Connecting the consumer to the factory.
What's so hard about connecting consumer demand to factory production? How is it that after decades of successful cases demonstrating the value of demand-driven supply chains in business-to-business (B2B) settings, most retailers still lack technologies that can maximize sales at the shelf or Web portal?
There is no single answer to these basic questions. Collaborative planning, forecasting and replenishment (CPFR) techniques have provided clear process roadmaps. Until recently, however, enterprise systems that could support large-scale alignment on a consumer sales-driven demand plan fell far short of the vision. Manufacturers have understood it for decades, but bricks-and-clicks retailers are only now beginning to realize that demand-driven systems and processes are key to defending and growing their share of market--not just against other bricks-and-clicks competitors, but against an explosion of Internet retailers and business-to-customer (B2C) dealers using exchanges such as Ebay and Amazon.
Two factors are galvanizing change. To begin with, customers are raising the bar. Demand-driven systems and processes will help retailers to offer the "omni-channel" service experience that customers are learning to expect from their online and mobile shopping activities. Secondly, recent technical breakthroughs have made possible the deployment of reliable and economically scalable store-level distribution resource planning (DRP) capabilities for retailers. The newest retail DRP systems support detail planning over an extended horizon--a big leap forward in capability over earlier systems, which were purely executional.
With these systems, retailers can calculate demand three levels into the supply chain, driving factory output from what consumers are actually buying at the shelf or Web portal. The benefits flow outward: retailers' trading partners can synchronize supply and demand and incorporate all of the logistics constraints or product flow from the factory floor to the store shelf with unparalleled accuracy
In effect, retailers and their trading partners can now coordinate their supply chain as if only one company were managing it, enabling them to jointly reduce inventories and cut the costs of manufacturing, transportation, distribution centers, procurement, and selling. The gains can be substantial: for instance, Sony (Canada) reduced its supply chain inventory by 20 percent and saw its in-stock levels improve from 87 percent to at least 95 percent. Lowe's reported a significant reduction in excess inventory, and its in-stock stepped up from 92 percent to 98 percent.
My company, West Marine, Inc., was the first retailer to use an enterprise solution for driving supply chain demand from consumer sales. In 2001, we set up a store-level DRP system supporting a 52-week forward planning horizon. More recently, IKEA, Meijer, OfficeMax, Lowe's, Mark's (Canada) and Sony (Canada) have implemented store-level DRP solutions and reported improvements in service levels, cuts in supply chain inventory, and better coordination with suppliers on new products and marketing.
For instance, Mark's reported an 8 percent reduction in inventory and record sales growth. IKEA said it achieved cost savings from its longer-term commitments with suppliers. And in 2009, Walmart told its suppliers that it would implement a store-level DRP system with the intention of modeling the supply chain with its trading partners and providing manufacturers with schedules of its planned orders. For their part, suppliers including Scotts, Vtech, Kraft, Sony (US), and LR Services (Europe) have also implemented store-level DRP capabilities to improve their planning and execution with key retail partners by building out demand and supply plans based upon sales by location.
Store-Level DRP Still Not Widespread
Even though store-level DRP implementations continue to spread, the fact remains that store-level DRP today still accounts for only a small part of the total volume of retail supply chains. Furthermore, there is still a significant gap between the demand planning and execution of B2C compared to B2B supply chains. B2B supply chains have been forecasting sales and orders starting at the end customer for many years. Computerized materials requirements planning (MRP) systems started in manufacturing organizations in the 1960s and led to the development of DRP systems. Over the past 40 years, countless manufacturers have deployed DRP solutions to drive the flow of products across their distribution networks and to align their supply chains around a single forecast. DRP systems allowed manufacturing supply chains to maintain an accurate and forward-looking bottoms-up computerized model of their extended business that accurately predicted sales, purchases, receipts, manufacturing, and distribution from source to customer.
Also, manufacturing supply chain participants found that they could rigorously compare all of their bottom-up predictions to their aggregate targets; they could manage exceptions and make adjustments to deliver planned results with the least disruption and at the lowest cost. These DRP capabilities enabled the development of the extended forward planning process that we know today as sales and operations planning (S&OP). Without similar bottoms-up forward planning capabilities, retailers have been unable to plan their business in detail over an extended horizon and to assess and optimize these plans in advance. Store-level DRP solutions will close this gap and provide the basis for a similar leap forward in planning and execution capabilities for retail supply chains.
So why has store-level DRP worked for B2B but not for B2C until recently? My vantage point as a retailer and leader of both the former Voluntary Inter-industry Commerce Solutions (VICS) CPFR Committee and now the GS1 US CPFR Workgroup gives me some perspective on what has held back most retail supply chains thus far. There are four key hurdles:
* The computational challenges of a large-scale store-level DRP system were difficult for solutions providers to overcome.
