Does Perfect Customer Service Make Sense?

In supply chain management, it is widely accepted that holding enough finished goods inventory to fulfill every order is just too costly. The thinking goes that at some point, improving service levels means you are not adding inventory in a linear manner but rather exponentially. Ipsen, a global bio-pharmaceutical company with $3 billion in annual revenues, disproves that.

Ronan Stephens, the Senior Vice President of Supply Chain Management and External Manufacturing, explained how the company set out on a journey to improve customer service while also reducing costs. While the company also has a consumer healthcare business, Mr. Stephens focus is the specialty drug business.

Ipsen produces drugs to treat oncology, neurology, and rare diseases. Many of the specialty drugs need to be refrigerated and have short lifespans. Further complicating their supply chain is the fact that they operate in a heavily regulated industry.

“Over the last few years,” Mr. Stephens explained. “Ipsen’s strategic future was characterized by sales growth and frequent product launches.” As the supply chain team contemplated this growth, they realized that they had to envision a future, not that many years out, where the company could be generating up to three times their current sales. Further, after a couple of recent acquisitions, the supply chain was already feeling the strain. Their on-time in-full (OTIF) performance, at 99.5%, was very good. “Our inventories were in line with benchmarks, but we knew that intensifying the pace of launches could become challenging in the long run.”

That’s why the supply chain team proactively kicked off a series of meetings. The team began by thinking about what their goal or goals should be; then they could understand what kind of supply chain they needed to build to achieve that purpose. In examining the supply chain they would need, they were looking at not just the coming year, but what need to be done for the coming five years. The team came to the decision, Mr. Stephens explained, that rather than incremental improvements in service, the ultimate goal should be to NEVER stock out. “Our culture is to never fail.”

Ipsen had two additional goals. First, to launch on government approval of a new drug. Secondly, to drive supply chain savings generate cash for the business. This article will focus on the actions Ipsen took to never stock out.

It was thought that the company probably did need to improve their supply chain planning (SCP) capabilities. A member of the supply chain team ran scenarios to see how fast the company could respond to a market change like a surge in sales in a given nation. Manufacturing would not have been able to respond to that kind of event for two months. “We could not manage our business with that tool,” Mr. Stephens realized.

A new supply chain planning system would make their supply chain more responsive. Ipsen also need to reduce their lead times. Long lead times left them exposed, they could burn through safety stock and not have enough inventory to meet an order.

Ipsen purchased the RapidResponse supply chain planning product from Kinaxis. The RapidResponse planning tool gives both the internal supply chain and external stakeholders visibility to the plan and the ability to rapidly adjust that plan. The solution also promotes collaboration. In short, planning reflects existing and emerging constraints and opportunities across the end-to-end supply chain. That is key to planning in a manner that will promote a responsive supply chain. They also improved their sales & operations planning process. The Kinaxis solution also provides robust decision support for their S&OP process.

To reduce lead times, Ipsen decoupled manufacturing from packaging. Previously their manufacturing sites for specialty drugs produced finished goods. These finished goods were composed of the drug and the specific regulatory information different nations required on the packages and in the inserts. The semi-finished product, the drug, could be produced with a two-month lead time.  In a pinch, if a factory worked overtime, they could get the lead time down to three weeks. But the finished goods, inclusive of nation specific packaging, had a lead time of six months. Further, the higher inventory levels reflective of longer lead times, meant that some portion of their stock would expire. Ipsen simply could not afford to carry enough inventory for smaller markets.

Ipsen established a relationship with a responsive, packaging partner that takes semi-finished product – the drugs – and then adds the nation-specific packaging. Decoupling their manufacturing from packaging reduced the six-month lead time down to just a few weeks.

Further, previously Ipsen had done distribution in-house. Internal distribution was not aligned with the growing volume of shipments they would need to support. Ipsen worked to relax this constraint by working with a new, responsive, distribution partner that would be able to scale to support the growing volumes.

The results? Though service levels were already very good versus external benchmarks, the number of stockouts dropped a further 75% from 2016 levels.  All of those stockouts were extremely small; Ipsen’s metrics counts a stock out for any product not immediately available in any nation, even if typical monthly sales are just a few units. Often the stock outs are not a planning issue. One failure, for example, related to an importation document that did not arrive on time. Whenever a stockout does occur, the company works the root cause to help make sure that mistake does not occur in the future.

Ipsen’s approach to continuous improvement is interesting; it mirrors the process they used in plants to promote safety. To reduce injuries, a plant measures fatalities. The next level down in the pyramid are injuries that lead to hospitalization. Under that are incidents requiring first aid. Then, at the lowest level of the pyramid, are near misses. “If you are constantly working the bottom of the pyramid, you never have a problem at the top,” Mr. Stephens pointed out.

They are using a similar continuous improvement, pyramid approach around stockouts. At the top of the pyramid are stock outs, then comes red alert near misses, then comes amber alert near misses. Any time the company’s safety stock is consumed faster than expected, amber or red alerts are raised. An amber alert is raised when, within a forecast planning horizon, it is projected that three quarters of the inventory has been consumed or will be consumed; a red alert comes up when it is projected that 100 percent of the inventory will be consumed before the end of the forecast period.

The SCP solution is providing the data for these alerts. These near misses are categorized by root cause. The company has visibility to demand and inventory. The company now knows, because of this system, that 55 percent of near misses are driven by forecast errors. Important inputs to the forecast come from the commercial team. Ipsen is weighing whether it makes more sense for the commercial team to own the forecasting process.

Alerts are shared with country managers, commercial heads, and the supply chain team. Alerts “drive an all hands on deck, let’s pull out all the stops and solve this” approach to problem resolution.

Covid, not surprisingly, created real problems for the supply chain. During Covid, Ipsen saw 70% spikes in sales in some nations. As passengers stopped flying, planes were grounded, and the belly capacity in the planes disappeared. Further, border disruptions occurred making shipping into a nation with high infection rates much more difficult. Infections occurred in their own plants, which took the plants down for two weeks.

For specialty care, Ipsen only has two manufacturing sites. As a plant faced downtime, the company used RapidResponse to understand the demand situation. They needed to understand which nation was potentially exposed to a stockout. There was visibility to the inventory at rest in storage locations and in transit.  This allowed the manufacturing sites to focus manufacturing efforts with pinpoint accuracy on products for markets where inventory levels were weaker or where volatility was higher.   “Really,” Mr. Stephens exclaimed, “we had quite heroic people in our manufacturing and supply teams.” The company has experienced no stockouts during the pandemic!

Service levels around specialty drugs should be different. If patients don’t have these drugs when they are needed, the consequences can be dire. But to be able to virtually eliminate stockouts without requiring large increases in the amount of inventory held, that is remarkable!