The food and beverage (F&B) and fast moving consumer goods (FMCG) manufacturers that supply the major supermarket retailers share many common business challenges, along with a tough competitive environment. This holds true whether the product is food, drink, personal care, cleaning products, or any other product stocked and sold by supermarkets. The customers—huge, powerful, and demanding supermarkets and retail chains—want products manufactured "to-order," with lead times often measured in hours rather than days or weeks. In fact, in the sector it is routine to deal with delivery lead times shorter than the time actually needed to make the product. To top it all off, this circumstance is often bundled with highly variable forecasts, which shatters any vestige of predictability.
Most consumer packaged goods (CPG) manufacturers have a few very important customers that account for much of their output, and these customers will usually provide some form of demand projection or forecast. Such customers are large retailers which control a significant portion of demand. Examples include Wal-Mart, Kmart, Tesco, Kroger, and so on (see Challenging the Competition: Mega-mergers and Supply Chain Technology). Because of the tight timescales involved, manufacturers have to interpret forecasts astutely in order to set the production processes in motion, and order entry has to be very closely coupled with forecasting, demand management, and manufacturing planning. In fact, since every major customer may have its own way of managing sales order entry, the supplier's enterprise resource planning (ERP) system must be sufficiently flexible and workflow-enabled to support the user's particular method of managing customer orders. For more information, see Process Manufacturing: Industry-Specific Requirements. And for more information on typical process manufacturing show-stopping ("fatal flaw") concepts, see Process Manufacturing Software: A Primer, What Makes Process Process?, and The Fatal Flaws for Process Manufacturers.
Furthermore, these arm-twisting customers often require their suppliers to come up with new product ideas and to run test production, without any guarantee that the new line will be approved. As if that were not enough, the costs of the finished goods are constantly being driven downwards—but consistent high quality is mandatory. On top of that, increasing regulatory requirements mean a burden of increased levels of control and reporting: a failure to meet a single one of the retailer's stringent demands means that the business will inevitably go to a competitor. Indeed, of late, the stricter new European Union (EU) regulations make food processors legally bound to have traceability systems, even if their customers do not necessarily require them. This is applicable to the entire supply chain (production, storage, purchasing, quality control, and so forth), and to everything that contributes to food safety (including packaging, closures, seals, bottles, jars, and the like). This is in contrast to the former requirement to identify only the source of a ingredient (see Is Intentia Truly Industry's First In Food Traceability?). There is also the need for backward traceability for multiple ingredients, as well as forward traceability, for recall purposes.
CPG manufacturers have to measure the most important metric of all: orders delivered on time, in full (see The Perfect Order—Inside-out or Outside-in?). Because supermarkets typically have no warehouses, all finished goods (whether perishable or not) have to be collected from the manufacturer or the supermarket's distribution center, and routed to the supermarket shelves within hours. If a vehicle is going to collect a certain number of roasted chickens or chilled apple pies, at precisely 2 p.m., say, in order to meet a delivery slot at the distribution centre, then everything in the manufacturing process must be geared to that deadline. And because there is no "finished goods" stock, the end of the production process often becomes the loading bay. Rather than conventionally handling products that are made in batches, for this market ERP systems should be able to deal capably with a combination of continuous processes and high volume, rate-based manufacturing (where one might manufacture to forecast, but package to order, in a mass customization manner). Indeed, an ERP system designed for CPG manufacture has to support an order-less, rate-based manufacturing process, and still retain full traceability.
CPG manufacturing is typically fast moving, high volume, and fairly simple, with relatively low-cost ingredients. Nevertheless, full quality control and lot traceability is essential, which means that it must be possible to use ingredients without the overhead of constantly having to record stock issues. Many ingredients will be fed from bulk tanks or silos, and the ERP system must therefore support vendor-managed inventory (VMI), shelf life management, inventory locations and zones (for example, to segregate organic materials from chemicals, or to accommodate differing ambiance temperature requirements), bulk stock handling, line side stock locations, and backflushing. Throughout the entire process, the system must maintain full product traceability, with minimum clerical intervention. It is unnecessary to stress that these manufacturers need to have planning and execution systems that can adapt rapidly to changing circumstances—in minutes, if necessary. And they also need data logging capabilities, in order to provide real-time quality functionality that can track materials right through the production process and into finished goods.
