Decision Automation
Why is decisioning important? Decisions are the critical mechanism by which we choose one action from the set of all possible actions. The purpose of a decision is to select the action that is most beneficial to the decision maker in a given situation. A decision is 'made' when we resolve the available data to a single definitive outcome - a decision cannot be multi-valued, ambiguous or tentative. The derivation of the decision outcome is achieved by applying discipline, knowledge and experience to the data within a specific context. It is this application of discipline, knowledge and experience that most closely defines the unique character of any business.
Business Decision:
A single definitive outcome that is the result of applying business knowledge to relevant data for the purpose of beneficially directing the activity of the business.
The desire or need to regulate and control system responses to give a zero touch process is ensuring that decisioning is becoming a core function of many modern systems.
In these systems decisioning drives the discovery of data and its relevance to the business - the need for the decision drives the need for the data. And because the decision outcome predicates the process response, it also implicitly drives the process definition.
Businesses do not make a decision merely because they have available data, nor do they act without making a decision. Decisions are clearly identifiable as the heart of any automated systems response.
If decisions are core drivers of the data and process elements of a system, they are also drivers of other decisions. For instance, the decision to accept insurance business depends in turn on decisions regarding the customer, the insured risk, price, and underwriting terms. Each decision can also depend on other decisions until a tree structured model of decisions is formed - the decision model.
Our experience has shown that decisions are the fastest and easiest of the system design elements (i.e. of data, process, decision) to discover and model. The decision model is the core of business knowledge. The decision model then provides a fast-track to the discovery and modelling of data and processes, giving rise to a 'decision oriented' system.
The systemisation of decision discovery, definition, and deployment, including the execution of deployed decisions, is the basis of the new discipline of 'decisioning'.
Decisioning fits naturally into an automated systems framework. External events occur that require business responses. Where the business has a selection of possible responses, a decision is required to determine the most beneficial response. In fact, the business already knows how to select the most beneficial response based on either experience or declared policy. Decision automation is about encapsulating that knowledge into decisioning 'components' for insertion into computer systems.
In legacy systems the decision maker is usually human, so the process must be interrupted to acquire the decision. Modern systems seek responses that are 'zero touch'. To achieve zero touch, we must automate the decisioning that was performed by human 'actors' within the process, replacing them with automated proxies that immediately return accurate, regulated and auditable decisions.
It is necessary but not sufficient to incorporate decisioning know-how into the system code-base. The decisioning knowledge results in a different class of system component to the data and processes that define the rest of the system. Events, data, and processes are generic by industry - they are relatively stable and generally do not differentiate individual businesses. For example, all insurers have customers and risks - both of which exist completely and independently of any particular insurer. Similarly, all insurers have processes to issue policies. But all insurers will use different decisions to determine how to accept customers and risks, and how to price and qualify the policies.
Decisioning demands a new development approach. When implemented as plug and play components within the comparatively static application framework, decisioning can allow the system to respond rapidly to business changes without affecting the underlying code base. The decisions driving business behaviour can adapt as fast as the business can learn.
For systems vendors, this plug and play approach also allows 'mass customisation' by mixing and matching bespoke decisioning components within a common application.
For management, decisioning is a critical implementation of business policy and practice and so independent testing, verification, and audit-ability of decisioning components is essential.
For the developer, the components must integrate easily, execute quickly and robustly, and not otherwise constrain architectural options.
All in all, decisioning components look more like programmable logic controllers that exist entirely independently from the systems they control - they are remote controls that allow the business to directly govern systems operation through controlled decision making.
Find out how the Idiom Decision Manager can help with automating your business decisions