Is Optimized Supply Chain Planning Useless? Not If You Understand These 6 Points.

“I am mad as hell, and I’m not gonna to take this anymore!” #SlayingSCBS

This post has been brewing in my mind for quite a while as I continue to hear misrepresentations of the role and value of optimization in supply chain planning processes. “Doesn’t work”, “Isn’t agile” and other smears have been bandied around for too long and represent a lack of understanding by the folks making those claims. Dig deeper, you find that the reasons for these misrepresentations have nothing to do with the technology itself, but rather how it is used and maintained. My concern is that these kinds of comments confuse supply chain leaders and cause them to underutilize a critical part of the supply chain tech arsenal.

Source: Supply Chain in 5 and ChatGPT

Here are six-points supply chain leaders should understand to be successful with optimized supply chain planning¹ solutions.

Supply chain plans are temporal. Optimized planning can create the best plan at a given time. However, the quality and relevance of the plan degrades over time because many of the underlying assumptions change. As businesses become more dynamic, planning cycles need to be compressed.

Take strategic planning as an example. Most companies have strategic planning cycles that are measured in years. What businesses today are static for that length of time? None! It doesn’t matter how good the optimization technology and planning process are. Strategic planning processes don’t “work” when they are not synched to the rate that the underlying assumptions and business conditions change. Today’s best-in-class strategic planning processes are operated continuously to provide financial and operational guidance as business conditions change or new strategies are proposed.

Data quality and timeliness matter. Every hear the term “garbage in, garbage out”? This is especially the case when using optimization planning solutions. Data-related issues usually happen over time as data for customers, products, service and operating policies, etc. are not maintained. The optimization results degrade, crazy manual work arounds are put in place and ultimately, the users say that the system no longer “works.”. To learn how to address this problem read: How Well Do You Train Users AFTER the Initial Implementation?

One optimized plan isn’t enough. Instead, scenario planning must occur at all levels of planning. The longer the planning horizon, the more likely base assumptions will change, so it is imperative to consider the impact of potential changes in advance. Scenarios for competitive threats to major customer acquisition or loss, etc. can be optimally planned and a “play book” assembled to best guide the business when these changes happen.

Optimized scenario planning also provides insight into the business. For example, scenario planning can be applied to very short planning horizons such as daily planning. Producing three plans at the same time – one optimized for customer service, another for lowest cost and a third a balance of cost and service lets supply chain leaders make more informed decisions and trade-offs on a daily basis.

There are lots of kinds of optimization technology, every one of them has its limitations and none are perfect. Optimized planning solutions use mathematical models to solve particular business problems. The fact that they work well in one planning process doesn’t mean they will perform as well in another. The problem could be the base algorithm, but equally it could be the ability to model all of the operating, product, or customer requirements. It’s why the optimized planning market is so fragmented. The same is true when incorrectly applying the technology to the wrong type of planning problem such as using operational planning solutions for strategic planning purposes.

Even when the optimized planning solution is the right fit for the business problem, it will not address all of the planning scenarios. Optimized planning is an empirical science. Solutions are based upon mathematical models, built out based upon the knowledge of the product experts and existing customer requirements. However, even best product experts cannot consider all of the planning scenarios for the product or are not exposed to all possible requirements.

Supply chain leaders should expect that there will be edge conditions that the optimized planning systems do not address. Failing to address this fact with optimized planning users create unrealistic expectations, undermines confidence in the system and ultimately poor utilization and lower value delivered.

Optimized planning delivers the most value when the strategies, processes, etc. around it are optimized too. Yes, you can get some level of value if you insert optimized planning solutions into existing strategies and processes. However, understand that you are missing the chance to get much more. Optimized planning can’t fix outdated or broken processes, operating assumptions, strategies, tactics, etc.

For example, let’s say that your order cut off time for next-day deliveries was 3pm because the existing planning process took four hours. However, the new optimized planning solution that was installed cut planning time to one hour. Wouldn’t it make sense to move the order cut off time to 6pm to capture more business and better serve customers? There are numerous examples like this where optimized planning is the enabler, but the incremental value only came when things around it changed.

It's not optimization vs AI. It’s optimization + AI. Optimization and AI technologies are synergistic and there are many forms of AI that can add value. For example, machine learning improves the accuracy of optimization results by improving parameters and based data such as stop times, drive times, lead times, customer locations, replenishment quantities, and on and on and on. AI changes how users interact with optimization technology during the planning process and even configuring the solution to achieve specific business objectives. These are just some of the many ways AI is augmenting optimized planning.

Adding to the confusion about optimization and AI are marketing “weasel” terms like “AI powered” or “AI driven” planning. In no way do they describe what role AI plays and actually cast doubt on the use of AI. It’s best to get your current or perspective prospective vendors to explain exactly how AI works in their optimized planning solutions.

Unfortunately, the value of optimized supply chain planning solutions is one of those areas where there is an overabundance of bad advice. In a world of heightened customer expectations and collapsing margins, optimized planning is more important than ever to provide a superior customer experience at the lowest possible cost. To get the most value from optimized supply chain planning, supply chain leaders need to understand what optimized planning can and can’t do and the strategies, tactics, and processes that best exploit its capabilities.

Foot notes:

“I am mad as hell, and I’m not gonna to take this anymore!” is from Peter Finch in the movie “Network”

1 Optimized supply chain planning includes strategic, tactical, and operational planning, and has scopes that could be supply chain wide or functional (demand planning, sales & operations planning, inventory planning, transportation planning, etc.).