Forecasting and Rapid Response
CPG manufacturers are looking for astute systems that drive forward planning and deal with order expectations. These systems should be able to manage forecasts from a number of sources (which might be customer- or product-specific, or for pre-production runs, or for interplant or company demand). Other complicating factors in CPG forecasting include promotions. Occasionally a product may be promoted that requires special packaging, for example—as in the case where an extra 30 percent in volume is added to a product (say, mouthwash or toothpaste) in a non-standard package, but at a standard price. The forecast may expect 10 million units of this product to be sold over 3 weeks. But the true demand is actually determined by the success of the promotion. Perhaps the true demand is not 10 million units, but 3 million (in the case of poor promotion), or 14 million (for a good promotion). The manufacturer needs to be able to quickly detect the success of the promotion, alter the forecast, and adjust production to eliminate over- or under-production.
These systems should also allow the creation of a reliable forecast at a lower level in the product structure than the top-level (finished goods) item. This lower-level forecast can be consumed by production (as with all forecasts), and its purpose is to support the volume manufacture of products where a wide variety of finished options is available, based on a standard base product. In theory, the end product is made specifically to customer order—and the possible variations at the top-level saleable item are so great that forecasting at this level will always be very inaccurate. By forecasting at a lower level that is probably common across a variety of products, a more predictable forecast can be maintained, and manufacturing can go according to customer specific call-offs. When this is integrated into the product design and manufacturing planning, it is called "decoupled manufacturing" or "mass customization." For example, a manufacturer of cosmetic face creams may forecast total demand for a face cream regardless of how it will be packaged; the face cream will thus be made in bulk, and its packaging will be driven by specific customer orders.
All this is to say that across the manufacturing sector in general, companies are realizing that the demands of customers for rapid delivery responses, combined with corresponding requirements for greater variation of product (whether in terms of color, pack size, or specification) can only be met by manufacturing in separate stages. While this can be done easily enough (if one is prepared to stock vast quantities of the partially finished product and simply configure, finish, pack, and so on, when the customer places his order), this is hardly a practical option, not least because of inventory handling and cost issues. Supporting decoupled manufacturing is rather about forecasting more accurately at a common intermediate or part-finished level (as with face cream), and being able to finish—and therefore customize—to a particular order.
There is also a need for control of output material, which means that complicated material process flows can be defined, including by-products, co-products, waste, scrap, yield, work-off, and feed-back, all of which can then be the inputs into inventory, or into other production stages for other processes. These inputs and outputs should be included in the costing process as positive or negative contributions, which would allow the definition of the production process to be a close description of the actual way products are manufactured.
This approach includes delivering material supplies to the production process in line with manufacturing requirements according to time and location. This emphasizes the importance of better timing within a production operation, since few successful manufacturing companies in virtually any sector can simply dump a pile of raw materials into a machine and get their products out exactly when the customer wants them. A major capability would be the ability to spread demand across multiple production sites so as to increase resource usage, and minimize production and transportation costs. For companies that stock goods as intermediates (partially finished), or bulk goods (such as whiskey distillers or chemical producers), both finished and partially finished goods can be accounted for. All this is in contrast to the concept used by customary material requirements planning (MRP)-based systems, whereby material is kitted to a work order according to a list of discrete components or ingredients, called the bill of materials (BOM).
Holistic Inventory Optimization
The aim of specialized inventory optimization products is to enable clients to reduce investment in stock, while at the same time maintaining or improving customer service levels. Managing risk is about managing the cost of maintaining unnecessarily high inventory levels versus the risk of running out of stock at the crucial point when a customer actually wants something (the "moment of truth"). Most traditional ERP systems record transactions, run programs to suggest re-ordering (based on fairly crude algorithms or even more dubious forecasts), and occasionally come with some form of management reporting. Few, if any, include sophisticated inventory management tools of the level required for the consumer goods supply chain. For other research notes on this subject, see Inventory Planning and Optimization: Extending Your ERP System, and Lucrative but 'Risky' Aftermarket Business—Service and Replacement Parts SCM).
