Previous We all have a role to play
Next Hacking like crazy!

Storytelling within the art of Enterprise Architecture

Carlos Kelkboom
0 minutes


Kobe Bryant, arguably one of the top ten basketball players to have ever played the game, was a tremendous story teller. If you have ever seen one of his interviews you'll know how much impact his words have had on a generation of basketball players. Storytelling is what drives our imagination; storytelling is what took our species from being tribes trying to survive by gathering and hunting our food to a species on the cusp of interplanetary space travel.

We, as a species, see patterns. Patterns to help us distinguish a snake from a branch or patterns which help us detect early stage cancer from a blurry black and white image covering a well lit background in a doctor's office. Storytelling gives us a way to fill other people's minds with discernible patterns and guide them to a realization which had otherwise not occurred. A good story will help you communicate your goals and your values, through a good narrative you can help everyone work towards a common goal. Like with almost everything in life, a good story is incredibly important when creating software. It is really hard to get 200 developers to create functioning software without aligning your goals, a good narrative is crucial to this endeavor.

This story is not about Kobe Bryant, this story is about how we as an architecture community must focus on the story we are trying to tell. Architects are very good at finding and exploring patterns, but a pattern is not something we can convince someone of, or sell to the business. A pattern is just that, a set of seeming similarities which might be abstracted and given a name. A story is what brings the need for these patterns to light. We should not skip the story, just so that we can spoil the ending. This article will show you how our habits and our tools cause us to miss most of the story and because of that, deliver software which could be a lot better.

Diagrams vs Story

One of our most important languages, as architects, is ArchiMate. We tend to use it to tell our story. Our focus on ArchiMate and the tools implementing the ArchiMate language, muddy the story which we are trying to tell with our software. A diagram is not a story, a diagram might be a picture used to bring a scenario to life, but it can never be the story by itself.

The ArchiMate language is maintained by the Open Group. To quote the Open Group's website: "The Open Group is a global consortium that enables the achievement of business objectives through technology standards." They publish a modeling language called ArchiMate which can be used to generate consistent architecture diagrams. The following definition of what the ArchiMate specification is can be found on their website:

The ArchiMate Specification provides a common language and elements to enable enterprise, business, solution and technology architects, business analysts and modelers, and software engineers to describe, analyze, and visualize the relationships among business domains in an unambiguous way - ArchiMate Open Group

If the goal of the Open Group is to enable us, to achieve business objectives, then why do they not focus on storytelling? Are we to assume that a diagram really triggers our imagination? Let alone our understanding? Where have we described the business objectives themselves so that others will understand what our diagram describes? Where is the narrative supported by the diagram?

When we see an image like the Archimate example our brains go into overdrive and we start looking for patterns. For some reason the yellow blocks are at the top. Some seem to have more rounded corners than others and still others seem to look 3d instead of being 2d. Wow, there are different arrows.

ArchiMate tries to combat this natural reaction of our brains by setting up rules and definitions. Like every intellectual worth their salt the Open Group started whittling away at their problem space by writing down what the arrows, icons and corners mean. From the entire set of possibilities they defined their way to something they think is reasonable. But what happens next? John, from upstairs, really needs a business capability to be implemented through a certain data object by a component, and before we realize it we have a committee bloating our language with exceptions and clarifications.

This is, of course, an over-simplification of what happens; but the gist stands, nothing in the ArchiMate language is axiomatic, everything is an interpretation and because the storytelling surrounding the rules put forth by the Open Group is lacking at best, we end up with a set of rules people need to learn by heart in order to have a chance at understanding the diagrams individuals create.

It is safe to assume that we now understand why none of the large tech-companies bother using tools like Archi or Enterprise Architect by Sparx, in their day to day development. Other types of organizations still use these tools and as a consequence the gap between architecture and development has never been bigger. To make matters worse, every five years we get a paradigm which is a silver bullet. It might be Functional Programming or it might be Cloud. It does not matter if it is a framework, a language or an infrastructure. With a lack of a common sense architecture we are faced with a plethora of choices regarding our development strategies and the implementation of our stories.

Micro-services, for example, are a way of thinking about a solution. They are patterns, guidelines and best practices shared with us by inventors. They are tools without a story. Not every story leads to a micro-service architecture and not every story needs the Interface Segregation Principle. It is true that once you are bringing your story to life like a director bringing a play to the stage, you'll have a certain freedom to choreograph your story. At the root of the implementation, however, remains the story.

Storytelling within Architecture

Effective storytelling will save us from the day to day chaos of large enterprises; let me explain why, but like every good story we will need to start at a beginning:

As a consumer I would like to see how much electricity or gas I have consumed in the last month. - Essent Customer

There, our story begins. There is no need for diagrams, there is no need for rules or definitions, there is just plain language. With these simple words our brains can start finding patters; we can start finding truths within this statement and we can grow our understanding by talking to other people and sharing our patterns and assumptions with each other. Our tribal instincts can be met by human to human interaction. This storytelling, at least for me, is at the heart of Domain Driven Design. And I personally define a Domain as: "An area of expertise I can fit into my brain."

