Читать книгу The Demand Driven Adaptive Enterprise - Carol Ptak - Страница 11

CHAPTER 2

The Prerequisites for Relevant Information

The search for and use of relevant information to control variability, promote flow, and ultimately drive ROI begins with an understanding of four basic prerequisites. The absence of one or more of these prerequisites will dramatically compromise the integrity of the relevance of the information.

Prerequisite #1: Understanding Relevant Ranges

The concept of relevant range comes from economics and management accounting. Relevant range refers to the time period in which assumptions are valid. Trying to force fit assumptions (and metrics derived from those assumptions) into an inappropriate time range directly results in distortions to relevant information, creates variability, and thus causes a breakdown in the flow of relevant materials and services.

The assumptions and information that are valid and relevant within different time ranges will differ dramatically and these differing ranges are utilized by different personnel. For example:

Forecasts are highly relevant in thinking about the future, they are essentially useless with regard to what needs to be done today.

Fixed expenses are variable in the longer range, not in the close-in range (that is why they are called “fixed”).

Work order delays are relevant for the current schedule, not for an executive trying to determine the company’s strategy for the next year.

A machine breakdown is relevant for a maintenance crew’s activity, not what the company thinks it will sell in Q3.

There are at least three distinct relevant ranges that we must deal with: operational, tactical, and strategic. The specific time value of each relevant range will differ between companies depending on the circumstances involved in that company. Chapter 3 will describe a way to discretely determine these ranges for every organization.

Prerequisite #2: Tactical Reconciliation Between Relevant Ranges

While the assumptions and information that are relevant for decision making differ between ranges, there is still an absolute need to reconcile these differences on an ongoing and iterative basis; the signals at all levels must align and reconcile. Strategy must be influenced by defined operational capability and performance as well, considering how the operating model might perform under additional predicted scenarios. Operational capability must be influenced by predicted scenarios and/ or strategic expectations in future time periods to ensure that the operational capability supports the strategy. But how does this reconciliation actually occur in a true bidirectional fashion?

Between the strategic and operational ranges is the tactical range. As we will see in Chapters 3 and 5, this tactical range is unique, and for it to perform the necessary reconciliation it must become a bidirectional reconciliation hub for the key characteristics of a CAS.

Prerequisite #3: A Flow-Based Operating Model

A flow-based operating model is an operating model specifically designed on the fundamental principle of flow. Chapter 1 firmly established two things:

Promoting and protecting flow is the key to increased return on investment. This is the very essence of Plossl’s Law.

Flow can be a unifying factor, not just in operations but also between functions, to advance their individual primary objectives.

A flow-based operating model makes the following critical assumptions:

Assets must be as closely synchronized to actual demand as possible. The cost of being wrong has grown dramatically with the rise of complexity and volatility. Synchronizing assets as close as possible to actual demand minimizes that risk.

Variability cannot be eliminated but its impact on system flow can be effectively controlled and mitigated with the right operating model design. Every process, even when under statistical control, still experiences variability within its control limits; that is a given. The accumulation of this variability between processes is really the only relevant thing when it comes to managing system performance.

Focusing on precisely synchronized planning and scheduling across all materials and resources is largely wasted effort. Most MRP and shop-floor schedules are unrealistic as soon as they are released. Constantly rescheduling to attempt to account for variation actually introduces additional variability and subsequent confusion.

Some level of inventory is required at some stage in the supply chain because the customer tolerance time is insufficient to procure and make everything to order. Inventory cannot be entirely eliminated. At the same time, inventory should not be everywhere. The choice of where to place inventory is highly strategic since it determines the customer lead time and the level of working capital.

Batching decisions should be based on flow considerations instead of cost considerations. Batches are an actual bona fide limitation for most companies. How those batch sizes are determined, however, is a choice. Unfortunately, the vast majority of batching decisions are driven by cost performance metrics instead of flow-based determinants.

