Explaining EA: business architecture basics 3

Note: a revised version of these three posts is available at http://www.improving-bpm-systems.com/pubs/Explaining-EA-BA-basics_v7.pdf

Continued from Explaining EA: business architecture basics 2

5 Managing the complexity of VEB

The interactions between activities reveal the different relationships between them. In order to manage the complexity, the primary interest of any architecture is to bring structure to those activities and their relationships. There are several techniques (services, capabilities, and processes) which are discussed below.

Activities which are used by a number of other activities (i.e. commonly-used functions which are the result of specialisation) are wrapped as services (which function as some kind of independent building blocks). A service is a consumer-facing formal representation of a self-contained provider’s repeatable set of activities which creates a result for the consumer. (It is considered that there are internal [even within an enterprise] providers and consumers.) It is important that the internal functioning of a service is hidden from its consumers, so that some parts of the enterprise can be changed independently. For example, a “proper” service can be relatively easily outsourced. Services are expressed in terms of expected products, characteristics and delivery options (cost, quality, speed, capacity, geographic location, etc.) – this is the Service Level Agreement (SLA).

Complex services are created by means of the coordination of more simple services and/or activities (in the same way that an orchestra is a coordination of individuals and their actions). In this sense, an enterprise is a mega-service composed of a network of nano-services. Each service is associated with an owner who is responsible for delivering the promised results in all instances in which that service has been requested. That owner has
  1. to know/estimate the demand-side needs (the service may have many different consumers who will be using it with different frequencies), and
  2. to design/organise/create in advance the supply-side capabilities to ensure those needs are satisfied.

Capability is the proven possession of characteristics required to perform a particular service (to produce a particular result, which may include the required performance). Capability needs to “understand” the mechanics of delivering that service. The mechanics include the resources, skills, policies, powers/authorities, systems, information, other services, etc., as well as the coordination of work within the service.

So, how can one ensure that a service has the required characteristics? There are three options:
  1. by contract (“re-active” approach) – acquire a service with the required characteristics, use it, check that its performance is acceptable and replace it if something is wrong with it;
  2. by measurement (“active” approach) – implement a service, use it, measure it, improve or re-build it, etc.;
  3. by design (“pro-active” approach) – build a service model, run a simulation test, improve the model, build the service, use it, measure it, improve it, etc.
The first option works with some support services, the second option can work satisfactorily with lead services and the third option should be used for core business services. The core business services can’t be outsourced, can’t be bought and must not be “damaged” (otherwise the enterprise may no longer function).

One of the models of the mechanics of delivering a service is a business process – an explicitly-defined coordination of services and/or activities to produce a particular result. The explicit coordination brings several advantages.
  • It allows planning and simulation of the behaviour of a service to evaluate its performance. If that service uses other services, then the demand-side needs for those services can also be evaluated.
  • It can be made to be executable, thus guiding how work is done.
  • It allows control that the actual behaviour of the service matches its intended behaviour, thus pro-actively detecting potential problematic situations.
  • It allows the measurement within a service of the dynamics of different characteristics, e.g. valuing, costing, risk, etc.

So, there is a structure of services in which some services are composed from others via explicit processes. The use of explicit processes allows the objective definition of the capabilities of composed services.

6 Typology of business architecture artefacts

6.1 Motivation artefacts (why to do what)

Vision and related “ends” chain – desired result, goals, objectives

Mission and related “means” chain – course of action, strategy, tactic/projects

6.2 Value and profit proposition artefacts (what to do)

Value, value-streams, value-chain, value creation, value system, TOM?

Products or assets (tangible and intangible)

6.3 Organisation artefacts (who is doing)

Organisation structure

Governance structure

Supplier, providers, customers, and other partners

6.4 Execution artefacts (how to do what)

Process, Services, Functions

6.5 Knowledge/information artefacts (with what resources)

Terms, facts, rules, policies, etc.

6.6 Performance artefacts (how well to do what)

Capabilities, KPIs


Bruce said...

I have been long searching for something like this explanation. Unlike most architects you do not confuse "process" with "capability" or "value stream", and you put process in context at the end. I need more time to study all of this but an extremely valuable contribution. Would you say your framework is "conventional" EA/BA thinking or unusual?

Alexander Samarin said...

Thanks Bruce,

In the BA there is no "conventional" thinking yet, so any framework is unusual thinking.

Considering that "conventional" EA does a great job in describing the “enterprise genotype” (a full nomenclature of enterprise artefacts or assets) and there are many techniques to evaluate the “enterprise phenotype” (a set of observable characteristics such as performance), then a _bridge_ between “enterprise genotype” and “enterprise phenotype” is still missing.

I think that executable models of relationships between artefacts (with the use of BPM and SOA) can be a good foundation for such a bridge. My framework is about how to make processes as those executable models. It is an unusual thinking yet.