But, how can we create software if we just have a story? We will need something we can validate, we can test, we can deploy and which we can take ownership of and maintain.

Yes, this is true, but we will need to crawl before we can walk and we'll need to walk before we can run. A child needs to feel safe to get the grounding they need to face the challenges of life. Software is the same, if we do not feel safe in our assumptions, if we do not feel confident in our skills and if we do not trust our architects to work with us; we will never achieve the level of software we will need to lift our business to a higher level.

Start-ups vs Enterprises

Let us dive into a tangent; why do start-ups fail? And, why do certain start-ups surpass larger enterprises? An Enterprise should have the money to outlast the competition, right? An Enterprise should have the money to look around, hire the best people and quickly integrate great ideas into their own production, right? The answer lies in storytelling. An Enterprise has hierarchy which means that the story can become fragmented and corrupted. Within an Enterprise a lot of focus and time is spent on aligning ideas, which is nothing more than: "Getting your story straight." When hierarchy comes into play, the story of someone does not fit the overall narrative but is forced onto the narrative by reasons other than cohesion. Hierarchy is used when the story needs to be changed but the actors and actresses do not understand or sanction the changes. If it were otherwise, the story would just unfold and there would be no need for hierarchical decisions.

Start-ups are smaller, usually run by an enlightened despot. They have a single narrative, a single story and everyone knows the story. There is no ambiguity within the story, no doubt, no fragility and no conflicting message; there is only the story. But, they do fail, a lot. Just because the narrative is strong and unambiguous, does not mean that the story is good or that the story is gripping. There are plenty of books which I do not care to read but the ones which gripped me have never left me.

Our story continues...

From our story we can distill that we have concepts like: "consumption", "electricity" and "consumer". These story concepts will need to find a place in our architecture. If we follow the breadcrumbs left by our story, we will start to figure out the sequence in which the events of our story take place.

Our story starts with a consumer, consuming, electricity or gas. In this part we set up the actors in our story, we describe how they interact, what motivates our actors and what limits them. Clearly these actors need something; or there would not be a story to write. Instead of saying: "A consumer consumes electricity." we might write something like: "Winter is coming! An unexpected cold front will hit our lovely country with a fierceness which has been unparalleled for the last twenty years. Anna feels the cold seep through her walls and turns on her central heating which is powered by heating water through the burning of gas. This heating consumes gas at an initial flow rate of 1/2m per hour. This of course is only in times when it is cold. During the summer the consumption of gas might even drop to 0." This story is starting to help us understand what "consumption" actually is and why it is important for us to show it to Anna.

To further help us understand this process we might want to draw an additional diagram:

Our diagram describes the preceded story in a different way. We do this because everyone sees patterns differently. Some people see images and lines between images and with that they form a story in their minds. Some people do not see anything but lines and blocks and cannot read the story from these lines. To be honest, not many people can. The ability to abstract something into blocks and lines is not that common; but the ability to go from an abstraction back to the actual story is something I have seen even less.

This is where tools like Archi and Enterprise Architect fall short. They do not tell a story. They help people organise their mind in lines and blocks but leave a project unattended for a year, maybe two and put a different architect on the project and you will probably hear the famous words: "This is wrong, I will have to re-do everything!" This of course is the famous mantra of the engineer and it is due to the fact that the story is not clear and that the code does not reflect the story. It is why projects fail, it is why turn-over is such a destructive force in a company; to be fair, it is why gilded-cages were invented.

Agile Architecture

Have you ever tried "Cooperative Writing"? It is when you write a story together. Let me tell you, it is hard. Your plot will change in ways you did not expect but the result can be beautiful. Let's think of ways to make cooperative writing easier. We could give an outline per chapter; describing the start and the end of the chapter, leaving everything else free to be imagined by the writer writing the chapter itself.

This is exactly what we do when we create software. The architect creates an outline of what needs to be done. Component A should take the meter readings and calculate the consumption of a customer. Again, there is no need for diagrams. What this sentence does not tell you is how we need to deliver this story, but a combination of different diagrams might give us what we need. We might use a flow diagram to show the interaction with other components. We might use a logical data model to explain the data or information needed in the story.

How we save Architecture

If we, as a society, want to ensure that we retain the knowledge we have built over the last 500 years; we will need to become better at telling stories. We will need to train the same skill as our ancestors trained and sit together to thrill, entertain and bring people together. Our stories unite us as much as our software does. Just imagine 500 years from now, what will they remember from this time? Our ArchiMate diagrams? Or the stories we've written in which the core of our humanity and our human endeavors are cataloged?

I do not fault the Open Group for pushing their standards, I do not fault the architecture community for following them. I just ask everyone, business and IT alike to start communicating in stories before building a single piece of software. After the story is told and everyone can tell it to their, children, wives, neighbors and basketball friends; than it is time to start writing your first unit test. In the order of things, the story must come first.

Carlos Kelkboom

Technical Fellow

Carlos is a Technical Fellow at Essent. He enjoys writing, he writes stories, technical content, music and code.
No comments have been posted yet.
To continue, please enter data in the marked fields.
To continue, please enter data in the marked fields.