Success Impaired: Common Pitfalls of S&OP and Supply Chain Planning Technologies...And How To Avoid Them
The need for supply chain flexibility and resiliency has been a common theme in recent months, and for many companies, the ideal qualities of ultimate flexibility and maximum resiliency are much more realistic than they imagine. While each quality has multiple enabling components, supply chain flexibility begins with having timely knowledge of options, impacts, and risks, and the ability to use that information to rapidly implement a new, optimized operations and supply chain plan. Sales & Operations Planning/Integrated Business Planning (S&OP/IBP) and detailed production scheduling systems are among the supply chain planning technologies that are key to achieving the goal of maximum flexibility. These technologies have helped many companies improve supply chain operations and deliver positive business results to varying degrees, but they have also too often fallen short of providing users with the full insight and maximum flexibility possible thus leaving significant operational and financial benefits untapped.
The solution to achieving a large part of that goal of ultimate flexibility is not an extremely difficult or costly one—and it can be achieved with supporting technology that has been around for the last 20+ years. The root of the issue is in how the a specific S&OP and detailed scheduling technology fits (or doesn’t fit) with a particular operating environment, but this “fit” problem is driven by an extensive amount of misleading assumptions, ratings, and general information related to these technologies—with process industry companies being affected the most. This has resulted in many of the affected companies undertaking time and resource consuming implementations only to realize that they are still far from reaching their target state of operations insight and flexibility and are not achieving the maximum operational and financial benefits they might have envisioned.
Some organizations that experience this dynamic face critical usability obstacles early-on, while others may still achieve supply chain planning improvements, but those improvements are tainted by gaps in what stakeholders expected versus reality. In these cases, early frustrations are coupled with the sobering understanding that the political and financial cost of a near-term change is out of the question, so stakeholders resign themselves to accept the less than desired end-state and manage around any flaws. Meanwhile, a third set of organizations may view the incrementally improved state as at least a moderate success without fully appreciating the untapped potential left on the table due to supply chain technologies that are ill-suited for that environment.
These situations often result from subtle but critical nuances in supply chain technology solutions that can either greatly enhance or insidiously limit supply chain planning and optimization outcomes. The nuances are typically part of the core software architecture and design such that minor configuration changes and tweaks will not resolve the limitations. In order for stakeholders to select and properly implement solutions that are well-suited for the environment thus enabling maximum operational and financial benefits, they must understand what these nuances are and the business outcomes that they drive. Addressing these nuances and eliminating their limitations is also a prerequisite to achieving many of the promises and benefits of Industry 4.0.
Let’s explore more thoroughly what ratings, guidance, and assumptions I’m referring to, why they are often misleading, the outcomes they drive, and some methods you can use to help nail your next supply chain planning enhancement and unlock truly exceptional operational and financial benefits.
Misleading and Incomplete Guidance
This misleading and incomplete guidance is not a new problem; it manifests in the most vaunted and referenced sources like Gartner’s Magic Quadrant, extends to many supply technology vendors, and even includes some technology implementers and supply chain stakeholders. As we examine this multifaceted problem, let’s start with Gartner’s Magic Quadrants.
There are a variety of reasons why questionable recommendations from such IT powerhouses like Gartner might exist. While there likely are several organizational and analyst-specific drivers, one structural factor in Gartner’s case was their combining S&OP-related ratings for process industries, discrete manufacturers, and resellers/distributors, obscuring many of the critical differences in technical requirements among the three different groups.
Around 2005, Gartner separated its S&OP-related Magic Quadrants by creating special ratings for Process Industries and Distribution Industries. I considered this appropriate given the unique constraints and requirements specific to process manufacturers that can get overlooked if the technology ratings for process, discrete, and distribution/reseller industries were combined. However, by 2012 Gartner’s S&OP-related Magic Quadrants had done just that. As illustrated in Diagram 0.1, Gartner published special supply chain planning-related Magic Quadrants unique to these specific industry types from about 2004 through about 2010. Since then, Gartner has combined them into one Magic Quadrant with no distinction between industry types for any aspect of supply chain planning.
Diagram 0.1 – Gartner Supply Chain Planning-Related Magic Quadrants Over Time
The combination of industry types into a single Magic Quadrant laid the structural groundwork for these critical factors to be overlooked. While I appreciate that there are possibly some arguments in favor of combining the ratings, in doing so, Gartner would ideally highlight industry-specific factors more prominently in the narrative and further expand on them in client-specific discussions with their technology analysts. However, the combined Magic Quadrants have been overwhelmingly silent on these critical factors and, in my experiences, discussions with Gartner analysts did not yield any comments or concerns related to these criteria.
