If you are a supply chain executive, chances are that your organization already owns a solution that is either a dedicated “control tower” application or has a “control tower” module. So here is a little exercise for you. Removing all the jazz aside (like when a planning element uses MILP or AI-enabled algorithms), list your control tower’s functionalities.

Chances are that you will find (give or take a few), the following:

• Monitor supply and demand (may or may not be real-time)
• Identify risks leveraging real-time data
• Evaluate resolution options

You may have come across many jazzy buzzwords when the solution was sold to you, but at its core, all control towers do the tasks mentioned above. Now that you have the list let us move to the core premise of this article.

Again, keeping all the buzzwords like “intelligent evaluation, AI-enabled learning from the past” yada yada. You had all those capabilities within your portfolio already. Probably not in one single application, but you had. If you want to say that you did not have “real-time” data on your supply and demand, you may be wrong. Your “control” tower solution does not provide more real-time data than your ERP solution. They will all capture demand and supply through transactions. The latency comes from the data management infrastructure, not the application.

Let us come to the algorithms now. Many aspects of supply chain planning may involve Big Data elements. Unless you are FedEx or UPS, tracking billions of packages worldwide, your demand and supply data is not precisely Big Data (even though you may like to call it that). But even if your business generates that massive data quantity in near-real time, the challenge is in the data management infrastructure, not the algorithm side.

If you already have the appropriate and optimal data management infrastructure in place, you can build, train, and deploy an actual AI algorithm that does a far, far better job than any generalized algorithm in an off-the-shelf solution. Guaranteed!

So essentially, as you may see, you already had the capabilities that your control tower promises. But let us now explore the term “control.” While I understand that you can use all the tools packaged together to leverage your processes and people to control the supply chain, the solution does not control anything at all. Nothing. Nothing at all. It gives you data and tools to make sense of it, get insights from it, and then, YOU (your people, enabled by processes), take control.

If, at this point, you think that all I am saying is that these solutions should not be named “control towers,” then I may not have been clear. I insist that if absolute control is what we seek, there are better ways to build an actual control tower. Without human intervention, there will not be a control tower in my lifetime that will seamlessly control the entire supply chain. But with your current investments, your current portfolio of solutions, and inexpensive open-source tools, you can build control towers that actually have control elements.

As I mentioned in an article I wrote in early 2023, the key lies in the data strategy. If you formulated your data strategy with a forward-thinking vision, then exploded those visions into data capabilities you need and built an architecture accordingly, the rest is (relatively) easy.

The challenge is more painful for the solution providers. There is a limit to gaslighting and buzzword throwing. The barriers to building customized solutions have almost hit the floor. And that is why I insist that enterprise technology companies must look at their business from a different perspective.

If you offer a robust foundational platform and associated capabilities to build plug-and-play capabilities on this platform, you do not have to get into the rat race of trying to sell yet another supply chain planning solution in a crazy market with almost no differentiation.

But the differentiation you can build will come from the fact that a customer can build a planning tool and function, using your platform, plug-ins, open source tools, and partner solutions, within weeks, irrespective of function. Again, if the appropriate data architecture exists.

Why do solution providers don’t take that route? Because it requires you to sell individual stories to every customer vs. one single marketing narrative for an off-the-shelf solution. This lack of skill set to show the customer how to build a customized path using the foundational blocks offered removes the differentiation from the solutions space. Hence, any company that develops this capability will be able to showcase how they can help customers build something unique, customized, and valuable that no other off-the-shelf solution can match.

And if you can build teams and teams with this skill set, you can disrupt any category of enterprise software market.

It is all about a solution, not a software !