Next wave: how pharma analytics can be improved with new technologies

Pharmaceutical companies have always had access to a steady stream of data to look at what has happened in the past and to try to predict future prescribing trends.

Business intelligence (BI) departments have supported this throughout with timely and effective reporting, within an environment that has seen in recent years a bit of an ‘arms race’ with BI tools adding an increasing array of chart types and functionalities.

To date this has been typified by the visual approach of the ‘fish tank’ chart. But now technology – specifically artificial intelligence (AI) and machine learning – is poised to offer new ways of analysing and processing data, allowing the pharmaceutical industry’s use of analytics to step up a gear.

Business intelligence analytics today

BI provides key metrics for pharma companies to track sales performance over time, whether through market share, contact rates or other endpoints.

You absolutely do need to know what worked in the past when you’re making your future plans, but the various retrospective figures that have been available to pharma to date can only show occurrences that have been and gone.

Different metrics have come into fashion and then departed, with some even coming back around again. However, they only look at the traditional questions companies ask of their sales teams: Are we doing well? Are we hitting our targets? Are we growing? How do we compare with the competition?

Meanwhile, recent years have seen some major changes in the types of information that is available to those in pharma who assess sales and marketing performance.

Traditional NHS prescribing data has been augmented by information on biosimilar uptake across the health service, real-world data and other sources, while the data sets available to pharma have also increased in size. The advent of this big data means the typical pharma sales rep might now receive up to 4,000 data points a month, depending on the size of their territory and the number of competitor products or packs in their markets.

But there are limits to the insights that such large data sets, on their own, can bring to the industry – not least because diving fully into all of the data that is available would be a full-time job in itself.

Why we need to improve current BI tools

To make the most of modern-day analytics requires a new approach. The users of these data sets fall into a number of different types, all of whom must be catered for, but typically they’re all non-analysts. Our core users come from pharma sales and marketing, and it’s important we give them as much value from the data in the time they can spare from their regular duties.

In this way we can help up everyone’s game so that they can in turn have a bigger impact on business performance. What we’re trying to do as a consultancy is shift that curve a little bit, so that everyday users – as much as super users – benefit from these tools.

Timeliness is another area where improvements are needed. The worth of current business intelligence tools has long been proved, but they’ve had to focus on what has happened in the past and, within this, deal with time lags with the data.

Even the most up to date mainstream sales and market data will only arrive at the end of the following month, which in practice means a one to two-month lag on the period it covers. It’s great to learn from the past, and an important part of how analytics should be used, but it’s also a side of business intelligence that can be further enhanced.

New BI technology for pharma

To date, technology has been a limiting factor for development. Business intelligence has always been haunted by this to some extent, but tech’s continual advances mean that it will get better. As it does pharma should be looking for improvements to come from the insights it can uncover from the data, and particularly by combining large datasets.

With the ever-increasing size and number of datasets that are available, new technology can provide a hugely valuable ‘noise cancelling for BI’ role, allowing those in pharma to cut through the white noise to get to the relevant information. It’s here that machine learning can come into its own, doing some of the heavy lifting that your data requires; if the thousands and thousands of data points it offers are to be made sense of.

At the same, applying AI to the data can start to reveal the hidden patterns from the data sets in a way that just isn’t possible when an individual has to click through 100 bricks or 200 practices and look at every pack or product prescribed to try and decide if something has happened that’s interesting. There are a wealth of different hidden patterns in the data that the human eye won’t know are there, while the machine won’t rest until they are found.

“Further value might be found as we start to assess what the post-COVID future might look like, and combining AI and advanced analytics will allow pharma companies to measure, monitor and predict this”

 

Advancing analytics to provide future value

Looking for patterns in the data, and at what might happen in the future, is all about helping pharma to ‘find the interesting’ in the data, and the technology that facilitates this can also free up users’ time by providing them with quicker answers.

Among those answers might be directions to redirect the marketing strategy based on the data, or to institute a wider adjustment in sales and marketing team behaviour to drive tactical change on the ground.

Further value might be found as we start to assess what the post-COVID future might look like, and combining AI and advanced analytics will allow pharma companies to measure, monitor and predict this. Certainly no AI predicted COVID-19 and the devastation it would cause, but it could assess the virus’ impact on different diseases, therapies and NHS locations.

However, as with any use of new technology, it’s vital that pharma benefit from it and, with so much talked about in AI, there is a real need to avoid ‘AI atrophy’ when solutions are built and implemented before any assessment has been conducted of where they will add value.

Answering pharma’s big questions with tech-enabled BI

How will COVID-19 change prescribing patterns, what impact will a new formulary have on physician decision-making and how will market dynamics change when a new product is launched? These are some of the big questions that a tech-enabled approach to BI analytics might answer.

At the centre of this process will be the use of machines to guide and power-up human decision-making so that pharmaceutical sales and marketing teams can look to the future, as well as the past, processing more data, more quickly than ever before.

Technology is going to do a lot of the heavy lifting for BI professionals in the future and they will also be able to give it more lifting to do as they seek to solve specific problems for their organisation. As this happens it will also provide a welcome dose of ‘de-risking’, removing elements of human error that can sometimes creep into the data.

The future of pharma analytics is about getting people to answers – and questions – quicker, so the time they spend using the next wave of BI tools can have a positive impact on the future performance of their organisation.

About the interviewee

Lee Ronan is commercial director at CSL. Lee has worked in healthcare business intelligence since 2002, beginning as an analyst and CRM admin before spending time in an SFE role as well as working on secondment as a medical rep.

He has a passion for helping clients use data and visualisations to make informed decisions – Lee’s experience in the field gives him a unique insight into the challenges and opportunities offered by the healthcare sector.

Having previously served on the British Healthcare Business Intelligence Association (BHBIA) board, Lee is now a member of the Best of Business Intelligence (BOBI) committee with a focus on organising the BHBIA Analyst of the Year competition, as well as the Newcomer awards.