* Retailers had complex legacy replenishment systems that were difficult to abandon.
* Retailers lacked the practical experience and know-how required for developing organizational routines for coordination and forward planning such as CPFR and S&OP.
* Functional and organizational boundaries inhibited the development of robust relationships and trust among supply chain partners.
Benefits of Store-Level DRP Systems
As a retail practitioner, I can appreciate the process improvements that retailers and their manufacturing partners realize by implementing store-level DRP systems. Retailers benefit because they have to create only one robust forecast: the forecast for consumer sales. They are relieved of the task of creating robust distribution center (DC) shipment forecasts, because the DRP system automatically calculates the DC shipments through its time-phasing calculations. Automation of the DC shipment forecast essentially eliminates the function of DC forecaster. This saving either reduces operating costs, or is invested in hiring merchandise planners who are accountable for assortment planning and inventory optimization. The most important outputs of a DRP system are the forecasts of orders to suppliers. Retailers commonly refer to these as order forecasts; manufacturers call these time-phased forecasts "supplier schedules"-and manufacturers have been providing them to their upstream trading partners for decades.
In providing a long-term order forecast to manufacturers, a retailer communicates all the details of its demand plan. Usually this communication is weekly, but it can be daily or even real time. When retailers are as transparent as this with their supply chain partners, things happen very quickly, and with immediate consequences. (My analogy is that when we hiccup, it's on Facebook.) The information provided in a retailer's order forecast or supplier schedule makes everything about the retailer's demand strategies completely apparent.
For retailers, this level of transparency is challenging, but it facilitates commitment to reliable planning. That is good; DRP processes extend the trading partners' commitment horizons. However, it will not automatically increase trust in the relationships. As a senior planning associate at a large retailer once said, "Why would I provide an order forecast to suppliers'? That would just give them a stick to beat me with." For West Marine, it's clear that having a store-level DRP system has enabled and encouraged us to plan ahead and explore sales and marketing potential beyond anything we could have imagined without a bottoms-up planning engine. Using our forecasts, our suppliers improved on-time shipments from a dismal 30 percent to over 90 percent, and we maintained peak season in-stock at 96 percent overall and 98 percent for key items. More importantly, our suppliers came to recognize the value of coordinating their efforts with ours to drive consumer sales, customer satisfaction, and market share.
That said, the task of creating reliable sales forecasts is far from easy. The ideal situation for a manufacturer is that all of its retail customers provide order forecasts or supplier schedules. Today, manufacturers may be receiving store-level DRP supplier schedules from some retailers, which should be the most reliable projections they have. If these retailers are large customers, the forecasts are immediately beneficial for production planning and they provide the basis for additional CPFR activities between the partners. Other retailers may provide automatically calculated order forecasts based upon DC activities, but these forecasts rely upon historical DC shipments to stores rather than a multi-echelon store-level forecast methodology. A third group of retailers may not provide any insight to manufacturers on future orders. For this group, the manufacturer has no information, except its historical shipment data, on which to base a demand forecast. In this case, the manufacturer must independently forecast the demand plan for these retailers.
Until recently, the planning capabilities of retail supply chains were very different from those of most of their manufacturing counterparts. Most retailers' store-level replenishment systems can robustly plan consumer sales, but these systems generally calculate only today's orders. They have no capability to generate store ordering needs beyond the current order, much less orders next week, next month or next season. Retailers generally employ DCs to optimize freight handling and allocation of products to stores. Most retail DC replenishment systems forecast shipments to stores based upon DC historical shipments, without reference to a store-level order forecast and without accounting for current store selling or current store overstocks or under-stocks.
Without systems that support a collaborative demand plan established through a bottoms-up store-level DRP process, business execution visibility is short-term and orders are often a surprise (as if they were tossed over a wall). Each trading partner forecasts its d[e.sub.mand] independently, limiting category or market insights that could be shared between key suppliers and retailers. A history of past supply chain outages may drive each trading partner to build buffer stocks to avoid the risk of out-of-stocks. The inevitable supply chain overages and outages regularly result in adversarial buyer-seller relationships.
A store-level DRP forecast immediately accounts for any changes in end customer sales, with the result that increases in sales automatically cause calculated purchases to move closer to today, and decreases in sales automatically cause calculated purchases to move further out into the future. Updates to the network forecast are rapid, with weekly, daily or real-time updates as achievable norms. Such forecasts virtually eliminate bull-whip effects, enable participants to optimize capacity and throughput planning, eliminate waste, and improve value-added processes.
