Inventory Optimization: Win the War by Enhancing ERP and SCM Systems with Analytics

Oct. 6, 2011
The grandest product innovation and savviest marketing efforts can be undermined by a weak inventory management approach that results in lost sales or high holding costs.

I'm a history buff, and that comes in handy when looking at the problems of inventory optimization. The whole idea of getting inventory to the right distribution center at the right time bears a strong resemblance to a military's need to get supplies and troops to the right place at the right time.

In fact, modern operations research or decision management, defined as a quantitative methodology for operational management, came out of the WWII British Royal Air Force. Nobel Laureate Patrick Blackett and his select research team optimized the use of anti-aircraft ammunition and other problems. Even farther back, Alexander the Great conquered vast swaths of the known world by optimizing his supply lines. Later the Romans, learning from Alexander, perfected extended supply chains while devising new ways to cripple their enemies' supply lines. Napoleon understood the importance of supply chain management. He famously said that "an army marches on its stomach." (Ironically he died of stomach cancer.) But he knew a well-administered army is necessary for success.

The most dramatic example of supply chain dominance for the United States was the costly U.S. Civil War. Although not all historians agree, there is broad consensus that the Confederacy had better generals (i.e. CEOs), but the Union won the war. Why? Because the Union focused on the supply chain. The Union better managed its supply and lines, and destroyed the Confederacy's supply and lines. Both often are credited as determining factors in the final outcome of the war.

Yes, CEOs -- and the rest of the C-suite -- you need to pay attention to moving inventory. The grandest product innovation and savviest marketing efforts can be undermined by a weak inventory management approach that results in lost sales or high holding costs. And while both enterprise resource planning (ERP) and supply chain management (SCM) systems have made inventory optimization possible, without a forecasting component companies still risk suffering the "bullwhip effect."

Taming the Bull
The bullwhip effect, also known as the Forrester Effect, is an observable phenomenon of how order generation is magnified up the supply chain from consumer to raw materials suppliers resulting in unwanted quantities of inventory. The inflated stock levels set off an expensive discounting and promotional cycle on top of the expense of carrying costs. When combined, this means lost profits.

It goes like this: When one additional pair of a certain style of athletic shoes is purchased from Store A, it triggers an automatic order for a pallet load from Distribution Center 1. If that order is over the predetermined Distribution Center threshold, it sends an order to the Regional Distribution Center for a truck load of shoes, and so on. It affects everybody involved from the factories to the materials suppliers.

The move from the push environment, where suppliers have relative control over their inventories, to the customer-dictated pull environment, complicates matters even more. Going from push to pull adds extra volatility to the ERP and SCM systems carrying threshold triggers. While the bullwhip effect can be countered with regular, correlated stock orders instead of batch ordering and price smoothing to counter price-volatile ordering, these measures don't get to the root of the problem. Manufacturers hedge their bets by maintaining safety stocks to counteract the potential disruptive forces of unanticipated customer demand.

ERP and SCM systems started to offer advanced planning and scheduling functionality to counter the bullwhip effect, but these transactional systems never were designed to be analytical platforms. They deliver improvements, but they could be more robust with the addition of advanced forecasting and optimization.

Add Some Sophistication to your Inventory Management
A better approach is to work smarter by layering multi-echelon inventory optimization on top of existing operational system forecasts. Companies are beginning to look at this approach and are achieving excellent results: more than a 15% reduction in inventory costs.

Multi-echelon inventory optimization works because the algorithms at its core replace rule-of-thumb inventory policy parameters (min/max and sequential optimization). Algorithms customized to every product and location pairing effectively manage inventory by optimizing the costs of buying, holding, producing and selling inventory.

Multi-echelon inventory optimization operates by creating matrixes that produce organizational business rules. The customized policies link together those islands of efficiencies so they act in concert with the customer-facing location's service-level requirements. The inventory optimization system weighs the fixed ordering, unit, holding and potential penalty costs of not having enough stock for each product and location combination. It also accounts for the variables needed to create an inventory control parameter that determines inventory stock levels at various locations, timing for order placement, and order size for lowest costs (some inventory costs less to hold than others.)

The systematic multi-echelon inventory optimization process is data driven. Your organization needs the ability to manage large data volumes. In addition, the systematic approach should be able to achieve accurate forecasts at every level - top-down, bottom-up, middle-out - and also produce a variety of statistically based forecasts (short-term, long-term, new products, end-of-life, etc.) at frequent intervals.

In China, a company with retail and direct sales divisions used improved forecasting and inventory optimization to deal with a massive distribution issue. To predict demand, the inventory optimization solution looked at data from 70 million orders placed over three years. Based on relevant historical sales data, the system automatically analyzes, models, executes and adjusts predictions for products and regions at different levels, and forecasts demand by specific product and outlet. The result: A 20% improvement in replenishment times, a decrease in stockouts and reduced inventory costs. The stockout and replenishment improvements contributed to a 97% customer satisfaction improvement.

While military commanders have known for centuries about the importance of optimizing supply lines, many companies are just now recognizing that pumping up inventory safety stocks is not enough to be cost competitive and win needed market battles.

Optimizing stocks, just like Patrick Blackett did in WWII, means a more cost-effective approach that preserves precious capital while still defeating the enemy. Companies also can keep their inventory lean and avoid lost sales and market share. Doing so takes forecasting and optimizing, but that's a relatively minor investment in light of the huge costs already sunk into existing ERP and SCM systems.


Mike Newkirk is the, Global Manufacturing Product Marketing Director forSAS.

Sponsored Recommendations

Voice your opinion!

To join the conversation, and become an exclusive member of IndustryWeek, create an account today!