Some level of protective capacity must exist in any environment with multiple interacting resources. Resources have disparate levels of capacity relative to their respective required tasks and volume. The perfectly balanced line is still a mythical creature with an enormous price tag. This disparity means that precaution must be taken against misusing or wasting points of additional capacity.

Inventory is an asset and should be treated as any other asset on the balance sheet. Some lean enthusiasts like to promote that inventory is a liability; however, whether inventory is an asset or a liability depends on the position and quantity of that inventory. Production assets such as machines are added only when there is a positive return on investment. Inventory targets must be the result of a strategic choice for return on investment as well.

Capability must be established based on the expected future for the company. Since it is not possible to predict the future precisely, the capability range is established by the combination of inventory placement and protective capacity. The size of this range is dependent on the uncertainty of the future. The establishment of this range is dependent on a comprehensive risk assessment of the market and supply chain.

The objective of the operating model is to maximize margin by focusing on increasing service levels, enable premium pricing by providing a unique competitive market position, leverage constrained resources, and identify available capacity that can be used for incremental margin positive opportunities. This is in contrast to the more common expectation that the operating model should be developed to minimize unit cost and inventory while maximizing efficiency and utilization.

Thus, the question becomes: is there a well-defined and proven flow-based operational model that can scale to the multi-echelon enterprise level? Chapter 4 will focus on that model—the Demand Driven Operating Model.

Prerequisite #4: Flow-Based Metrics

If flow of relevant information, material, and services becomes the common focus for decision making in day to day operations and throughout the organization, then we need an appropriate suite of metrics for each of the relevant time ranges to support that focus. Appropriate metrics will allow us to maintain organizational coherence to that ROI objective now and in the future. Flow-based metrics should neither be a source of variability nor exacerbate variability.

Force fitting or failing to remove non-flow-based metrics in the deployed metrics suite will directly lead to conflicts and distortions throughout the organization—it will obscure what is relevant! Obscuring what is relevant directly leads to more variability, which in turn directly inhibits the flow of relevant information, materials, and services. Worse yet, when we do this, we are doing it to ourselves; it is self-imposed variation that directly hurts our ROI performance and causes great stress to the organization’s personnel.

Any suite of flow-based metrics must consider three additional prerequisites:

The metrics must fit the appropriate range.

The metrics must be reconcilable between ranges.

The metrics must fit the flow-based operating model.

These four prerequisites for relevant information define what it means to think, communicate, and behave systemically—the only way to protect and promote flow. If an organization and its personnel do not have this thoughtware installed, then the flow of relevant information and materials will always be impeded to varying degrees, leading directly to poorer ROI performance. Thus, before companies invest significant amounts of money, time, and energy into new hardware and software solutions, they must first consider investing in the proper thoughtware in order to gain visibility to what is truly relevant. Technology can only provide value if it addresses a limitation that the company is facing in terms of achieving its objective.

Convention’s Failure with the Four Prerequisites

Now that the four prerequisites for relevant information are established, let’s turn our attention to the failure of convention to establish and properly use these prerequisites. First, let’s describe what the typical conventional approach looks like.

The conventional approach to managing a company involves strategic, tactical, and operational perspectives. Strategy is deployed to the organization by a Sales and Operations Planning (S&OP) process. The S&OP process balances supply availability and demand requirements resulting in a Master Production Schedule (MPS). The MPS is essentially intended to be a tactical dampener to prevent the forecast variability from driving MRP directly due to the imbalance of load against capacity. In the process of preparing the MPS, S&OP integrated reconciliation performs a rough-cut capacity check. The only bidirectional interaction between the MPS and S&OP is an “aggregation—disaggregation” process in which disaggregated demand forecasts are rolled up from the item level to the product family level to create the new forward-looking production plan. Then the production plan is rolled down into item level requirements by date, and this is used as the gross requirements line for MRP calculations.