But Gartner is not alone in contributing to the lack of focus on these critical factors, although they have one of the highest profiles and carry a substantial amount of weight. In reality, there are hosts of contributing parties and circumstances including technology vendors, some supply chain practitioners and technology implementers, and even many buyers themselves. The inherently esoteric nature of this issue compounds the problem, and possibly even enables it from the very beginning. As a result, many buyers make technology decisions that miss essential criteria and impacts operational and financial outcomes in real and significant ways. Let’s take a moment to delve into exactly the what this issue is, what the misleading guidance entails, and how it affects business outcomes.
The Core of the Problem
At the highest level, the issue involves how well a vendor’s solution can model the buyer’s production environment. More specifically, it relates to the constraints, options, and permutations which themselves become more intricate in complex production environments thus requiring a more advanced level of software design and functionality for a system to properly account for these conditions. Meanwhile, too many supply chain planning technology vendors offer solutions with a level of design and functionality that renders them incapable of properly addressing many of the real-word operational issues that they profess to solve and even optimize.
Who cares about software design so long as it appears to work, right? Well, that’s where this issue gets a bit complex and esoteric and that’s why it has continued as a problem for so long. The details quickly get into technical and seemingly intangible concepts that many stakeholders overlook, minimize, or simplify to a fault, although they have very real and tangible business impacts.
Given that you’re reading this, odds are that you’ve already received a plethora of articles, emails, or opinions about the importance of supply chain resilience and flexibility in the wake of the coronavirus pandemic. This issue goes to the very heart of resilience and flexibility, particularly for manufacturers. In addition, odds are that you’ve also heard plenty – bordering on hype – about Industry 4.0 and IIoT (Industrial Internet of Things), how it will revolutionize everything under the sun in the next several years, and how you’ll be left out if you don’t start doing “it” today….right now, even. Fact is, there are additional real and substantial operational and financial benefits that Industry 4.0 and IIoT can deliver for all companies, but for manufacturers, those benefits are fundamentally dependent upon resolving this issue and properly modeling your production environment.
Does it sound like we’re getting into “digital twin” territory? Well, yes and no. The supply chain planning systems of the last 20 years were not necessarily intended to be proper digital twins in the way we speak of digital twins today. However, if the systems were designed well and properly matched to the buying company’s operating environment, they could have, at a minimum, served as a gigantic step toward that just by doing what they were theoretically supposed to do. While a “digital twin” was not (and is not) the objective or my focus in this article, it would have been (and will be) a by-product of selecting the right technology, implementing it well, and improving it over time.
From a business results perspective, what if a company could detect and precisely react to small, medium, and large changes in the market on a monthly, weekly, or even daily basis? What are the operational and financial benefits of having that flexibility and resilience—not just within your core production environment but up and down the supply chain? And what if, you could strategically harness the relevant benefits of Industry 4.0 and IIoT technology enabling your entire operation to cost-effectively turn on a dime in executing whatever strategic direction the CEO, COO, and other leadership required?
So those are the stakes. Now, let’s get back to the issue.
To reword the problem, the ability of a given supply chain planning solution to properly model scenarios, truly optimize outcomes, and offer powerful options directly contributes to and enhances the ability of company leaders to refine strategy, make accurate, well-informed, up-to-the-hour decisions, and quickly and precisely execute those decisions. Yet that supply chain planning solution’s ability to model, optimize, and provide powerful options is heavily contingent upon its ability to accurately account for the real constraints, options, and permutations that exist in the production environment which is the core of the business for manufacturers. And a solution’s ability to accurately account for constraints, options, and permutations—particularly those that exist in more complex production environments—depends on a relatively advanced, insightful, and purpose-driven software design and logic. This is represented in Diagram 0.2.
Diagram 0.2 – Technology, Planning, & Leadership Dependencies
Yet, in too many cases, this doesn’t happen. More often than not, a supply chain planning technology’s software design and logic is actually only capable of superficially modeling a subset of many production environments and, even within those limited parameters, only offering sub-optimized scenarios and options even when they say they’re optimized. Wondering if this applies to your company? Even if you chose the supply chain planning technology offered by a large, ubiquitous vendor or went with a vendor that is a Gartner Magic Quadrant “Leader”, odds are, it’s still affecting your company—especially if you are a process industry manufacturer or have a reasonably complex production operation.
What’s Causing It?
At a high level, what’s happening internally for many of these systems is that the design and logic deployed is rudimentary to a fault resulting in a system that’s functional on the surface and with its scope limited but which lacks the ability to handle the technical complexity required to model and optimize more complex production environments. As a result, the systems use rudimentary solving methodologies and require “customizations” (what I later refer to as “artificially adapted models”) to enable some functionality that otherwise should have worked out of the box. These customizations or “artificially adapted models” can make the solution even less flexible and less accurate. On the surface, it works (in a limited manner) and usually provides some benefit (within relatively narrow parameters and over a limited scope), but it falls far short of the full capability, flexibility, and resilience, that would drive even greater outcomes and which many buying organizations originally envisioned.