Store-level DRP systems are available to support forecasts for an extended horizon--over a 12-month or longer rolling horizon--which gives trading partners the data they need to coordinate more strategic and longer lead time tasks. Linking all activities to a DRP forecast creates a complete and dynamic calculated model of the extended enterprise. Since all future supply chain activities can be captured in this computerized model of the business, trading partners can share a single plan and work together to analyze and optimize their linked business activities.
Retailers that implement store-level DRP will find that their new demand coordination capabilities help each trading partner to work as part of a coordinated value chain. This will help trading partners to develop relationships of trust. Retailers and manufacturers build trust when retailers honor the forecasts that drove production. Manufacturers will build trust when they use available-to-promise (ATP) methodologies to protect product supply for retailers as if DRP forecasts were actual orders.
One leading SCM thinker put it this way: "Knowledge-sharing activities result in a production network that learns faster than other production networks about the best practices in production, quality, and management. High trust within the extended enterprise results in a production network with the lowest transaction costs. Less time is spent bargaining and haggling over the pie and more time and resources are spent increasing the size of the pie." (1)
How Does Store-level DRP Work?
Supplier schedules created by store-level DRP systems are the connecting bridge for the flow of information from buyers to sellers. (2) Supplier schedules are the information flow about future ordering from stores to retail DCs, from retail DCs to manufacturer DCs, and from manufacturer DCs to the factory. (See Exhibit 1.)
The form of the schedule is item numbers, dates, and quantities representing the expected orders from each node in the supply chain to the next, including from the retailer firm to its manufacturer. Building the supplier schedule begins with only one forecast: the independent forecast for sales to consumers at each selling location. All demands for nodes further up the supply chain are calculated and thus are dependent demands, based upon the consumer sales forecast.
Let's take the example of the retail chain for sales of 40-inch HDTV sets. The building blocks for the supplier schedule are the demand plan (sales forecast) and the dependent replenishment forecast (planned shipments) for the 40-inch TV unit in the Los Angeles store. (See Exhibit 2.) The planned shipments are calculated accounting for the planned safety stock, the presentation minimums, delivery schedules, and order minimums and multiples. The store-level DRP system calculates similar planned shipments for all stores serviced by the DC. The timing and quantities of the sum of the planned shipments to all stores becomes the dependent demand forecast for this DC for the HDTV.
The next step in the DRP process is to calculate the DC's total purchase needs in a time-phased manner that discriminates quantities and both shipment and receipts timing. An important data point is the calculated total receipts schedule for all four sizes of HDTVs carried in stores. (See Exhibit 3.) This data set also shows how the visibility of store-level DRP supports a comparison of the calculated receipt plan for all HDTVs with the approved receipt from the monthly sales and operations plan.
The ability to compare the S&OP plan approved by the executive team with the sum of the supplier schedules coming from the day-to-day operating system (store-level DRP) provides an extremely powerful capability for the retail management team to improve control. Let's say the management team sets a tolerance for purchase variance for the HDTVs of plus or minus 2 percent. The variance between the automated order forecast for the current month is 30 TVs or 1.6 percent, which is within the tolerance. But the variance projected for next month is 450 TVs or 21.4 percent, a difference that requires resolution. In this instance, the buyer must change his planned purchases for the second month or seek approval for the additional purchase before releasing orders.
By starting at the lowest level (item store) and leveraging the continuous calculations in the DRP solution, retailers can drive the supply chain from the consumer demand forecast. Tolerances can also be set and efficiently managed to close the loop between S&OP and the day-to-day operating plans generated by the DRP system.
Another key data set shows how a retailer's DRP-generated supplier schedule links the retailer and manufacturer and also aligns the two partners' S&OP processes. (See Exhibit 4.) Remember that without the retailer's supplier schedule, the manufacturer's demand plans and S&OP are based upon historical shipments. Using the retailer's supplier schedule, the manufacturer replaces its own shipment-based demand plan for this retailer with the retailer's demand plan calculated from consumer sales.
Notice that the two-way red arrow in Exhibit 4 represents the flow of actionable information between the retailer's supplier schedules and the manufacturer's demand management. Demand information goes in both directions because it flows from the retailer to the manufacturer, while information about the manufacturer's supply issues flows back to the retailer and is reflected in its plans.
This quick tour of DRP functionality makes the process seem very simple. In reality, the challenge of accomplishing calculations for millions of store-item combinations across distribution networks and extended time horizons has, until recently, prevented the development of effective enterprise store-level DRP solutions.
