In search of the perfect business visualization

How many ways can you 'see' a business?

Once you start to think about how businesses make sense of themselves, you quickly get drawn into a world of myriad diagrams and assorted business visualization techniques. If one wants to try to make sense of how these different visualization techniques and approaches hang together a consistent and cohesive way to compare and talk about these techniques is needed. The best way I have seen is Whitla and Whelan's Concreteness and Constraint Framework. This framework, enables us to take any visualization technique and compare it against other techniques, and easily identify its strengths and weaknesses.

The framework is outlined here, using the example of how one might choose to represent various aspects of a house:

The vertical axis represents the degree to which the visual model actually looks like the thing it is representing, thus in the top left we have a realistic, instantly representation of a house, whereas towards the bottom of the model, we have images that represent some aspect of a house, but are not necessarily easily recognisable as a house.

High levels of concreteness, usually mean that the representation is easily understandable by all stakeholders, and can therefore can be used to quickly generate shared understanding and alignment

The horizontal axis represents the degree of constraint placed upon the representation by the inherent rules and logic of its visual language. As a result in the bottom left, you can see a highly abstract, surrealist representation of a house, in which the creator of that image was free to imagine and represent their vision in any way they liked. In contrast on the bottom right we have a wiring diagram of the electrical system that is in the house. Diagrams such as this are highly constrained by recognised standards which mean there are distinctly right and wrong ways to represent electrical wiring plans, thus the creator of this diagram would have been highly constrained by those rules.

Representations that utilise highly constrained visual languages are useful for specialist groups, who have been trained to understand the rules of the language, such as engineers who can immediately utilise the diagram to ensure they are building the right thing.

In the middle of the diagram we have a 2D floor plan layout, it is recognisable as a house, albeit probably not as easily and quickly as the image in the top left, and its 2D nature immediately means as a representation it is more abstract that the top left image. The 2D format also carries with it some formal rules which one needs to stick to and therefore mean the representation is less freeform than the surrealist image in the bottom left.

Whitla and Whelan's framework gives us a useful start point for thinking about visualization techniques, but it raises the question of...

Could a visualization technique exist within the top right of the quadrant, thus being rule bound and useful to specialists, whilst also having sufficient concreteness to render it easily understandable by non-specialists?

Whitla and Whelan argue that top-right models do exist, and they point to the examples of digital twins and Building Information Modelling (BIM) type dashboards, such as the example below:

The thing is these digital twins and BIM dashboards, don't actually create a new type of visualization, rather they combine 2D graphical, 'normal' dashboard type inputs with 3D representations of the subject matter such as a building or a plant, sitting on top of a data model that allows an accurate representation of the underlying reality. There is no actual new type of representation here, it is an amalgam of existing diagrams.

This shouldn't be that surprising, if it was easy to create a single diagram type that was understandable by any audience, whilst also giving experts the specific information they need in order to be able to do their work, it probably would have been created by now, the issue is that it isn't easy. To see where the difficulty lies, consider the range of diagram types out there, and how they compare, Whitla and Whelan identify 4 groupings of diagram types, and place them on a visual continuum:

The reality seems to be that if you want to be able to provide an accurate, easily understandable perspective on a business, that is useful to specialists and non-specialists, there is no single diagram type that can do this. You will almost certainly need to be prepared to utilise multiple diagrams from all four of the diagram type groups, seen in Fig 2 above.

What this means therefore is that...

...in order to describe anything remotely complex or complicated, you're going to need a wide range of visual artifacts, many of which will be largely unintelligible to anyone other than an expert in the field with which that artifact is associated.

The question therefore becomes, can we do better than this? With advances in visualization techniques driven by gaming technologies, in enterprise analytics and in AI, could we create a experience that can seamlessly describe every aspect of an enterprise, to any audience cleanly and concisely, which gives them the level of situational awareness they need, without either bombarding them with too much detail, or abstracting away important contextual detail?

Think of this super-diagram, as an 'enterprise-wide meta-map'.

To deliver the experience I describe above, as a minimum an effective 'Enterprise Meta-Map' would need to be able to do the following:

• Deliver any of the diagram types detailed in the visual continuum, from BPMN models to architecture diagrams to story boards.
• Be able to seamlessly transition between diagram types to help the viewer maintain the context of what they are looking at.
• Show and explain the relationships between different diagram types, ie: Is one diagram a parent or a child of another diagram?
• Nest and/or decompose one type of diagram across/into other diagrams, ie: A BPMN swim lane, that can be viewed as swim lane in isolation (as we normally view such diagrams), or broken out across an architecture diagram, which shows which elements of the architecture are delivering which swim lane.
• Showcase (and maybe predict) how elements shown in one diagram relate to elements in other diagrams. For example, I am examining 'Thing X' in Diagram 1, which might be an architectural schematic, and I want to see what processes will be affected if I remove 'Thing X'. Diagram 1 is not a process diagram, so I will need to look at Diagram 2, a process diagram to understand the impacts of removing 'Thing X', but could the tooling be smart enough to update process diagrams and flag issues/risks/opportunities/threats automatically when the underlying architecture diagram is changed and 'Thing X' is removed?
• Be based on live data, so diagrams are not continually going stale as soon as they've been created.

These requirements are exactly the ones we are working to deliver at EMMa3D. Indeed the reason we're called EMMa3D is because it is our shorthand for Enterprise Meta-Map in 3D, and we're busy creating the category of 'Enterprise Meta-Maps' as we speak.

Right now we are heavily focussed on the ServiceNow ecosystem, and working with a number of experts to take the data and insights from the ServiceNow platform, and bring them to life as an Enterprise Meta-Map. The starting point in this activity is the ServiceNow CMDB (Config Management Data Base) module and our 'SeeMy CMDB' product that gives you an automated health check and maturity assessment based on ServiceNow's CSDM (Common Service Data Model) that has been developed by the brilliant Mark Bodman.

CSDM provides a single lexicon for creating coherent, common services across all of your IT management domains, indeed attempting to deploy a ServiceNow CMDB instance without adhering to the CSDM recommended standards is likely to lead to problems and limit the effectiveness of your deployment.

Over time we will be adding further 'SeeMy...' modules that will integrate and layer on top of our CMDB visualization, thus creating a true Enterprise Meta-Map based on one's ServiceNow instance.

DM me if you'd like to know more or get involved.