This might beg the question why so many supply chain planning solutions lack the ability to handle the higher technical complexity associated with more complex manufacturing environments. While I don’t have a certain answer, a few themes emerged after speaking with several colleagues and vendors:
- Developer Frame of Reference. Many vendor’s leveraged junior developers or developers with no operational frame of reference to design and build their supply chain planning solutions, thus they inadequately account for production complexities that would otherwise require 20+ years of practical knowledge to fully appreciate. If the vendor doesn’t recognize and prioritize incorporation of those complexities, the solution ends up too rudimentary and ill-suited to function well in moderate-to-highly complex production environments.
- Economic Disincentive. Many supply chain planning technology vendors charge for their solutions by headcount. However, process industry (or “continuous”) production operations typically have less headcount than discrete manufacturing operations. Therefore, the added technical complexity required in modeling process industry production operations receives less attention from vendors.
- Competitive Disincentive. IT Leadership’s bias towards standardizing around the ERP vendor gives some ERP vendors a significant advantage in selling their supply chain planning solutions. Also, incumbent ERP vendors have greater access to and sway over the buying company’s stakeholders. Each of these factors can conceivably create less incentive for the vendor to invest the time and expense into addressing any shortfalls or technical nuances that may plague their solution. Of course, this would only apply to the subset of vendors that offer ERP systems in addition to problematic supply chain planning solutions.
The answer could lie in either, some, or all of these possibilities, but whatever the actual reason(s), the result is that there is a relative small number of supply chain planning technologies with the capability to fully address the process industry and other more complex manufacturing environments.
To make matters worse, the differences between the ill-suited or incapable solutions and those that are well-suited and fully capable are obscured with marketing lingo, imprecise verbiage, elegant websites, and sales pitches that can make one solution practically indistinguishable from another. With buyers already desensitized to this issue, the critical details that differentiate the solutions often seem overly technical to some stakeholders. As a result, the marketing is effective and buyers decide to purchase ill-fitting solutions with stakeholders often left to discover sizeable gaps or substantial limitations later.
While I have described the issue at a high level, in the next sections I’ll disassemble the different facets of this problem and explain each in detail to convey more specifically what the problem is, how to identify it where it exists, and how to avoid it when selecting a supply chain planning solution. But before diving into the details, let’s look at a real-world case study to illustrate how this issue commonly unfolds and how it can affect business outcomes.
Real World Scenario: Case Study
My interest in and struggle with issue dates back over 15 years when I was working with a client—a $2.5B consumer products manufacturer who had a process-type manufacturing operation. This case represented the beginning of my ongoing quest to highlight and articulate these critical but far too often overlooked details.
Background. The client’s burning platform was that late deliveries were becoming increasingly problematic with key customers, and the client initially perceived that their hectic and often late production was due to insufficient production capacity. Their solution was to build up extremely high levels of inventory during slower periods of the year, essentially hoping to brute-force their way through the busier periods. I assessed the client’s current production schedule, inventory levels, and delivery rates and created a future-state model showing the anticipated production output, new inventory levels, and on-time delivery rates with improved forecasting and planning processes and better production scheduling methodology. The model illustrations demonstrated that the improvements would radically increase on-time deliveries, keep inventory at relatively moderate levels, and deliver substantial improvements to financial results—all without the capital investment in new production lines which the client had already planned as a solution to their perceived capacity problem. Given the model and business case, the client agreed to implement a new S&OP/IBP process, the more effective production scheduling methodology, and supporting supply chain planning technology.
Buy-In & Technology Selection. The basic S&OP process wasn’t completely dependent on supply chain planning technology, and the scheduling methodology that resolved the “burning platform” problem could have been resolved more cheaply with a custom application that might have 5-10 days to create. We decided to pursue more formal supply chain planning technologies because they could support the new scheduling methodology thus resolve the burning platform problem while additionally: (a) creating more flexible, optimized schedules in less time thus enabling the plant schedulers to support the plant managers in more strategic capacities, (b) enable the plant schedulers to model changes, incorporate R&D and maintenance requirements or changes proactively, and re-generate schedules quickly, (c) provide better tools to support the S&OP/IBP process and enable all the benefits of modeling, optimization, and schedule regeneration across all plants laying the groundwork for the client to progress to a higher level of supply chain maturity, and (d) address IT’s desire to avoid yet another custom in-house solution that they would have to support.