A Roadmap for Getting Started--Quickly
The advent of effective store-level DRP systems creates a competitive advantage. Companies in retail supply chains that embrace disciplined internal organizational routines (S&OP) and external routines with their trading partners (CPFR) will multiply the benefits of robust and dynamic automatic demand planning. The GS1 US CPFR Work Group recommended in its most recent guideline that retailers undertake a streamlined investigation to assess the benefits of these new solutions for their businesses. (3)
Retailers can launch a well-organized assessment project in approximately one to two months. We suggest the following six-step launch program:
1. Audit and Assessment (one week). An executive team with appropriate operating managers and outside experts is formed to analyze the systems and processes used in current demand planning and replenishment in retail stores and DCs and in coordinating with manufacturers. The assessment will cover what's not working well and what needs to change to make the organization more competitive. Current process design, technology, and planning competency will be addressed in this step.
2. Establish the Vision (a half-day workshop with the same leadership team). The objectives of this session are to achieve clarity about what will be accomplished in the store-level DRP project, particularly to instill forward planning practices throughout the organization and with the supply chain partners. The vision will be documented by the leadership team and communicated to the organization.
3. Establish Performance Goals (a one-day workshop with the same leadership team). This session will establish the performance categories that need to improve, how they will be measured, and to what levels performance will be expected to rise. The plan will set goals and measures for continuous performance improvements.
4. Building the Plan for Costs and Benefits (a one-day workshop with the same leadership team). This session will review all the anticipated costs and benefits related to adopting new processes and the store-level DRP solution.
5. Project Authorization (a half-day executive meeting). The authorizing decision should be taken at the conclusion of this meeting or as soon thereafter as possible.
6. Project Organization (one week). This step involves forming the project organization, allocating appropriate resources, and delegating required decision-making authority. It effectively launches the project and outlines the project's organization.
This succinct roadmap will launch the project and provide the structure for project design and governance. The executive or the team in charge of the project will establish appropriate timelines and milestones to implement and roll out the new technical solution. The implementation will include a rollout of new processes, roles, and responsibilities appropriate for the organization and for coordinating with its manufacturers.
Embrace the Change
Retail supply chains will become more competitive if retailers and manufacturers work together to reduce supply chain costs over time and share gains from eliminating waste or from using more effective value-added processes. In particular, they need to abandon win-lose approaches, which imply that a slight cost advantage by another supplier will cause an order to be lost or a contract not to be renewed.
So is your organization ready to adopt store-level DRP systems? Some companies resist the idea of forecasting, let alone planning. They believe it is impossible to predict the future accurately, so why bother? This attitude not only risks making a mockery of planning but it ignores the current realities of customers' expectations and technology's advances. Management guru Peter Drucker gave a crisp view about predicting the future: "To try to make the future happen is risky; but it is a rational activity. And it is less risky than coasting along on the comfortable assumption that nothing is going to change."
Customers are forcing things to change. New tools are enabling things to change. If your company doesn't acknowledge those truths soon, there's a very good chance that your competitors will.
(1) Jeffrey H. Dyer, Collaborative Advantage: Winning Through Extended Supplier Networks, Oxford University Press, 2000.
(2) Adapted from the GS1 US CPFR guideline "The Ultimate Retail Supply Chain Machine: Connecting the Consumer to the Factory."
(3) Adapted from the GS1 US CPFR guideline "The Ultimate Retail Supply Chain Machine: Connecting the Consumer to the Factory."
Larry Smith is the Senior Vice President of Planning and Replenishment at West Marine, Inc. He can be reached at LarrySm@westmarine.com
EXHIBIT 2 Forecast of TV Sales from Los Angeles Store Past Due 5/1/xx 5/8/xx 5/15/xx Forecast    In-Transit Projected On-Hand     Planned  Shipments-Receipts Planned  Shipments-Ship 5/22/xx 5/29/xx 6/5/xx Forecast    In-Transit Projected On-Hand    Planned  Shipments-Receipts Planned Shipments-Ship Source: DRP: Distribution Resource Planning EXHIBIT 3 Summary Schedule of Receipts vs. the Monthly S&OP Past Due 5/1/xx 5/8/xx 5/15/xx 40" HD TV   46" HD TV 52" HD TV     56" HD TV  Monthly Totals TV Category Monthly S&OP Plan 5/22/xx Monthly 5/29/xx 6/5/xx 40" HD TV  [1,500]  46" HD TV  [1,400] 52" HD TV   56" HD TV  Monthly Totals [1,930] TV Category Monthly [1,900] S&OP Plan 6/12/xx 6/19/xx Monthly 500 40" HD TV  46" HD TV  [1,040] 52" HD TV   56" HD TV  Monthly Totals [2,550] TV Category Monthly [2,100] S&OP Plan
|Printer friendly Cite/link Email Feedback|
|Publication:||Supply Chain Management Review|
|Date:||May 1, 2013|
|Previous Article:||To succeed at sustainability, first learn to communicate.|
|Next Article:||The new basics of supply chain management.|