From the time we had computers, we have always had analytics. Whether we called it MIS (Management Information System) or Business Intelligence, we always wanted the ability to look back into the data that we had collected and get some information out of it.
We went from Business Warehouse to Business Analytics to Business Intelligence.
We were able to tell:
- We sell a large percentage of our flu vaccines at certain times of the year.
- We consume less insulation for packaging medicines in the winter months.
- Raw materials sourced from some select vendors results in lower product quality defects.
- The faster shipping lane is more expensive.
We used canned reports from the system, downloaded data into spread-sheets and even built our own reporting.
We even did some regression analysis to predict sales (or a different variable), depending on a few other pieces of information.
And we have extremely fast (HANA) databases, immense computing power and business understanding.
However, we were still falling short on answering some key questions such as:
- Which part of our marketing campaign contributes most to a sale? By how much?
- What should the price of a new drug to be launched in Europe be? Which market should it be launched in first?
- What other products can be bundled with a top-selling prescription medicine?
- What is the likelihood of getting FDA approval for the product that the new acquisition would bring?
Answers to questions such as these can be expected of Big Data analytics.
What is Big Data:
Big as in massively big data. Hundreds of millions of records. Some clean, mostly not-so-clean. Some structured, others unstructured. Some text, some videos, some images, etc. Some from within the company, some from externally available sources, still some others that we purchased (such as prescription data).
Big Data is not data that you can collect in a few tabs of an excel spread-sheet.
Big data must have an analytical engine (a set of complex algorithms) that can collectively examine this data for patterns. Such analytical engines would have artificial intelligence built-in, i.e. the more data they crunch, the better they get at crunching the same type of data.
Most importantly, the resulting insights should be predictive and substantial enough for significant business gains.
Here is a definition of big data from Wikipedia.
“Big data is a term for data sets that are so large or complex that traditional data processing application software is inadequate to deal with them. Big data challenges include capturing data, data storage, data analysis, search, sharing, transfer, visualization, querying, updating and information privacy.”
Where can we get Big Data?
In Life Sciences, here are some of the areas where companies can expect to collect Big Data:
- Product movement at a unit level can yield billions of data pieces on product movement (vaccines, biologics, small molecule drugs, veterinary medicine, medical devices, etc.)
- Customer/patient engagement with your online sales channel – drug page visited, question asked, drug added to online shopping cart, cart abandoned, check out page opened, but not submitted, etc.
- Customer/patient feedback on social media.
- Clinical trial patient enrolment and engagement.
- Machine performance over a period of time (x-ray, MRI scanner, dialysis, etc.).
- Historical data collected by organizations such as Centers for Disease Control and Prevention (CDC).
The value of big data analytics lies in the varied nature and sources of data that the analytics engine can process. Identification of multiple sources, therefore, becomes valuable.
Benefits of Big Data Analytics:
Sampling analysis is a key subject within Statistics. We use a representative sample from a population of data for analyses and then extend the findings to the larger population.
With big data analytics, I believe that sampling will no longer be needed, since we now have the tools to analyze the entire population of data, no matter how large.
For Life Sciences (and HealthCare) some of the benefits that accrue from Big Data analytics are as follows:
- Understand what patients do before they develop a condition that requires certain therapies. This can help the (manufacturing) company develop a new product for the patient that complements the existing offering. Or it can provide a suitable target company for acquisition.
- Understand how to price drugs for a particular market. Predict the highest price at which a company can win the tender say, in Europe. Reference “Tender systems for outpatient pharmaceuticals in the European Union” by Panos Kanavos, Liz Seeley and Sotiri Vandoros (London School of Economics).
- Know which step in the customer journey contributes most to the sale, and then invest greater efforts at that step.
- Based on a patient’s lifestyle and food habits, predict a future condition. Follow up with treatment that will prevent or delay that condition from occurring.
The evolution of technology and its current state is the biggest factor that enables such analytics.
Customer engagement and commerce solutions such as SAP Hybris help us understand customer behavior at a granular level. Propensity of customers to engage in online commerce and their preference for using mobile device makes a lot of big data available.
The quality and capability of IoT sensors (they can monitor, record and relay information ranging from temperature and humidity to geographical location, time spent at each stop along the route and even exposure to light). When such information is related to a machine learning equipped analytics engine such as SAP Leonardo, we can get rich prescriptive analytics.
Of course, the back-end ERP system for any enterprise is still necessary for managing core functions such as order management, master data, financials, inventory, pricing, etc. In the age of big data such systems must not only process exponentially larger volume of data, but do so within a fraction of second. Simplified processes in S/4 with the blazingly fast HANA databases become indispensable.
I am always on the lookout for creative applications of technology. If you come across any big data analytics stories or use cases, please share.