IntegriChain CEO Kevin Leininger participated in a panel discussion at Phorum Philadelphia 2015 addressing how big data can improve the customer experience. We asked Kevin to also offer his insights here for those who could not attend the conference.

Due to a variety of factors including changing healthcare laws and economics, inventories have shrunk dramatically in the US healthcare channel in the last few years. As a consequence, stockouts and product availability at point of consumption have been an increasingly big problem.

The driving question is: how can large healthcare manufacturers optimize the performance of their distribution channel given smaller inventories, less safety stock, more generic product at much lower prices, etc. while not increasing risk to the consumer of not having a product when they need it? When the cancer patient shows up for an infusion, how can healthcare manufacturers ensure that their customers (i.e., patients) have the product on the shelf when they need it? That is a difficult challenge.

How Big Data Addresses Healthcare Industry Challenges
The IntegriChain Cloud takes in millions of transactions per day, totaling more than $175 billion in commerce and 1.5 billion transactions in 2014 alone. We provide an analytics environment that allows healthcare manufacturers to mine the data by running reports but also – perhaps more interestingly – by running analytics against this data in real-time. We keep everything in memory so the speed is tremendous. Because of this, users –of which there are hundreds in the system at any time – can query the data and get insights into activities in the channel quickly and easily.

Big Data Use Cases for Healthcare Manufacturers
A typical use case involves the real-time monitoring of inventory positions across the entire US healthcare distribution channel. Healthcare manufacturers mine this data each and every day watching inventory positions across the entire us healthcare channel. If and when inventories drop below a defined threshold, they call their trading partner, which immediately moves inventory from one location to another and orders more product from the manufacturer the same day. One customer has taken service levels from 97% to more 99%…THAT is customer service.

Real World Use Case 1: While analytics is important, making a difference to healthcare consumers is all about taking action. A little more than 2.5 years ago, a top-five healthcare manufacturer approached us and said, “We pay our trading partners hundreds of millions of dollars per year and with the new regulation and shrinking inventories, this is becoming a challenge we can no longer deal with ourselves. We need to use our data to be more data driven and efficient. However, there is a lot of data. Help.”

So, we built a system we now call IntegriChain Scorecard, which is in use at a growing number of healthcare manufacturers. IntegriChain Scorecard is used to pay for performance in the healthcare channel. It analyzes huge amounts of data, all in memory, to allow for near-real-time performance measurement of the channel itself. Manufacturers use it not only to pay their trading partners for optimal performance in the channel but more and more in a CPFR (collaborative planning, forecasting and replenishment) context. Given that we have access to so much data and can operate on the data so quickly, manufacturers are starting to run the scorecards weekly to monitor the performance in the channel and proactively make calls to their trading partners sharing where there are inefficiencies and problems brewing in the channel. This allows them to proactively address challenges with product availability in the channel before problems occur. All of this is about improving the quality of customer service in the healthcare channel.

Real World Use Case 2: About two years ago, another top-five healthcare manufacturer came to us and said, “It is great that you can help us manage our inventories, reduce stockouts, and improve patient health. We love it. However, what we really want to do is have your cloud use demand sensing to automatically control order flow when we purchase products from you.”

So we built an “outside in” order management system that, using those millions of daily transactions from the channel, senses demand changes and inventory positions and either automatically approves purchase orders or holds those purchase requests for the customer service team at this manufacturer to review. Instead of having to go, as they previously did, to eight different systems to review inventories, sales, and inventory trends, prior purchase patterns from the requesting trade partner, and then determine what to do with the order, they now simply refer to the contextual reporting built into their customer service experience. With the click of a button they can see order patterns, changing demand patterns, and inventory positions across their channel instantly. From this, they make a decision in seconds that used to take them hours. They either deny the order, approve the order, or allocate the order based on demand changes and inventories as they see them from the cloud. All of this is based on analyzing hundreds of millions of their own commerce transactions in real-time to help them make that determination. This is big data and has a big impact on the customer experience downstream in the channel.

Real World Use Case 3: One final story. Just more than a year ago, one of customers came to us and said that they loved what we had built. They could now see inventory positions and demand changes and had begun to manage their distribution channel in near real-time. They had improved service levels by more than 1.5%, reduced costs, and dramatically improved customer service since using their big data in the IntegriChain Cloud. However, they wanted to do something more. Instead of just taking that purchase order in and analyzing it, through customer service, they wanted the system itself to make a recommendation as to what to do with the order. In addition, they wanted the system to tell them not only what to do with the order sitting in front of them, but also, they wanted to know what to expect to see in the coming weeks for orders. They wanted the system to, by using the vast amounts of data they now received from the channel, tell them what they should expect as orders from their trading partners for the next few months on a daily basis.

So we built a predictive Order Planning engine in Hadoop and Revolution Analytics R. That system processes daily store level inventory and point-of-sale information against 10 predictive models each and every day. It scores the results and makes a daily recommendation for the orders they should expect in from their channel. By the end of 2015, we will be running more than six billion computations in a four-hour window resulting in one Demand Forecast and one Order Plan that the customer service teams use for the coming day and daily order plans and demand forecasts for each and every location in the US for the coming 13 weeks. THIS IS BIG DATA.

As we roll this into production, we expect this to have a tremendous effect on our customer’s ability to make sure that the right product is in the right location at the right time. This will significantly improve the industry’s ability to ensure that their patients and consumers never get to a location looking for a critical healthcare product and not have it readily available for them.

The cost and quality of healthcare impacts each and every one of us. Big data is playing an increasing role in making healthcare more data driven, thereby reducing costs and improving execution in the healthcare channel.

Tags: big data, channel collaboration, Distribution Agreement Scorecarding, Forecast Accuracy