As-is Modelling, the Sweet Wasteland of Enterprise Architecture
Alex Jouravlev
Data and Enterprise Architecture veteran and practitioner with up to date strategic knowlege and hands-on skills in AI. Proponent and enabled of Data-Driven Enterprise. Everything Graph and Metadata
Enterprise Architecture is under attack. On one side, the Service Design people are “planning and organizing people, infrastructure, communication and material components”, which is exactly the same as what Enterprise Architecture should do, only elegantly and with greater focus on customer. On another side, Data Scientist and Data Analysts promise performance and flexibility of Data Driven Enterprise.
Neither side wants to talk to Enterprise Architects. That could eventually relegate Enterprise Architecture into an irrelevant bunch of people who produce pretty pictures nobody really needs while other people deliver real improvement by placing real change and real data in front of the executives. Enterprise Architecture is losing the fight for attention of decision-makers.
The core of the problem is in irrational emphasis on Certification, and in obsession with as-is Modelling.
One has to admit, As-is Modelling is an ideal Consulting activity. One doesn’t have to be very smart, experienced or knowledgeable to ask a local SMEs about their processes, systems, integration etc, and record it with full compliance with say Archimate. At the same time, As-is Modelling requires a lot of work like that. A lot of work with hardly any responsibility, which is nearly impossible to fail. So if a Consulting outfit managed to score some As-is Modelling business, they can deploy a lot of very cheap resources, and receive massive margin for very long time.
Actually, As-is Modelling is an infinite activity. There always be more to model. However, while it happens, Service Designers are running workshops discovering what the enterprise does and should do, and Data Scientists are discovering what data adequately describes performance, opportunities and bottlenecks of the enterprise.
How long that is going to last?
The task of Enterprise Architect is not to model complex, ineffective mess, but to create a future which is a pleasure to model.
Sometimes, that quest would require to model how the things are done now. However, that activity should be performed to the extent it is required to develop safe and effective transition to a convincing future state.
Sometimes there are questions that cannot be answered without investigating the current state. Then the current state should be modelled to the degree required to be able to answer the questions.
Enterprise Architects are often some of the smartest, informed, experience people the enterprise has. They should not be reduced to modelling the junk other people create.
We need to talk about what Enterprise Architects should do to make Enterprise Architecture where business vision comes to find ambitious yet realistic answers. But the fist step is to admit that full As-is Modelling should not be the first or the dominating activity of the Enterprise Architecture function
Mostly Retired Business Architect and HE Lecturer.
6 年Very sensible piece.... In reality EA lost the plot years ago....in many cases its is not EA at all but just some I.T. systems mapping and if your lucky some correlation to basic business components.
Principal at Legacy Software, Ltd.
6 年Granted I dropped out of the EA debates a while back... primarily because the EA world appeared to be entirely uninterested in the AS IS domain.? The AS IS is what's in use today & is 100% guaranteed to be machine readable. How can one dream of the TO BE, if you don't know where you are today?
CEO at Clarasys. Enterprise lead to cash and order to cash transformation
6 年Part of the goal of an Enterprise architecture function should be to ensure that as-is doesn't ever need to be created from scratch - it should be maintained so that there is never a need for a colossal effort to map it; and so that change can be managed against it. Not understanding enough as-is leads to expensive mistakes when making changes to complex processes and systems and can stifle innovation as people become scared of the impact of a change
Architect ? Digital Transformation ? Microservices ? SOA ? Methodologist ? Practitioner
6 年I agree As-is Modelling is 'Sweet Wasteland', as long as model is being used as document. 'As-is' modeling is useful to understand complexity about the enterprise system, identify risk, impact, etc in planning phase of any change/CR.