once the business goal has been selected, the discovery begins with an attempt to identify the activities and the actors who interacting with them. For example, an Insurers claims process goal is to successfully handle its clients claim requests. By “successfully handle” we mean the process itself and not the success or failure of any specific claim request. Here the process goal of “successfully handle a clients claim” is the feature of the process. A process feature is separate from the outcomes of any of its activity. It is necessarily a unique feature of the process, simply because no two distinct processes can have the same goal. Distinct in the sense that at least one activity is different and used, otherwise your organisation has processes with a multiple personality disorder. At the activity level the outcome of an activity does not necessarily have such a measurable business value. It is only in the aggregate that business value accrues when the goal is achieved.
- At the highest level of the hierarchy of the process constituents are its activities. The properties of activities include the objects that are necessary for the execution of the activity. For example, objects representing business entities, such as, Insured, Policy, etc., (which in turn may have further properties). These are necessary objects of some of the activities in a claims process. They in part provide the history of the activity up to the particular activity. Business rules are like constraints on the freedom of the actor to choose the follow-on steps available to him, these rules use the state history of the objects relevant for the process.
- Two types of structures in the process connect the activities. One of the structures depicts the activities as a designed rigid chain; this is typical of most BPM tools. The traversal of the chain is dependent on the particular instance of the process, and is bounded by business rules and actions of the actor. Thus this rigid flow structure (diagram) provides a potential static sequence of actions fixed in time. There are many reasons why this type of rigid chain is not always a good implementation. It not only denies change, it also makes it increasingly difficult to manage change. This is true even after taking into consideration that the tools are be getting better at managing change.
- The second type of structure is dissipative in the sense that, the process flow is known at the end, and it is then saved for historical and analytic purpose only. The chain of activities varies based on the process instance, and the selections and actions of the actor. This is close to the understanding of adaptive case management. Not having a rigid chain of activities means that the actors have more freedom to choose which activities are required based on the process instance, and the history of decision made up to the current point in time. New activities can be added to enhance the process definition, and increase the choice to the actors.
- Not all process instances are defined by all the activities. For example, a business rule, such as, – a claim below some number should be sanctioned without any further checks beyond validation – would invoke straight through processing of the claim, since all underwriting checks will be bypassed and only validation would be the step required. This is true irrespective of whether we have a rigid or dissipative structure defining the process.
- The actors (in the guise of roles) are also elements of the process but at a level below the activity elements. This is not to minimise the importance of the actor element in a process, but this separation is necessitated by the differing structure and hence role definition and hierarchy within organisations. .
- In many processes the actors may be external to the organization. The boundary of the activity extends beyond the organization. Security issues and limiting the scope of visibility of the process history become a much more important consideration.
- In many cases the outcome of an activity may not be only consumed internally, but may also act as trigger for activity in a another organisation, say, a partner organization. Think of this as extending the boundary of the output channel of the activity. It is also possible that the process may need input from an external source, for example, a partner organisation. Implementing these as pull or push transposes the mode of thinking about the extension.
The process flow thus defines the time line of the activities in pursuit of the goal. It then follows that the goal is the resultant of a chain of activities defined in the process irrespective of its type. In the static structure we know the chain at the beginning, in the dissipative we know the chain only at the end when process goal is attained. We can also compare the execution times, at both process and activity level, for process instances with a static (rigid) flow structure. This comparison is a lot more difficult when dealing with processes with dissipative structure without full knowledge of the whole process and its activities.