All was going well as everyone from senior corporate leadership to plant managers to R&D and maintenance were brought on board—sometimes even with a bit of hard-won enthusiasm given the anticipated benefits. The plan was to implement the new S&OP/IBP processes and technology at Plant #1 and then roll them out to Plants #2 thru “N” in Business Unit A and then do the same in Business Units B and C. All three business units had process-type manufacturing operations. We began screening supply chain planning technology vendors and quickly shortlisted three vendors for more detailed reviews. My ultimate guidance and strong recommendation to the client was that the solutions from either Vendor A or Vendor B would work well, just avoid Vendor C. The client ultimately chose Vendor C.
What’s key is the rationale and missed opportunities. The client chose Vendor C primarily because the corporate logistics stakeholders on the selection team were most pleased with Vendor C’s presented logistics abilities and that portion of the team ultimately convinced the project sponsor that it was the better vendor choice for that reason. The problem was that Vendor C’s software design was fundamentally incapable of modeling many of the unique constraints that existed on the client’s production lines and therefore the planning, scheduling, and optimization capabilities—the core functions that would help to address the most critical and central needs—would be compromised at best, and rendered inoperable at worst.
As much as I attempted to convey this concern to the client, they ultimately made their decision based on the wrong set of factors. The specific nature of the constraints and why they were important was challenging to articulate in a manner that was relatable and convincing to the project sponsor. That combined with the heavy influence of the logistics stakeholders and the back-channel which Vendor C had to those logistics stakeholders short-circuited the decision logic. My conundrum was that explaining the issues in too technical or detailed terms would return glazed-over eyes and looks (if not actual words) that said perhaps I was aiming for more sophistication than they actually required at this point. However, explaining the issues in terms that were too general and insufficiently detailed would frustrate attempts to differentiate the solutions in the eyes of the stakeholders and enable Vendor C to simply insist that they meet the requirement—though I suspected they were fully aware of their limited capability in this case.
Implementation. During implementation, I quickly was proven correct in that the software could not account for certain critical constraints that existed on the production line as I expected. It was a Pyrrhic victory though. Having now assumed the role of implementation project manager and having made my objections as clear as possible, I essentially saluted by this point and did my best to make the client’s bad technology decision work. To rescue the decision as much as possible, I had to have the vendor bring two of their core platform developers on site and guided them as we re-coded (not configured) portions of their software to handle the key constraints. There was still only so much that I could reasonably achieve given the core software design. To preserve all functionality, this project “audible” would go from being a multi-day software patch to re-writing practically the entire application. Ultimately, the limited patch was successful in getting it to factor in those critical constraints along with the new scheduling methodology in a first pass. But by factoring in those constraints without recoding the entire application, it rendered the solution’s ability to dynamically adapt the schedule to subsequent changes, model multiple scenarios, and optimize inoperable.
Outcome. While we successfully resolved the main issue with the new S&OP processes and scheduling methodology, many of the enhanced planning, scheduling, cross-plant modeling and optimization, and process benefits anxiously anticipated by the plant manager, scheduler, R&D, maintenance, and others were lost. In fact, given the manual adjustments that were necessary after the technology solution created the initial schedule, the plant scheduler could say that that the new technology-enabled scheduling process was actually more time consuming than when he simply created the schedule manually. Subsequent plants likewise looked forward to any change much less enthusiastically and possibly even resisted it to some degree. And while the client’s operational and financial benefits technically improved as compared to the previous mode of continuous fire-fighting and massive inventory build-up, the benefits they realized were still far less than what they could have achieved if they selected one of the vendors whose solutions were appropriately designed to account for the unique production constraints. To add insult to injury, the constraints that posed the problem were relatively simple given the range of constraints that commonly exist in process industry production operations.
Since then, I have been keenly focused on conveying these nuanced but critical modeling and constraint handling differences in ways that client stakeholders and business leaders can fully appreciate so they avoid making ill-suited supply chain technology selections that bite them later. This case was not the only example, but it was the first. My interest in this issue has only grown with additional cases that I’ve experienced myself, discussed with others, or read about where companies make similar mistakes—choosing ill-fitting supply chain planning solutions only to realize the limitations later.
As promised earlier, let’s get into the specifics of the issue and how it affects outcomes, how to identify the problem, and how best to avoid it. Since it is a multi-faceted problem, I’ll dissect it into components and explore each based on five different but compounding principles and then wrap up with how to make the right choice the first (or next) time.
- If a solution cannot adequately model a production/operations environment, then it cannot create realistic scenarios and it cannot optimize those scenarios.
- Solver algorithms add an obscure boon or barrier, but true optimization and synchronous solving approaches deliver the best outcomes.
- Different industries have different levels of production complexity and therefore require different levels of production modeling capability.
- Accurate modeling and optimization capability should be the primary factor driving a detailed scheduling or S&OP/IBP technology purchasing decision for manufacturers.
- The entire production operation should be accounted for when selecting detailed scheduling and S&OP/IBP layer technologies.
> Next Section: Part 1 – Realistic Scenarios Require Accurate Modeling
< Previous Section: Executive Summary