Grow top-line revenue and optimize inventory in pharmaceuticals drug manufacturing using SAS Analytics

Pharmaceutical organizations in the business of manufacturing formulations/APIs face a common challenge - too much of inventory locked across raw materials, finished products, spare parts at various plants, warehouses, in-transit. Raw materials include APIs, excipients and others. Finished product includes oral solid dosage, injectables, inhalers, ointments, etc.

Furthermore, organizations operate in multiple business models – “make to order (MTO)”, “make to stock (MTS)”, etc. Add to that dimension like which SKU is likely to be produced for which market – this determines the procurement location of raw material, its storage, and which plant it should be destined to.

And the impact of global macro factors affecting the supply chain – economic activity, warzones, disruptions in shipping routes etc.

The key essence for an organization is to maintain a lean working capital profile which helps in reducing inventory related revenue risk and creates a positive impact on top line and bottom-line performance.

What is the need of hour

Planners plan their demand 12-16 weeks ahead. Based on their projections or plans, the upstream supply processes prepare to fulfil the demand plan. However, volatility in customer ordering patterns, near term changes in environment, affect the accuracy of these demand projections. All the planning systems have built-in tolerances for such inaccuracies. What hurts enterprises the most are systemic biases in the demand projections. These biases are within tolerances on a period-to-period basis, but in medium term, they start straining the system. Chronic over forecasting builds up stocks that they cannot sell and creates an inventory risk while chronic under forecasting results in low order fulfilment, loss of revenue and creates a revenue risks.

In a survey conducted on 130+ Pharma companies, more than 50% supply chain executives responded they still use excel for their demand forecasting process

For instance, let’s say a generic Telmisartan API has 4 grades, US FDA, EMA, IP, and WHO. Between IP and US FDA grades the price could vary up to 100% with EMA and WHO grades in between. So in reality, Manufacturer must maintain separate stock of the API by grade and by source. This all adds up to the locked capital.

Then there is week to week price difference. Knowing the demand accurately in advance can help timing the purchase and hence save some procurement costs.

Forecasts have inherent uncertainties. The reason: natural error within the demand signals which are more noise than signals.

In MTS business, another retailer provides you with their forecasts but month on month their adherence to their own forecast is between 40-70% and then sometimes it is 150%.

This is where, the traditional tools are getting challenged.

How Enterprise Analytics Platform can help

A good fulfilment system factors these uncertainties to a level and establish stocking norms that balance the cost of losing money – revenue risks, and cost of capital, obsolescence, and expiration costs – inventory risks.

For domestic markets, the supply chain is very deep and multi-tiered. Imagine inventories built-up at different tiers, and the noise getting amplified upstream. The worst part is that the demand signals lose their shape by the time they reach the supply point.

Best in class enterprises invest in Data Analytics technologies that can maintain close synchronization between demand and supply. They ensure that their planning process is efficient, and accuracies are predictable. This requires a close collaboration between the frontline and the demand and supply planning team.

SAS Intelligent Planning solution helps in demand and supply forecasting, demand-supply planning and consensus management, new product forecasting, Product optimization, and advanced dashboards and reporting.

One of the key foundation pillars of intelligent planning solution is to provide better results with low efforts – people, processes, technology, or services.

An industry survey has revealed that Supply chain executives have less confidence on statistical forecasts of their SKUs; more than 70% responded either they make lot of changes in the statistical forecast, or they use it as a guidance and end up creating it manually.

Demand - Supply plans can be expressed with different unit of measures, be converted to financial plans in different currencies, provide comprehensive reconciliations, and integrate profit and revenue planning as well as price optimization. Forecast value add guides stakeholder demand planning with optimal collaboration.

SAS Intelligent Planning Solution is designed for helping customers meet these objectives and it is trusted by hundreds of customers all over the world.

What about existing investments in demand planning and ERP solutions

A typical IT operating model employed by a customer is to do material planning in their ERP. For S&OP planning they often use other third-party software like O9, IBP, Kinaxis, or excel sheets. Or they are building such solutions using open-source technologies.

While SAS offers an integrated solution – including data ingestion, transformation, building analytical models and deploying in production, it can be deployed in a modular manner.

In fact, many of SAS’ forecasting customers use the forecasting platform in conjunction with their existing tools.

SAS augments investment in ERP and MRP systems.

SAS has time-tested constraint programming procedures or tools and algorithms to explode the BOM. Using these algorithms, you can augment your MPS, RCCP, and Material Requirement Planning processes.

On Open-source technologies, see they are free, but they are very costly. This may sound contradictory, but this is the reality. For smaller scale, as experiments, they are extremely effective. But for an enterprise class use, these open-source technologies must be augmented for orchestration and operationalization.

So, we are back to a situation where we require enterprise platforms like Data Bricks for orchestrating the opensource Apache Spark to support enterprise class applications.

Another practical issue is with the cost of operations. Many of these technologies are inherently inefficient and require massive parallelization to scale. This results in higher cloud costs.

But these technologies are extremely popular with a large citizen scientist ecosystem.

SAS has python extension. So, you need not learn SAS scripting. Using python, you can leverage SAS platform orchestration capabilities and its analytics procedures, either using a notebook or it’s no/low code interface.

Summary

Enterprises can leverage SAS’ advanced analytics platform to supercharge their sales and operations planning process. SAS Intelligent Planning has cutting edge capabilities to reduce manual effort at the same time, provide an agile ecosystem to design advanced statistical forecasting models thereby increasing trust and transparency in the business functions.