Executive Roadmap: Integrating Simulation Modeling into Your Business

In the previous blog, I defined the meaning of the words "Enterprise" and "Simulation" in the context of the Enterprise Simulation concept. We concluded that applying the idea of simulation modeling on an enterprise-wide scale enables prospections into different futures and unlocks the exceptional potential of knowing the future in advance.

In this blog, we want to dive further into the idea of Enterprise Simulation and share our decade of experience with the different implementation paths that our team has explored and implemented with a significant number of mid to large-cap businesses.

What is Enterprise Simulation?

I expect your first impression of the term “Enterprise Simulation” to be either:

1. One All-Encompassing Model (Monolithic Approach): A type of simulation model that covers the entirety of the business core in one encompassing model.

Or, if you are thinking a bit more modular:

2. All Part of a Whole (Modular Approach): A collection of simulation models, each representing a part (node) of the business, and together they cover the core aspects of a business.

So, I think that you have already imagined one or both of the definitions above and have a pretty solid understanding of Enterprise Simulation before I even start explaining this term and why we felt the need to coin it a while back. It might be obvious that both definitions point to the same goal: a dynamic “causal” model (or set of models) that, either in one model or collectively, covers the core elements of an enterprise or business.

Monolithic Approach – Usually the Dead End

The first definition, in practice, is not achievable yet! Although it is very desirable to have one giant model that includes everything, in practice, we hit the limitations of current general-purpose simulation engines that were not originally designed for such a scale. Even the most scalable engines, such as AnyLogic, which considered scalability from their inception, had a much lower target. Without delving into the technicalities of how to quantify the scale, and just to give you an intuition about it, imagine that even the best-in-class general-purpose simulation engines are designed to deal with thousands, not millions or billions, of objects, events, etc. So, as I mentioned, in most cases I have seen over the years, this approach is impossible until we have a significant breakthrough in simulation engines (I will write about this topic in the future).

Just not to leave out any possibility, I have to mention that if you are willing or know that a higher level of abstraction can solve the computation problem, it is theoretically possible to model even the entire world in one model—but that requires many stars to align! In practice, such high-level models can be the steppingstone for the implementation of a more detailed version and to test some architecture before fully committing to building the modules. Our team of experts is available to evaluate this path for you based on your requirements and strategic goals.

 Modular Approach – The State-of-the-Art

One thing that may not be obvious is that the second definition is actually a more natural progression toward achieving this goal within the current constraints. Let me explain this in more detail.

In the evolution of the idea of Enterprise Simulation, regardless of your ultimate scale, the natural starting point is to model one of the nodes first—such as one fulfillment center, one data center, or one restaurant location. Usually, people start here because there is a pain point in that specific location, or it is a significant node that is worth improving or attempting to optimize. Over the past ten years, I have encountered thousands of teams from mid-size to mega-size corporations at this stage.

Veterans already know this, but newcomers exploring the field quickly realize that building a useful model is likely significantly more involved, time-consuming, and resource-constrained than they initially expected! So, there are two paths from here:

1. Focusing on solving pinpointed and immediate problems, not at scale or at the enterprise level. In this path, if you are only thinking about creating a few models per year and do not want to invest in building an internal team, it is very likely that it makes more sense to hire consultants to deal with the issues, get the results, and end the engagement until a future similar need arises. I have also seen internal consultants (center of excellence teams) build many of these types of quick and useful models based on the immediate needs of their business, but they are still not investing in a larger-scale encompassing approach.

2. The other option is to be more ambitious and, while being aware of the challenges ahead, increase conviction in the usefulness of simulation and try to justify the required efforts at scale! In other words, the company decides that they want simulation and clearly see its expected return, but since it requires significant investment and mobilization of resources to do this in-house, it is better to do it at scale and for significant nodes of the business that are naturally very similar, to get the targeted ROI with this approach. This path is viable both for existing internal teams that are used to small-scope models and now want to outgrow and cover more of the business, as well as for executives with higher convictions who are thinking about taking the scalable path from the start.

Regardless of whether you are considering monolithic or modular models, if you have an existing internal team, are thinking about hiring a consultant, are already engaged with consultants, or are thinking about building an internal team, we can add significant value by being your advisor and sharing our experience and technical know-how with you. If you are interested in knowing more about your options and the experience-tested approaches, please schedule a call or ask for more information on our Contact page.