The Art of Strategy
‘Mastering The Unpredictable’ covers unpredictable, operational processes. I want to focus here on the most important knowledge work in the top hierarchy of the business: defining a business STRATEGY. I did post a few weeks ago about my belief that management has a strong link to ART, which is even more true for creating a strategy. Thus, ‘The Art of Strategy’ is adapted from ‘The Art of War’ by Sun Tzu (approx. 500BC) who said: “Creativity must go beyond generation of new ideas; it must become an ongoing activity.”
The term strategy, from the Greek strat-egos (army-leader), appeared in the 18th century as a term for the “art of the general.” Adapted from military jargon, a strategy requires a number of principles:
- Define simple to understand, attainable objectives
- Assign a single commander for each objective
- Take the initiative to act rather than react
- Prioritize to concentrate efforts and not dilute action
- Remain agile with your forces to react to changes fast
- Communicate securely to avoid information in the wrong hands
The above principles are a good approach for any planning level and any situation in life. Given that Sun Tzu was also a Taoist master, this is not surprising. It does contain the core elements of what I see as requirements for adaptive processes in a business. The last point is a warning call for social networking advocates.
Knowledge processes such as creating and executing a strategy don’t focus on flow, but on architecture and goals. In today’s business environment it simply is no longer feasible to define a strategy without having a business model and architecture that enables its application and execution. There would be no terminology to define the strategy with! As there aren’t any industry-standard business architecture frameworks on the horizon that would allow that, businesses need to iteratively create unique business architecture models for their needs.
A good strategy does not rely on analysis and the future being predictable, but much rather takes unpredictability as the only fact. A great strategy certainly does not make assumptions about initial conditions and then enforces a related course of action to achieve some goal. A great strategy is not even invalidated by changes to environment or technology. Such changes should only have an influence on tactics only but not on strategy itself.
With all the communication technology available, executives seem to believe that they can exert more and precise control remotely and fall into the micro-measure-to-managment trap, which is also a clear sign of mistrust. Sun Tzu said: “The ground determines the distance. The distance determines your numbers.” Communication is valuable, but just because data can be collected and reported from everywhere, distance does not become irrelevant. That is however the BPM and BI approach: ‘These are the numbers and this is what we are going to do about them, regardless and if it doesn’t work we try to improve later.’ The relevance of numbers is distorted by distance, time and processing. Information and numbers are like food: only local, fresh and carefully cooked food is valuable. Distance can only be counteracted by empowering people to act local, where their information is recent. Obviously, they ought to follow simple and attainable tactical goals.
To allow this, people have to be nimble and open to change, which is only possible in a bottom-up, evolved hierarchy, where each layer acts towards the strategic goals independently and according to their own best knowledge. That kind of people empowerment was already suggested by Sun Tzu as he wrote: “When one treats people with benevolence, justice and righteousness, and reposes confidence in them, the army will be united in mind and all will be happy to serve their leaders.” Empowerment is about trust.
With the ever growing complexity of the economic environment and the lack of transparency that this brings along for the executive, various ways to improve information have been explored. One strategy is the dreaded data warehouse with complex predictive analytics on top. I see this as the ultimate fallacy, because statistical patterns from the past do not predict the future of an individual situation. The ability for better decisions can thus only come from getting closer to NOW and not by moving further into the past and remote data to try and reach out to the unknown future.
How could a business react to certain situations NOW?
One option is to use business rules to define a library of events that require immediate action. That needs a business architecture to provide a master meta-data definition that rules can be applied to. Mapping rules to data structures to identify complex events can be a substantial piece of work. Many of those events can be seen as process exceptions that do not point at the execution being wrong, but at the PROCESS BEING TOO RIGID! An adaptive process environment will not experience exceptions, lacking a predicted flow. So how do we know that the process is right? That is not done by enforcing sequential execution, but by defining goals with multiple solution options. Here event defining rules makes sense to test for COMPLIANCE to regulation or business principles. As I have mentioned before, in an adaptive process environment exceptions no longer break the process, as the system or user can add conflict resolution goals to the process as needed.
If rules would be too complex then a business architecture can be used to create a taxonomy of classified event scenarios in the repository. The ontology of the business data is a prerequisite to taxonomy classification. What if those event patterns could then be captured by a pattern matching algorithm that is trained by user action? Pattern matching technology like the User-Trained Agent is capable of identifying repeated patterns that signify a certain event taking place without rules having to be coded. The meta-data in the repository ensure that all incoming business data are plausible for the agent that performs the event discovery. Rather than a complex technology implementation that has to create and manage the schema, signatures, and interfaces of the external events, the business event patterns should be identified transparently from the execution environment by setting up agents that monitor the process execution. Especially in an adaptive process/case environment such a capability is essential, because it is even more difficult to create rules for unpredictable processes. That is not post-mortem process mining, but happens in real-time and does not require intermediate knowledge engineering. It can be typical applications such as claims handling or purchase-to-pay resolution, but it can also be tactical and strategy planning activities that are suggested because of certain business data patterns.
Is all that really a relevant issue for strategy? More than that! Taking the initiative and assigning priority to processes as the need arises is a key element of a great strategy, military or not. Defining a business architecture that does not fall apart when the conditions change and empowering users to deal with real-time discovered business events, may be the most strategic element that an executive can add to the business he/she is responsible for. It makes the business resilient to change. Orthodox BPM may reduce cost and increase measurable quality for predictable szenarios, but it makes a business brittle and vulnerable given today’s rate of change.
Adaptive processes for strategy planning need process definitions that are goal-oriented and a business architecture model that can be modified by the business users. Those abilities empower executives and management to plan and execute strategic, tatical and business goals for processes that span today several application environments. The consolidation provides the seamless link to strategy, while on the other hand integration of existing systems just causes a lot of pain in implementation and maintenance.
To define a strategy it is not necessary to develop an enterprise-wide business architecture first. The same is true for business event patterns. Business entities and structures linked with strategy can be created where the need arises, but with a central meta-data repository it is possible to ensure that all models will be aligned. Strategy is about aligning business architecture with business need. The strategy starts with the vision and mission statements and outlines the core tactical principles. It uses the business architecture to describe multiple model layers, from capability maps, strategy maps, business models to finally process goals/milestones. The business fills bottom-up the required process goals with execution knowledge. All models interrelate and all should be derived from strategy and kept in one place for top-down transparency. Only then it is possible to define simply, attainable objectives for the commanders.