Embracing AI in pharma sales

The future of pharmaceutical sales will rely increasingly on the combination of technology and data. Rasim Shah, Director of OKRA discusses embracing AI in pharma sales.

Those who embrace artificial intelligence (AI) will thrive and leap ahead in the intelligence race, those who don’t will fail to keep up.

The intelligence race is real; companies across all verticals are investing heavily in AI to drive sales and generate efficiencies across commercial functions. Algorithms that drive consumer purchasing decisions within the B2C environment have long proven to be successful for Netflix, Amazon and Uber. Thanks to this power to improve sales performance, McKinsey analysts in Harvard Business Review estimate that AI can create $1.4tn to $2.6tn of value in marketing and sales1.

The life science analytics market is set to experience staggering growth over the next five years, reaching an estimated $33.2bn by 2024/252. However, for those in the sales team, the growth in the power of analytics is coupled with practical challenges. The increasing difficulty for sales representatives is well documented, with up to 20% of doctors in the UK not accepting visits3 and at least 50% restricting access4. The question is, how does the pharma industry invest in the right AI solutions that support the process of delivering the right drug to the right patient at speed?


AI today

Delivering generic brand messages is no longer good enough, sales representatives must bring value to healthcare professionals (HCPs) that is personalised. Conversations need to be data-driven, based on truth rather than assumptions or just gut feel. Critical to supporting this personalisation is the ability to have systems that layer data, both structured and unstructured from a variety of sources – delivering powerful insights based on a granular understanding for example of patient populations, the environment and historic activity that drives outcome.

Much has been said about supporting the pharma sales representatives with CRM-based suggestions termed ‘Next Best Action’. But to make sales teams more intelligent, we need systems that deliver evidence from all the data and empower individuals with a broader view of the right set of actions. What we must avoid is telling the sales representatives what to do and risk removing all human intelligence. We must focus on addressing the real problems faced by sales representatives and help them answer their burning questions:

  • Who is the most urgent HCP to visit and why?
  • Where is the best opportunity to drive sales and why?
  • Who is my competition in this practice and how are sales growing?
  • How should I contact this practice and why – phone, face-to-face or email?

Successful implementation of AI systems beyond ‘Next Best Action’ will help representatives to truly support HCPs in navigating an increasingly complex and niche prescriptions market. Empowered by country-specific data, representatives must remain the vital link between innovative therapies and medications and the patients that need them most. In this increasingly competitive market it’s becoming clear that maximising the impact of each customer interaction is the best way to break through the noise and connect with HCPs4.

What does the future hold?


In 2020 and beyond explainability will underpin the outputs made by AI systems. However, these systems will only make sales teams intelligent if they can action the output, and they will only be empowered if they can trust the output. Why has the prediction, suggestion or recommendation been made?

In an industry built on facts and evidence, AI systems must be able to explain the reasons behind the outputs or they will fail to build the trust required. This is critical if you consider the difference in data sets and data granularity across different markets in Europe. Without explainability, the representative and the system can’t learn together.


Empower, not replace

Thousands of sales representatives on the road, having thousands of conversations every day across therapy areas means tens of thousands of data points. We must encourage representatives to give feedback directly into the system. This communication will use the representatives’ knowledge and create feedback mechanisms that generate a rich source of data to fuel future insights. This coupled with explainability will forge trust between the representative and the system. The best AI systems will be designed to empower representatives to take action with confidence, not replace them.


Be agile or get left behind

Sales teams can become the catalyst for life science companies to adapt quickly to change. The tools that representatives use must reflect the dynamic environment in which they work. 12-month, large scale, full country pilots that take years to implement must be a thing of the past. Validation of AI pilots for sales teams should not take longer than a few weeks.

AI can and is fueling competitive advantage within weeks and seeking out external AI vendors will help leaders of sales force teams accelerate speed of business through smarter decision-making, leading to faster execution. From the acquisition of data, system design, data modelling, feature extraction, prediction accuracy and testing, all can be done within a matter of weeks or faster, including the heavy lifting and cleaning of data.


Rasim Shah is Director of OKRA. To learn more about OKRA solutions email hello@okra.ai



1 Chui M., Henke N. & Miremadi M. (2018) Most of AI’s Business Uses Will Be in Two Areas. Harvard Business Review [online] [Accessed 11 February 2020] | 2 Markets & Markets (2019) Life Science Analytics Market by Type (Predictive, Descriptive, Prescriptive), Application (Marketing, Compliance, R&D, Pharmacovigilance, SCM), Component (Software, Service), Delivery (On Premise, Cloud), End User – Global Forecast to 2024 | 3 PricewaterhouseCoopers (2009) Pharma 2020: Marketing the future – Which path will you take? [online] [Accessed 11 February 2020] | 4 Khedkar P., Kalyan N. & Scott E. (2016) Sales Force Effectiveness in Pharma Is No Placebo [online] [Accessed 11 February 2020]


Read more about AI in pharma here >

Read more from this month’s Pf Magazine >