In an ideal world, one would of course hold exactly the amount of inventory required to never be out of stock, for anything. However, real life is obviously a far cry far from this, as there are both financial and physical constraints. Some products have a shelf life, or new products might flood into the market. And there are a host of other reasons for having to strike a balance between anticipated demand and expected supply. Supply chains are indisputably complicated, and one can never know all. The only certainty is the fact that supply chain efficiency depends on minimum buffer stocks, integrated processes, efficient use of resources, and shorter lead times. Nonetheless, experience shows that companies either struggle with a surfeit of reliance on local knowledge, chance-taking, and effective fire-fighting, or else spend time and effort developing their own supporting systems (usually glorified spreadsheets) to help to better manage inventory levels.
In other words, because demand and supply are inherently unpredictable, most companies employ some method of forecasting, even if it only entails using intuition or a spreadsheet. However, almost universally, they encounter one or more difficulties: they scribble their forecasts on scraps of paper and promptly lose them; or different people maintain unhelpfully individual forecasts. Or they may spend valuable time on extracting suitable data on a planned and regular basis, only to have to struggle to rehash and format the data. None of these forecasts can easily be integrated into planning systems to drive the replenishment process.
In any case, management needs an effective mechanism, or an understanding of the methodology, in order to establish macro inventory targets (expressed in days of coverage and availability) in accordance with corporate strategy and then to link these targets to line item replenishment. Too often, little effort is directed at scientifically establishing the macro level of inventory targets, let alone the line item level.
Keeping Planners, Buyers, Executives, and Suppliers on the Same Page
The other side of inventory is the supply side, and the unpredictability of supply is often as large a contributor to excess inventory as difficulties in demand forecasting. Any attempt to reduce inventory must also focus on managing the supply chain, and for this the ability to manage suppliers and their lead times is the key. Most good ERP packages record supplier performance data, but few include pre-built supplier performance analysis. Often this is limited to a few reports: in order to make the most effective use of the data, expensive custom reporting suites have to be built around business intelligence (BI) reporting tools.
The common theme running through the challenge of optimizing inventory levels is the need to identify the real causes of inventory. Simply treating the symptoms can often hide the root cause of shortage or excess stock. Organizations that hold stock at multiple locations have the additional challenge of not only determining optimum stockholding levels, but also of deciding where to hold the inventory within the nodes of the supply chain network. Again, transaction histories can provide misleading information, since stock movement between locations may indicate artificial demand and supply, whereas real demand may exist at only one location.
But even the best inventory optimization methodology may be worthless unless it is applied constantly. A once-off exercise might make a difference, but without constant adjustments, fine-tuning, or warning of impending shortages and inventory build-up, the same problems typically recur. The key to effective ongoing inventory management is an applied methodology, tied to visibility of its effect. This needs to be set up in a hierarchical manner, whereby key performance indicators (KPIs) are defined overall, and strategic product grouping is performed. Management reports (preferably online and near-real-time) should show the total inventory picture against these KPIs, and reports should then be requested to show the detail of any discrepancies. These reports, requested online directly from the summary KPIs, are known as drill-downs—a quick and effective way of identifying potential problems before they even occur.
About the Authors
Predrag Jakovljevic is a principal analyst with Technology Evaluation Centers (TEC), with a focus on the enterprise applications market. He has nearly twenty years of manufacturing industry experience, including several years as a power user of IT/ERP, as well as being a consultant/implementer and market analyst. He holds a bachelor's degree in mechanical engineering from the University of Belgrade (Yugoslavia), and he has also been certified in production and inventory management (CPIM) and in integrated resources management (CIRM) by APICS.
Olin Thompson is a principal of Process ERP Partners. He has over twenty-five years of experience as an executive in the software industry, and has been called the "father of process ERP." He is a frequent author and award-winning speaker on topics such as gaining value from ERP, SCP, e-commerce, and the impact of technology on industry.
He can be reached at Olin@ProcessERP.com.