Bridging the Reality Gap Between Planning and Execution Part Two: The Manufacturers' Perspective

The Manufacturers' Perspective

Despite the fact that many manufacturers have invested in enterprise resource planning (ERP) systems and supply chain management (SCM) systems, most continue to use inopportune batch reports and pesky spreadsheets to manage their operation's performance. These have proven to be inefficient and error-prone methods of supporting decision-making, resulting in a reliance on "educated guesswork" rather than on accurate dynamic analysis to align decisions with strategic objectives. For that reason, some innovative enterprise software vendors intend to do the same for manufacturing and operational decision-making as has already been done to financial decision-making by some business intelligence (BI) applications. (See Financial Reporting, Planning, and Budgeting as Necessary Pieces of EPM).

Manufacturers today need to react quickly in order to remain efficient and competitive, given that the biggest problem they face is that change is the only constant in manufacturing. For those who are lucky, only minor changes will happen between the "as planned" and "as executed" worlds. These changes are the usual minor but endless variances that exist between planning and forecasting in the "ideal world" and manufacturing in the "real world." For all the investments spent on sophisticated supply chain planning (SCP) and manufacturing-planning tools, almost proverbially, the only sure thing about a forecast is that it will be wrong, by and large. Years ago, in business, many minor variations could be ignored as margins were sufficient to accommodate many suboptimal decisions or manufacturing processes, such as keeping increased safety stocks or frequently expediting. But in today's world of often razor thin margins, the variances between plans, real customer demand, and forecasts must all be spotted, and manufacturing success can be determined by the speed and effectiveness of the response to even minor changes.

Sometimes, however, the more serious changes, which disrupt routine tasks of planning and forecasting, manufacturing, delivering, and invoicing standard orders, result from significant unplanned events that occur either within the company or elsewhere throughout the extended supply chain (e.g., a supplier is late, supplies are wrong or of unacceptable quality, a manufacturing line is still unexpectedly busy or is down, there is a last-minute order change, etc.). The risks can be high, ranging from lost margins to loss of customers or erosion of competitive market positioning. Speed and effectiveness in response to these changes can determine the difference between profit and loss in manufacturing an order, or they can determine margin size in a manufacturing cycle or run.

In addition to the above minor and major changes, manufacturers face a third and typically more powerful change to the business itself. Namely, to remain viable and prosper in today's ruthlessly competitive global markets, manufacturers must continually reinvigorate themselves and even occasionally reinvent themselves. Fundamental business changes range from new product introductions and promotions to new corporate goals and objectives; to new target customers, continually changing customer demands and new supply chain partners; and even through to merger and acquisition (M&A) activities and structural changes to the business itself.

Some manufacturers try to cope with possibly the most difficult yet pervasive of all forces in business—never ending change—by trying to leverage the strategic information systems they have been using to control, manage, and optimize the "as executed" state of their business, such as ERP systems, or SCE systems available respectively from their ERP vendors and a broad array of pure play SCE players. However, it has been extremely difficult for them to deal with day-to-day events in real-time with a transaction-based ERP system that relies on batch processes to extract the data they need. It can be daunting for a manufacturer to get real-time notification of a supply chain event, understand its impact, and take the appropriate action given traditional ERP and or execution applications do not really provide this capability. Almost every enterprise has huge volumes of information flowing through their enterprise systems, but only few possess the tools to quickly exploit their wealth of data and thereby optimize their operational and corporate performance.

Most manufacturers will have gone a step further in addressing this problem by deploying some level of strategic planning or forecasting systems to help them forecast and model the "as planned" state, in solutions ranging from a variety of business intelligence tools, such as Cognos, Hyperion, SAS, or Business Objects, all the way through comprehensive SCM solutions, such as those from Manugistics or i2 Technologies, or again from their ERP providers. If one considers traditional business intelligence tools, they merely do a prodding analysis of historical data after the fact, but one cannot count on the future to look like the past, which has been the shortcoming of some forecasting methods as well. Therefore, more up-to-date information is a requirement.

