What keeps you up at night? The answer for a lot of supply chain professionals in healthcare and life sciences is how to deal with the rise in high-impact uncertain events and demand volatility. Regulatory approval risk, uncertain product launch uptakes and rising product-mix complexity make planning harder than ever, driving up stock levels and eroding profitability.

How a company responds to these challenges can create a distinct competitive differentiator in terms of service and financial performance. Leaders are turning to other industries for best practices and field-proven solutions, and particularly to the consumer products sector, where low margins have led manufacturers to aggressively pursue demand-driven supply chain strategies.

It turns out that automated data analytics solutions, originally designed to understand consumer behaviour in fast-moving markets, are highly applicable to predicting patient demand for drugs and medical devices. And, in doing so, data analytics provide welcome relief with a healthy dose of certainty in an increasingly uncertain market.

Signal in the noise

In the past few years, the supply chain for virtually every industry has become flooded with data; this torrent is a blessing and a curse. While the sheer scale overwhelms traditional systems, valuable information hidden in the clutter provides the key to understanding complex customer behaviour and making better decisions in a rapidly changing business landscape.

Meanwhile, shorter product cycles, rising complexity and increased volatility are putting new pressure on supply chains to become more efficient and agile. Investor calls for growth and profitability, combined with the looming patent cliff, threatening billions in sales, are starting to reshape traditional practices in favour of more agile and leaner business models.

The days of masking inefficiencies by simply throwing inventory at the problem are coming to an end. It is more important than ever that supply chains drive competitive advantage, and it starts by making sense of all that data.

The use of automated analytics has now become a strategic priority for manufacturers. A recent survey from The Consumer Goods Forum and KPMG found that 56% of supply chain executives from all industry sectors ranked data analytics as "very or critically important" to their strategy this year, and according to IDC Manufacturing Insights: "Big data analytics will remain one of the top investment priorities for manufacturing organisations of all sizes for the foreseeable future."

Pharmaceutical supply chains are traditionally manufacturing-driven with large-batch, centralised manufacturing designed for blockbuster drugs. Manufacturing efficiency is maximised at the expense of responsiveness, and service is protected with a heavy-inventory model.

A closer look shows a complex supply chain comprising an active pharmaceutical ingredient component that is relatively easy to forecast and a finished product side that is much more difficult to predict.

Demand planning efforts are most often centred on the manufacturing flow of active ingredients, with finished-goods-inventory built on a poor demand signal and pushed downstream to distribution centres, hospitals and pharmacies. High safety-stock levels hurt return on capital and cash flow, and erode profits with high carrying costs and expensive write-offs. Inventory is often in the wrong location, contributing to localised shortages and missed revenue opportunities, as well as unplanned costs from trans-shipments and expedited freight to meet service commitments.

The financial impact is felt across the traditional high-margin, high-lead-time and date-sensitive prescription business, and low-margin, high-volume over-the-counter sales.

Senior management has started to recognise that the decades-old ways of forecasting used by their teams are simply not working and, more importantly, fail to meet strategic corporate objectives. They see peers and competitors embracing new technology to overcome these barriers and question how they could make better business decisions with accurate forecasts.

The answer, according to McKinsey & Company, is that the healthcare sector has the potential to improve margins by $130 billion by adopting supply chain advances that are established in other industries, such as fast-moving consumer goods. The study found that consumer products companies are more than a decade ahead of life sciences in terms of supply chain capabilities and performance, carry half to a quarter of the inventory, experience six times less obsolescence and are 15 times faster in manufacturing responsiveness.

A new way

The shift from a push to a pull network repositions the supply chain from a cost centre to a core strategic advantage, as does the demand planning transition from using prior years’ performance or extrapolated annual budgets to advanced analytics driven by current data.

The systematic use of current data provides new visibility to what is actually happening in the supply chain, and allows companies to make the right decisions first time for key deployment, manufacturing and sourcing activities, instead of taking a best guess and constantly replanning.

Equally important is forecasting at the right level. Monthly national demand forecasts for product families may be sufficient for capacity planning but fail to provide the granularity required for efficient supply chain operations. Aggregated forecasting is much easier and looks impressive but can give a false sense of achievement.

"The days of masking inefficiencies by simply throwing inventory at the problem are coming to an end. It is more important than ever that supply chains drive competitive advantage, and it starts by making sense of all that data."

For instance, a monthly base national forecast with greater than 80% accuracy typically translates into weekly item-location forecast accuracy of 40% or lower. Most importantly, it is far from the patient ‘moment of truth’; a patient undergoing an emergency procedure doesn’t care if the specific medical device required is available at a hospital in a neighbouring town an hour away – it is needed right here, right now.

Advanced demand-sensing analytics overcome the limitations of traditional statistical engines by identifying patterns within the wealth of available supply chain data to create accurate demand predictions for each item at each stocking location. Sensing demand is, by nature, automated. The vast amounts of data, the increased granularity and daily frequency requires automation – there is simply too much to do manually without an army of planners and statisticians. The software needs to automatically process all this data within a few hours, create accurate forecasts in tune with market realities, and publish them directly to supply planning systems without human intervention and review.

Findings from an annual forecasting benchmark study encompassing $150 billion in consumer products sales show that using current data within the manufacturer’s network to sense demand provides an average 38% reduction in weekly item-location forecast error. Expanding the dataset to include downstream data from retailer partners typically lowers forecast error by another 30% or more. For life sciences networks, think hospitals and pharmacies.

In addition to accuracy, manufacturers gain two to three weeks of forward visibility. Advance notice of fluctuations in demand provide the opportunity to adjust manufacturing schedules before the receipt of an influx of otherwise unexpected orders. For example, during the H1N1 outbreak, a multinational tissue manufacturer was able to sense rising demand as the virus spread across the country, ramp-up production in time to ensure on-shelf availability and successfully capture market share from competitors caught unaware.

The bottom line

The financial advantages are compelling. Fast-moving consumer goods companies that sense demand have a consistently stronger year-on-year performance. To illustrate, 30 multinational manufacturers in the packaged goods, food, and health and beauty sectors were split into two groups: companies choosing to sense demand and those still relying on traditional statistical forecasting methods. Published financials from annual reports were compared for the five-year period of 2009-13 to show relative performance.

  • Inventory: demand-sensing companies (DSCs) showed a considerable inventory advantage, cutting five days of inventory, or 7%, compared with just one day for the other companies.
  • Cash conversion cycle: DSCs were more effective at converting their resources into cash, with the cash conversion cycle metric cut by 18 days, 70%, compared with just 11 days.
  • Revenue and gross margin: likewise, DSCs had impressive revenue and gross margin advantages, with sales growth of 20% in the past five years compared with only 13% for the other companies. They also maintained consistent margins rather than the 6% decline seen by companies that stuck with traditional techniques.

The financial potential for life sciences companies is even greater. Using industry inventory and obsolescence averages from McKinsey & Company, each 1% reduction in forecast error for a €10-billion company frees roughly €20 million in cash through one-time capital reductions and €3 million in annual savings. With a conservative 30% reduction in error by sensing demand, a typical life sciences company could free roughly €600 million in cash and increase profits by €90 million for each €10 billion in revenue.

The prescription for profit is clear: sensing demand leads to healthy performance in volatile markets. The stakes are high. First-movers in life sciences will enjoy a tangible competitive advantage, while companies that cling to traditional approaches risk being outpaced within in a few years, only to realise it is too late to catch up. So why not get ahead of the pack?