It would be so easy—wouldn’t it—if the forecast was right on the money, and demand changed only slightly—or not at all. But as the old saying goes, there are only two sure things in life—and neither of them is a perfect forecast.
Supply chains can be disrupted by weather, geopolitical upheaval, labor stoppages, poor quality from a sub-tier supplier, consumer whim and more. In response to the added pressures, supply chains have increasingly shifted to a demand-driven supply chain.
According to the Boston Consulting Group, a demand-driven supply chain “offers real-time information on demand and inventory levels to all supply chain participants, so that they can react quickly and effectively when unexpected changes arise.”
Demand-driven supply chains have been gaining popularity over the past couple of decades, says Jeff Bodenstab, vice president of marketing at ToolsGroup, calling it an evolution—not a revolution—that started in small ways and picked up steam.
“If I had to pick one industry early on that was most aggressive, it would be consumer goods, like CPG and food and beverage,” he says. “They had the will and the way. They had a need because [they were] dealing with highly demand-driven supply chains and lots of different channels and sources of demand that tended to fluctuate quite a bit.”
Since then, he says, use has been broadening to retailers, wholesalers and industrial companies who find a need and are seeking ways to make it happen. “There was a lot of earlier work around making the transition from inside out to outside in—translating downstream demand. In its simplest form, most companies, 10 to 15 years ago, were looking at a very gross aggregate: How much product did we sell last month versus this month? When you deal with aggregate, there are a lot of variations that balance each other, or wipe each other out. You lose various demand signals.”
Keith Baranowski, global vice president and general manager of direct materials sourcing at SAP Ariba, says the term was broadly coined by AMR Research, which was acquired by Gartner in December 2009. From a profit perspective, he says, it was under the category of lean manufacturing, and worked well for certain industries to reach zero inventory. That doesn’t necessarily work today.
“I think if you look at extreme practitioners—hard line, never use a forecast, only go off of customer expectations—larger companies have the discipline to make it work. For some, it wasn’t practical,” he adds.
In the semiconductor industry, for example, the production process might take 12 to 16 weeks. That’s difficult with a fully lean approach.
“Customers want it in less time,” Baranowski says. “So, you have to forecast and pre-build. One way is to build up a wafer bank by forecast, but do the packaging by customer demand. We’re seeing a lot of that kind of thing.”
In what he describes as “mixed model” demand driven, some companies will use forecasts as a signal, but not a legal commitment like a purchase order. Then, after demand materializes, the purchase order (PO) can be written.
Eric Pfeiffer, senior director of supply chain integration at HAVI, says his supply chain management company began the journey to demand driven in 1997-98, triggered by limited time offerings (LTOs) in food services, menus, toys and gifts. “Demand and supply were not in balance,” he explains. “Retailers were not getting value in inventory, and marketing was not where it should be. We improved the system by using consumer demand at the retailer to establish a forecast to better enable the marketing plans. Have the items in the store in the quantity it is needed.”
Use the Data
Once upon a time, companies used spreadsheets to look at past results and make their forecasts. How long have we been doing this? How much did we do last month? Then, says Bodenstab, they smooth it out and allocate an extra 2 percent. “That’s 1990s thinking. They’re still thinking top down.”
With demand-driven supply chains, a more granular approach is required, and the data is there.
“What’s happening in the big box channels? At the convenience store?” Bodenstab asks. “Look at orange juice. Does Vitamin D enhanced sell more? Are people buying in smaller units and with more pulp? When do you sell, and in what form? Leverage that extra data, and create a more demand-driven supply chain.
“That doesn’t require big data,” he adds. “Just take advantage of [the data you have]. Maximize the data you can acquire. Then look at what’s happening on social media.”
Consumers are willing to share their opinions, so if you have a new product line, new flavoring or new packaging, social media can add to the information you have. Response to your Tweets and “likes” on your Facebook page can be early indicators of demand.
Pfeiffer agrees, pointing out that knowledge of the “general ecosystem” is important, leading to better predictions. That means visibility from production schedules, inventory, distribution centers, suppliers and outlets. “Use your plan to know where and when it needs to be there and make adjustments along the way,” he says. “You can quickly jump in to best support the network if there are interruptions. Monitor and know what’s going on with your suppliers, retailers, transportation, DCs or manufacturer.”