In summary, the MPS is a statement of what can and will be built recognizing available capacity. The MPS sends this forward-looking plan for the planning horizon to Material Requirements Planning (MRP). MRP calculates the necessary supply order generation dates and quantities necessary to synchronize to that plan through a level by level explosion. Orders are then released when required by a date that has been calculated precisely from this multi-level explosion process. Figure 2-1 illustrates the conventional approach.

FIGURE 2-1 The conventional approach

Now we will turn our attention to the failure of convention regarding the four prerequisites to relevant information.

Convention and Relevant Ranges

The conventional approach clearly supports the need for relevant ranges. Figure 2-1 clearly shows a top-down linear approach that recognizes strategic, tactical, and operational emphases. The problem with convention, however, is that it improperly manages these ranges. This will be demonstrated through two examples that are devastating to relevant information and flow.

Our first example is conventional planning’s reliance on forecasting item level demand for the planning horizon. Obviously, predicting market behavior and conditions is a necessary component for successfully managing any business. The better our forecasts, the better leadership can define and manage a company’s path to success—there should be no doubt about that. However, bringing those predictions into the immediate operational range by tightly synchronizing order generation directly with a demand signal containing known error creates an immense amount of distortion and waste. There are three rules about forecasts:

They start out wrong.

The more remote in time the forecast is extended, the more wrong the forecast will be.

The more detailed the forecast is, the more wrong the forecast will be.

Despite these well-known facts, convention continues to drive actual supply orders to forecast and then attempts to make corrections as better information is available. This means that capacity, capital, materials, and space are committed to signals that have significant rates of error associated with them. This is the very definition of irrelevant or at the very least distorted information and is one reason why forecasts are irrelevant in the short range. This is a clear mismanagement of the short term relevant range, demonstrating a lack of comprehension of this critical concept. Readers wishing to know more about this issue should consider reading the book Precisely Wrong—Why Conventional Planning Fails and How to Fix It (Ptak and Smith, Industrial Press, 2017).

Our second example of improperly managing relevant ranges is the use of fully absorbed unitized cost metrics for operational decisions. Fully absorbed unit cost means that all manufacturing costs, fixed and variable, are absorbed by the units produced. In other words, the cost of a finished unit in inventory will include direct materials, labor, and overhead costs.

Direct materials are variable costs. Variable cost is tied to unit volume, not resources. Variable costs rise and fall with unit volume but do not change on a per unit basis. Labor and overhead are fixed costs. Fixed costs are not affected by volume changes in activity level within the operational relevant range (a specific short-range period). Using fully absorbed unit cost related metrics directly creates the false impression that fixed costs can and will vary within the short range. They do not; that is why they are called fixed costs. The unitized cost equation obviously improperly mixes relevant ranges. This causes significant distortion in relevant information at the operational level and directly relates to disruptions in coherence and flow. Once again, this is a clear mismanagement of relevant range, demonstrating a lack of comprehension of this critical concept.

Convention and Tactical Reconciliation

In convention, tactical reconciliation is not bidirectional—it is essentially a one-way street in an assumed linear predictive world. This oneway direction limits the ability to drive any meaningful adaptation and additionally, any attempt at periodic reconciliation is incredibly painful. For example, every MRP run results in massive cascading schedule changes as date and quantity changes at higher levels affect all connected lower level components. This is an effect called nervousness and leaves most companies in a huge dilemma about how often they run MRP and how much they can actually trust its output, causing planners to export data to spreadsheets in an attempt to find the truth.

At a higher level, monthly S&OP updates create massive shifts at the beginning of every month that are then compounded by new MRP runs. Planning personnel struggle to stabilize the requirements picture and plans only to have the new version of the prescribed future reality dropped into their laps at the beginning of the next month. In convention, this makes tactical reconciliation more akin to a constant and repetitive cycle of tactical demolition and reconstruction; both are incredibly messy, wasteful, and cause great confusion.

Convention and Flow-Based Operating Models