On the other hand, best-of-breed SCP solutions may not be the answer, as they require a significant investment in both software and integration. Sometimes the solutions to improve manufacturing effectiveness are so complex and costly that they overwhelm any benefits that they might provide, such as when engaging consultants to, for example, scrutinize modeling revenue, cost, and supply chain capabilities, with breaking products into families and analyzing the channels they are sold through and the geographies they cover. Frequently, the exorbitantly high cost and complexity keeps companies from realizing the potential benefits that these systems promise.

Additionally, there are still significant barriers to an easy deployment of SCP systems, as they are based on cumbersome proprietary algorithms and heuristics that take a long time to master and harness to work, forcing companies to have full-time rocket-science' expert consultants on the premises to interpret the results and to keep the application in tune with the business processes it supports. Therefore, the use of a traditional APS method that is non memory-resident and latent in itself, as a basis for all decision-making, is becoming increasingly unsound. Due to the growing visibility of supply chain information, the necessity of SCM has also progressively become more the provision of real-time information. Still, early supply chain event management (SCEM) systems, while crucial to increase visibility and raise flags, have lacked the ability to figure out resolution processes in the applications and their subsequent impact on operations.

This is Part Two of a two-part tutorial.

Part One presented the Problem.

How Manufacturers Cope Today

As mentioned earlier, the unpleasant truth is that most manufacturers today cope with the ongoing problems of change through a combination of people, paper, and tools. For any attempt to align variations in forecast demand (i.e., planning) with the execution side of the supply chain, it is often not enough to detect and understand the variation, but it is also necessary to react promptly to either take advantage of an opportunity or minimize the negative impact of a demand exception. Traditionally, such events would send business analysts scurrying back to spreadsheets and paper-based manual processes to consider various ways to address exceptions. They might look at scenarios for alternate warehouse transfers, expedited shipping, working overtime, or even adding an entire plant shift. These hurriedly explored alternatives and then quickly driven decisions are usually a compromise or a suboptimal decision, given that spreadsheets are slow, often out of date, and that they provide different versions of the truth.

Sometimes there is one "kingpin" person, a true "power user", who has custom-built a large and complex spreadsheet, and who has to labor under time and business pressures to work and, often empirically, rework alternative solutions in search of an acceptable answer. Other times a broad team of people, both throughout the company and even across the extended supply chain (i.e., trading partners), will hastily develop a variety of alternatives, sometimes on the back of envelopes and sometimes with the help of various desktop tools to manage demand events, though these tools at their best create merely a "swivel chair" exercise (i.e., due to the lack of integration, one would go from, for example, an event alert to the back office ERP or SCE system, possibly identify a problem, then to a separate planning solution and run a scenario, maybe identify another plant or warehouse as a solution, go back and enter a transfer order, and so on endlessly and iteratively). This team then compares their data, assumptions, and potential solutions via the phone, fax, and e-mail.

The Reality Gap

Thus, despite the use of spreadsheets and other desktop tools, this "reality gap problem solving process" is still mostly manual and hunch-based, and thus hardly ever rapid or effective. But the problems of responding to the inevitable reality gap go beyond just minimal automation and insufficient speed and optimization. One of the most troublesome reality gap problems is the "multiple versions of the truth" situation, since multiple people working on the same problem scenario, both within the company and across the supply chain, will often have differing data sources. Even with the same data source, they often acquire the data at differing times and, as data inexorably changes over time, they will have different data, which in turn yields different versions of the truth, like in the Japanese classic movie "Rashomon". As a result, multiple people will likely each reach differing "optimal" solutions, whereby there will seemingly be more than one correct answer. Under the intense time pressures of the reality gap, a mostly slow manual decision-making process is further complicated by the "which version do you believe" problem.

Conversely, in a nearly ideal world, the strategic solution would start with a single version of the truth, then continually update this version of the truth with real-time data feeds, and finally drive a collaborative decision-making process broadly across the action team.