What the Consumer Wants
Gaming is big business, and Microsoft, of course, is a major player. The tech giant recently announced that it’s shifting from a forecast and allocation-driven model to a demand-driven model to “provide guaranteed availability of the right product to the right place before the customer’s tolerance time.”
In a webinar with SAP Ariba and Intrigo, Microsoft’s Ali Khaki, principal product manager: supply chain engineering (order management, fulfillment and logistics), explained why the move was being made. He cited inventory reduction, the ability to capture upside potential to make more of the products that are selling well, better balance of Microsoft’s demand and supply, as well as more reliable order confirmation dates and an improved overall customer experience. As part of the shift, Microsoft is using the Ariba Network as the backbone of its Xbox and Surface tablet line of products supply chain.
Under the legacy system, a forecast spreadsheet was sent via email, then prioritization was conducted primarily by telephon, and supportability was conducted by meetings and emails. That resulted in limited traceability, a highly manual process, error prone communications and limited visibility for contract manufacturers to changes in demand/priority.
With a demand-driven supply chain, traceability is improved, the discussion revolves around exceptions, and the process is highly automated and scalable to multi-tiers. Additionally, there is better visibility of consignment parts to the contract manufacturer as the process is being leveraged.
How has it worked out? Here’s what Microsoft reports:
- The supplier supportability process improved by more than 90 percent (1 to 2 hours reduced to approximately 5 minutes).
- Planner efficiency improved 95 percent (order follow-ups/expediting was reduced from 1 to 2 days to 1 to 2 hours).
- Supplier on-boarding time improved 75 percent (time reduced from 3 to 4 months to 3 to 4 weeks).
“There are now new tools and mechanisms to digitize the process between buyer and supplier,” says Baranowski. “Where is your inventory? What state is it in: finished goods or raw products? One benefit is that a digitized process changes the way you collaborate. It’s more exception based. You’re looking at the same numbers, but it’s more value-added, rather than chasing spread sheets. There’s more response time. They buyer and supplier are equal now.”
Now Hear This
Amplifon Group has a 9 percent global market share in the hearing solutions space, operating in 22 countries on five continents. The company operates in a highly competitive, diverse and fragmented market. It’s diverse because it has 2,100 of its own retail outlets, 3,200 “shop-in-shops” and “corners” within a widespread network of third parties (pharmacies, opticians and medical practices) and a network of more than 3,100 indirect points of sale, franchises and affiliates.
It’s fragmented because it operates differently in each region. In the United States, where it is known for its Miracle Ear brand, Amplifon’s suppliers manage all purchasing and logistics in a B2B model. In Europe, the Middle East and Africa (EMEA), suppliers and central warehouses in the Netherlands and Italy distribute products directly to retailers in a B2C model.
Over time Amplifon lost full oversight over its inventory, particularly which products needed replenishing and the assortment of products to be distributed to the stores.
Alessandro Nobile, EMEA supply chain director, says, “Amplifon’s supply chain doesn’t end at the point of sale. It’s a forensic process that involves monitoring the whole product life cycle, and includes reverse logistics of trial products. Our top priority is to automate and streamline procurement and logistics at the point of sale in order to reduce in-store stock levels and significantly improve the rotation of peripheral and central warehouse inventory.”
In 2014, Amplifon reviewed its logistics and business processes and implemented a demand-driven supply chain with ToolsGroup to manage planning and replenishment. Amplifon today has end-to-end control of the supply chain and complete visibility of all operations because data is sensed at every stage of distribution, including point of sale (POS) and by channel. Planners are able to identify demand trends, trigger alarm signals, and improve responsiveness between planned and actual supply chain events, while providing detailed forecasts and determining optimal inventory levels.
Amplifon is able to determine the right mix of products in stock as well as those in high rotation, such as hearing solutions, connectivity tools, batteries, cleaning accessories and headsets, for each point of sale. Replenishment is done automatically according to each retailer’s requirements, pace and sales strategy. All this also increases service levels across the supply chain while achieving balance between inventory safety stock and holding costs.
The results? By allowing Amplifon’s sales forecast data to be integrated with short-term demand data, forecast error has been reduced by 30 to 50 percent, while still allocating stock across its distribution chain. Amplifon reports that it has slashed inventory by 18 percent and obsolescence by 40 percent, while improving quality.
“Today, branch logistics is much more streamlined,” Nobile says. “And the work carried out by our most experienced logistics teams is completely integrated into the other business functions. This amplifies our ability to deliver real added value throughout the supply chain.”