When you’re trying to figure out a better way of doing things, having a model can improve your ability to see what really matters in your business…and what the impact of your change will be. However, that will only work out well if you pick the right modeling tool. The first step in getting the right tool is to look at the model’s dimensions and relationships. The last step is to validate that model.
When your organization has a new opportunity to pursue or a problem to solve, modeling your business can be a critical tool in helping you decide what’s the right thing to do. However, there are a ton of different ways you can model your business: You can create an organization chart, a business process flow, a strategic capabilities graph (just some of the many different kinds of modeling tools we discuss in Learning Tree’s Modeling for Business Analysis course). Which, of course, raises two questions: Do you need all of these models? And which ones would actually help you?
The answer to the first question is “No”–you probably don’t need them all. The answer to the second question is more interesting and requires understanding how a model can be useful to you when you’re trying to solve a problem or pursue an opportunity.
A model, of course, isn’t the real thing: It’s a scaled down replica of whatever reality you’re modeling. Any model leaves many things out…which is why models are so useful. A model allows you to focus on the components or relationships that matter in your organization, at least as far as this problem or opportunity is concerned. A model clarifies the problem by focusing on the essentials and eliminating details that won’t make a difference. Furthermore, a visual model can allow you to quickly assess a problem just by looking at the result of your modeling process. A really useful model will even let you go one step further to do experiments and simulations — to ask the question “If I make this change here, what will the result be?”
This means that, to pick the right model, you have to know what problem you’re interested in and what questions you want to ask. A model that doesn’t address your problem can’t answer the questions you’re interested in.
Models can be looked at as being made up, first, of components (called “dimensions”) and, second, of the relationships between those dimensions. A model that’s useful to you will include the dimensions and relationships that make a difference in your problem or opportunity.
There are six fundamental dimensions that a business model can include: Functions (what’s done), Things (what’s used), People (who does it), Time (when and how long it takes), Locations (where it’s done), and Motivation (why it’s done). No model includes all of those dimensions, so you want to use the model that includes the dimension (or dimensions) critical to the problem you’re trying to solve or the opportunity you’d like to pursue.
There can be five relationships between those dimensions:
The dimensions that you’re interested in will probably drive the relationships you’re interested in. Models that focus on people, for example, are frequently interested in the reporting relationships between those people or the activities those people perform. But, as with the model’s dimensions, you’ll want to pick a model that incorporates the relationships that matter to your problem or opportunity.
For example, an organization chart has as its dimensions people and the departments those people belong to; An organization chart’s relationships are reporting (who has authority over who) and aggregation (what departments go together to make the organization). You can see why an organization chart is the default modeling tool that people use when they’re interested in finding a way to streamline the organization’s relationships among its people.
On the other hand, a business process diagram has activities as its dimension and sequence as its relationship; A business process diagram shows what activity follows another activity (and which activities are performed in parallel). If you’re ever drawn a diagram to describe an existing process so that you can find inefficiencies or opportunities for improvement in that process, you’ve probably created something very much like a business process diagram.
That doesn’t mean that a business process diagram is the best tool for your problem, however. For example, if you’re interested in who performs a task in the flow so that you can ensure that the task is assigned appropriately, you’d want to integrate the people dimension and the activity relationship. In that case, a better choice for your model would probably be a responsibility assignment matrix. One of the benefits of a responsibility assignment matrix is that it provides instant visual feedback as you move responsibilities around, allowing you to ensure that you don’t overload (or “underload”) the people involved.
On the other hand, you may be more interested in looking at the dimension of time in your process and how the process develops. A state diagram lets you look at what triggers the next activity or change in a process so that you can improve issues in those areas.
Picking the most useful model is only the first step, of course. You’ll need to do the research required to gather the real world data that will let you build your model by filling it in with the correct information.
But even after you’ve picked and built the right model, you shouldn’t expect your model to accurately match your real organization…at least, not on your first try. No matter how many people you talk to and how much research you do, you’ll probably have left out something that matters or failed to give enough weight to something you’ve included (and you may have included components that don’t exist or don’t matter). You’ll always need to validate your model against the knowledge and experience of the stakeholders who are familiar with the reality you’re modeling.
Don’t be discouraged if your initial feedback shows that you need to make a number of changes to the first version of your model. The first version of your model will probably be only about 60% accurate. However, once you incorporated that feedback and done a second review, you’ll probably find that your model is about 85% accurate (according to your stakeholders). Incorporating the feedback from a second review, will probably give your model a 95% match to reality — that’s probably “good enough” for you to start using it.
That 95% accuracy is predicated, of course, on you picking the right modeling tool to begin with. If you get that right, you’ll find solving your problem or grasping your opportunity much easier to do. If you get the wrong model, however…well, just make sure that you’re being paid by the hour.