We regularly speak with Pharmaceutical execs around the globe and in the past few years have asked the following question to well over a thousand Pharmaceutical executives, “What is keeping you up at night?”
Top 7 reasons Pharma sales and marketing execs cannot sleep
1. Launch Success
Many people we speak with are concerned about launches and getting it right. The world has changed and having a good drug, and a large growing market, does not assure you of success. Some key concerns are around getting the strategy and tactics right for products in therapy areas or segments that they do not have strong data for, knowing the optimal sub-segments to target, as well as resourcing and ensuring that they are focusing their resourcing on the right channels in the optimal amounts.
Many teams have only ever had linear analyses, so they feel that without historical data on their brand and the market they are in, they cannot conduct analytics to understand the optimal channels and optimal budget allocations. Once upon a time that was true as companies only had access to linear analytics. Now we are in the age of non-linear artificial intelligence analytics, and all that was previously difficult due to a lack of historical data is now not only possible but yields impressively reliable results.
2. Pricing and Reimbursement
Almost every launch team has concerns around whether they have got the price right and whether they will get reimbursement. This is not an unfounded concern. The majority of the top 10 launch failures failed due to issues relating to these. When one examines how Pharma pricing teams price their drugs, the approaches tend to begin with an analysis of current reimbursement followed by research on willingness to pay based on various endpoints, and then advisory board interviews on real-world decisions and endpoints, costs and return analysis, and finally end with basic statistics on the data.
What is wrong with this picture? Firstly, “willingness to pay”. When, in the history of Pharma, has anyone ever got good results from asking people what they intend or what they are willing to pay? Really? Ever? It is not that people are deliberately lying, but there are a lot of factors that come into play that the interviewees are not consciously aware of which are influencing them in the real world.
In addition, statistics in linear approaches fail to deliver an adequate analysis of complex interactions between factors, so you will never get a strong representation of what a likely result is using these approaches. Do you think that all of the spectacular drug launch failures (e.g. Makena / K-V Pharmaceuticals, Provenge / Dendreon, Zaltrap and Multaq / Sanofi, Benlysta / Human Genome Sciences, Horizant / Xenoport, Krystexxa / Savient, Silenor / Somaxon, Brilinta / AstraZeneca and Anascorp/ Rare Disease Therapeutics) did not use a sound research and statistics based approach to their pricing?
I am sure they all used decent linear approaches with their data. I believe that all of these companies had a strong team which probably used pricing agencies to assist them, but all failed to understand the fact that the old approaches are insufficient to understand the complexities of today’s market environment to achieve a good result. To get really strong and reliable results you need to go ‘non-linear’ and move to artificial intelligence approaches.
3. Marketing Measurement and Cross-Channel Resource Allocation
Despite all the advances in marketing measurement, many in Pharma do not feel confident that they are accurately measuring their marketing, nor their marketing channel interactions, well enough. They feel fairly confident they can measure the individual activities to a degree but how these interact with each other, and how to allocate accurately taking account these interactions, is currently beyond most companies’ abilities. The reason for this is also obvious to me. Linear analyses are not capable of dissecting complex channel interactions with any level of accuracy.
4. Finding Pockets of Opportunity in brands that have Plateaued – Overall or in Segments
This is a common request and one that is simple with artificial intelligence approaches, which take every variable and sift hundreds of millions to billions (depending on the data used) of combinations in order to pull out the top differentiators that will grow a brand – either overall or within various segments. In fact, it now is the only approach we would recommend for this type of activity if you want accurate results.
5. How to Personalize Marketing Content across all Channels, both Digital and Traditional
Personalization is a challenge for many Pharma marketers but it really does not need to be anymore. Many people know how to personalize websites and digital assets according to the individual. This has been done for websites for almost 2 decades; I remember in the 90’s we were using BroadVision to do this for Pfizer. Later, companies added in Hubspot and Marketo as well as a plethora of other software which covered the personalization and tracking of their web properties.