What vendors offer today is functional capabilities that tend to reflect the given vendor's birthright, so that the vendors coming from the logistics place exhibit strong collaborative planning capabilities in transportation management, but are typically lighter in demand planning or in intelligent response to constraints and exceptions, and vice versa. On the other hand, the existence of so many vendors, with so many nuances in offerings, presents a challenge both to them in differentiating their products and to the users in selecting a vendor.

Indeed, many vendors can enable companies to efficiently manage trading relations, demand management and fulfillment processes. Companies such as Webplan, Prescient Systems, Escalate, Demand Management, Descartes, Arzoon, RiverOne, WorldChain, SoftChain, Logility, Demantra, John Galt, PipeChain, VCommerce, Ortems, SeeCommerce, and Tradec (now part of Agile Software) in SCEM, visibility, and operational performance monitoring are able to connect disparate systems to provide all the parties with near-real-time information on current movements and trends.

Adexa too has been posting growth and good financial results lately, economic hardships notwithstanding, by providing a homegrown-like solution that is flexible and integrated on a single data model and can be incrementally implemented as required. One should also note the capabilities of Viewlocity, which came as a result of a three-way merger and that combines former SynQuest's "quick-solving" planning and execution capabilities that have been linked with former Viewlocity and Tilion's event management capabilities (see Merger Mania at its Extremes). The combined solution aims at supporting the real-time enterprise through adaptive supply chain capabilities, including explicitly planning to maximize supply chain profit in a dynamic environment, while meeting or exceeding customer expectations.

Furthermore, pure-SCM leaders, i2 and Manugistics, have made forays into marrying these components of SCM while leveraging the best practices across multiple industries. In a mass manufacturing environment these systems may still provide deeper granularity for operational planning, while Webplan is more oriented to problem identification and resolution, even if it is not closely tied to the SCE layer yet. Not to mention the former core-ERP gorillas' (i.e., SAP, Oracle, PeopleSoft, SSA Global, QAD, etc.) ever increasing footprint in the SCM market, and their displeasure with the likes of Webplan "piggybacking" on their install bases.

At the same time, other best-of-breed SCP vendors have been stripping down and streamlining products sets, given their cumbersome and confusing offering in the past, which have consequently resulted with a bad image in many cases. An emerging SCP approach espoused by the likes of Optiant, SmartOps, and LogicTools has been to use inventory optimization techniques that create plans to minimize inventories across the network, but with achieving desired customer service targets.


Even without the extraneous market drivers, like recession, the future of a plethora of SCEM vendors remaining as viable stand-alone entities would be highly dubious. Like the former material requirements planning (MRP) providers, which are nowadays all but extinct due to either their evolution or merger into materials resource planning (MRPII) first, and eventually in manufacturing-oriented ERP providers (or the demise of the less decisive others), today's SCEM vendors simply cannot remain as viable stand-alone component providers. Consequently, as MRP today is part and parcel of ERP, SCEM, being rather an enabling technology in a supply chain network than an application on its own, will gradually be blended into the SCM family of matter-of-course' modules, losing thereby its functional sovereignty.

The challenging irony is that not many companies are yet able to connect SCM operational metrics to the strategic corporate metrics. Given human nature, any vendor, by bringing greater transparency to supply chain operations, may even be stonewalled by some managers because they perceive that it will prevent covering up unattractive figures and pushing the blame to someone else (see Why Most Balanced Scorecards are Subverted).

User Recommendations

In order to optimize their extended supply chains, manufacturers need to address the planning, execution, control, and analyzing aspects of SCM. Neither planning nor execution tools can be a panacea for a dysfunctional supply chain, but both areas have to be concurrently analyzed and optimized. Successful enterprises with appropriate performance management metrics in place are characterized by their ability to articulate strategies into easily understood action plans. They also emphasize data gathering from monitoring the performance of the action plan and fine-tuning it. Hence, in any case, the prospective customers have to be aware of how actionable any solution is. Also, standardized and centralized data, tighter collaboration and trust among trading partners, and tighter alignment between corporate strategic goals and SCM operational metrics are key success factors (KSFs) of contemporary supply chain management.

comments powered by Disqus