However, we have come a long way since those earlier technologies with many new software systems available today to take the data generated from these systems and put it on steroids with AI. Today, companies want a lot more than just personalization of their sales force and website. Their ambition is to have an intelligent system which allows every sales and marketing interaction with individual target customers (physician, patient or payer) to be personalized and optimized for that customer – be it their sales calls, their website, events, advertising campaigns, forum interactions, bloggers, social media interactions – like Twitter – and more.
The way we approach this with our AI system is to use the data collected from all these touch points from a client’s existing systems (and there are already many software systems available that collect all this data now), link this to IMS data, Cegedim data, CRM data, advertising data, or data from any quarterly market research you are already doing, and then create an artificial intelligence powered app to analyze all this data constantly as it comes in and serve up the best content for the right targets as they interact with the any of these touch points.
6. Creation of Sustainable and Engaging Customer Experiences
This is another area of concern to many Pharma marketing execs, although we notice a lot of lip service but very little real results. How exactly can one do it so well that we outperform our competitors? One company we worked with in the US recently really did walk the talk. The project was for a new drug that actually was the same compound as the leading generic. What choice did they have but to ensure that their drug created the strongest customer experiences with the patient, the physicians and the payers. This was achieved by starting with focus groups to better understand the basic pain points, then creating a more focused research effort with each target group, and finally combining these with existing company data and sifting through it with AI to outline very detailed programs for each target group that would make the most difference in sales. The results proved the approach.
7. Sales Force Effectiveness
Despite dropping down in the list of top concerns, it still makes the top 10. The costs for sales forces are high, and easy access to physicians becomes less and less daily, so this is a key issue. What companies are seeking more of are automated sales force systems so that their reps deliver the right messages to the right customers at the right time to change that doctors prescribing.
Every company has data that can allow this issue to be rectified. We can create an artificial intelligence powered system that automatically pulls up detail tailor-made to that specific doctor’s drivers, and what one needs to convince them to use that brand. It would automatically change the content of the call depending on which doctor the rep was seeing.
The system is artificial intelligence based and gets more powerful day by day as every interaction with the CRM data (from Veeva, etc.) would feed into the AI engine and would be learning daily how to improve each interaction by tweaking each presentation as more data becomes available. It would become more and more powerful in terms of understanding how to change the doctor prescribing.
Conclusion
These areas essentially break down to money (spending it wisely to fund activities with limited resources available, ensuring your customers want your products from all your channels to get more revenue), customers (engaging and keeping them), and innovation (many items are innovative areas that few companies have mastered).
All of the areas above are a concern for an obvious reason; the methods being employed to inform the decisions made are not sufficient in today’s complex environment. Every single concern outlined is solvable with artificial intelligence techniques available now.
However, many companies are still relying on single and multivariate regression analysis that determines results based on correlations. We all know that correlation do not equal causation, and often correlation does not mean anything at all. An example proving this is the high correlation found between the Per capital cheese consumption and the number of people who died becoming entangled in their bed sheets. Another example is the number of people who drowned by falling into a pool with films Nicolas Cage appeared in (both examples by Tyler Vigen).
Are you currently seeing some flaws with linear based approaches? Are your results informed by analytics but still not performing? Once upon a time this is all we had, so we used it. However, now we have moved mathematically far beyond simple correlation and linear approaches and actually can determine all of these factors with astonishing accuracy using non-linear artificial intelligence approaches.
Anyone still relying on linear regression style approaches to inform decisions is essentially gambling with their brand(s). Maybe you will get lucky, but chances are, even if lucky now, you will not stay that way as these approaches are extremely unreliable. The last few years have advanced our knowledge so well, and yet Pharma teams in the main are failing to keep up. If you are not using non-linear artificial intelligence powered approaches to measure aspects based on your brands, then you are potentially lining yourself up for gross negligence – especially if you are an Analytics Manager, or Business Insights Manager.
For more information on anything in this article, please contact the author – Dr Andree Bates at Eularis: www.eularis.com. If you want to know how to get out of the linear analytics trap and power your analytics, speak with Dr Bates or anyone on the Eularis team.