2021’s crystal ball: Six AI predictions that will shape a new commercial model
By Aktana management
Artificial intelligence (AI) is no longer a “nice-to-have” in life sciences today, especially with increasing pricing pressures, global competition, and operational complexity. Even COVID-19 has not dampened the AI fire as 38 percent of technology leaders report that their AI investment plans are unchanged, and one-third (32 percent) say the pandemic has accelerated their plans.
For more than a decade, the industry struggled to get its volumes of data – the fuel for all AI engines – in order, working to gather it all from across different places, often confined in spreadsheets and disconnected systems across the enterprise. Now, modern data warehouse solutions have emerged to help aggregate and structure data so that it is usable. And, in recent years, many life sciences companies piloted AI for different commercial use cases with promising results.
As we ring in the new year, AI will continue its evolution as it becomes more scalable across all customer-facing teams—including sales, marketing, and medical—as well as across brands, regions, and channels. Here, Aktana experts forecast upcoming AI developments that will help shape a new commercial model, making 2021 the year when AI shifts from early exploration to broad adoption.
AI will expand functional flexibility to drive more coordinated customer engagement.
Alan Kalton, general manager, EU
Many AI deployments in commercial life sciences have been isolated within individual markets to one brand or channel (i.e., sales). Now, convinced of AI’s power to help improve the customer experience, companies want to roll out their AI solutions on a larger scale across the portfolio, embracing more teams to enable a coordinated and optimized customer experience.
In 2021, flexible AI platforms with industry functionality that meets the specific needs of field and brand teams, as well as medical science liaisons (MSLs), will leapfrog generic solutions. These platforms will apply multi-faceted algorithms and business rules that accommodate the unique needs, actions, messages, and strategies of all customer-facing teams, providing everyone with the necessary visibility to better orchestrate communications to healthcare professionals.
This is how AI can rapidly scale. Plus, groups that have not historically been able to benefit from AI’s data-driven suggestions can become better positioned to coordinate their communications. MSLs, for example, will now be able to exchange rich customer insights with the company in a more structured way, adding data into the AI platform so outputs continue to get smarter.
Actionable AI will remain the holy grail but getting context right will ensure user adoption.
Jim Anderson, chief customer officer
The first intelligence platforms were meant to remove the burden of data analysis from life sciences sales teams and give brand marketers an effective way to communicate strategy to the field. Today, the same issues early AI solutions set out to solve still apply, but on a much grander scale. In 2021, AI must adapt to handle more channels, data, and complexity – evolving to capitalize on deeper levels of human insight and address the new challenges that arise.
Deep analytics has been around for a long time, but the challenge has always centered on the inability to make information actionable for the users “on the front lines.” AI has helped to speed analytics processing so that information is now delivered in real-time, which contributes to its utility. Advanced solutions also deliver AI outputs to users where they are when they need them. But, the most actionable AI layers machine learning algorithms with business rules that reflect on-the-ground context for greater relevance.
In 2021, expect greater focus on AI that is continuously learning to balance ideal outcomes against real-world constraints and capture the necessary context to orchestrate meaningful interactions between the brand and each HCP. Further, companies will leverage new tools like Explainable AI, which enhances user trust by increasing transparency in the AI “thought-process.” Adding plain-language explanations to AI suggestions improves the chance that users will act on the system’s recommendations, but it also empowers them to provide clarifying feedback if those recommendations do not feel right. This, too, will help the AI platform get smarter – more data in, better data out.
In 2021, data science teams will begin to shift from centralized to specialized.
Pini Ben-Or, chief science officer
The pharmaceutical industry remains behind other industries in applying advanced data analytics, but it is catching up quickly now. Pharmaceutical companies have begun to decentralize data science efforts, opening the door to specialized teams who can focus on developing dedicated solutions that are in-production over time. For the most part, however, life sciences’ data science teams are still smaller than they should be to successfully apply AI across a wide range of problems in different business functions – from clinical research, product manufacturing, and pharmacovigilance to market research. This will change in 2021.
This year, the data science team will expand meaningfully, and specialized subgroups will emerge within the broader team. For instance, today’s data scientists are not directly focused on specific marketing issues like the complex problem of managing communications to various stakeholders across various channels, but it’s critical in today’s digital world and will demand their attention. As these teams become more specialized, they will also move from reporting to the CIO or CFO to whichever functional area lead is most relevant to the problem they are working to solve. The decentralization of data science teams throughout the organization will be critical for scaling up AI.
AI will start to be applied to content management.
The industry is experiencing an explosion of content. A shocking 78 percent of surveyed life sciences leaders report that their organizations produce “moderate to enormous” amounts of digital content, and nearly 60 percent say they spend more than $50 million on content each year; all expect this to continue to grow. Managing content efficiently is a big problem, but in same way that AI has enabled sales and marketing to navigate an increasingly complex space, AI can help.
In fact, AI can not only improve this process, but also bring back insights from physicians and even patients. As more organizations start to re-engineer their marketing materials as digital assets, they will naturally open the door to more data and insights. However, this adds further complexity – to the point that AI will be the only way life sciences companies will get control over their content assets, optimize content development, and leverage the digital breadcrumbs left behind by the doctors and patients interacting with these materials.
AI will enable in-the-moment adjustments to transform product launch execution.
Rapid learning and adaptation during the product launch phase is critical for pharmaceutical companies but historically lacking for many of even the largest organizations. Consider COVID-19 as an example where a new channel – the virtual visit – suddenly became the norm. The more agile an organization becomes, the faster it can adapt to changing on-the-ground conditions, and drive towards the most effective launch execution.
AI is the perfect accelerator at the central point where managing commercial execution across the complexity of time, channel, customer, region, and message becomes a data-rich but dimensionally intensive challenge. It has a short learning curve and will get even shorter in 2021. With COVID-19 driving most companies to go all-in on digital, accelerating the pace of change further, marketing, medical, and sales executives will lean on ever-improving AI technology to execute and adjust faster during the critical product launch cycle.
Life sciences’ commercial operations will lead the way in scaling AI.
AI is being explored in many different functional areas across life sciences organizations, but progress is slow and results are mixed. It is being trialed in drug development, clinical trial management, claims management, pharmacovigilance, and clinical diagnostics – which are all seeing a slow but steady evolution that will continue for years to come.
Commercial, however, is distinct because the analytics technologies used to parse customer data and the platforms that leverage it—like CRMs or marketing automation solutions—have evolved over the last decade and are now positioned to overcome the underlying scalability issue. This focus on building large-scale, repeatable processes and the incremental evolution of commercial AI has created a more mature platform that is now being tweaked to incorporate different blends of analytics technologies and business rules. Innovations in this area may not seem as compelling as a new algorithm that can detect cancer. However, if commercial teams can ensure the appropriate HCP is notified of the right therapy for a specific patient faster, the impact on patient care and operational efficiency would be dramatic.
In 2021, with this technical foundation in place, organizations will move on to tackle more complex problems and structure data in a way that can be incorporated into sales strategy, marketing models and other areas. This is the year we make the leap from hype to broad applicability.
Life sciences companies will learn to balance data and intuition.
One of the typical pitfalls in applied AI comes from relying excessively on a model, without applying common sense to the output. Anyone with deep experience in the space can tell you about a time when an algorithm produced a result that was wildly inaccurate for one reason or another. It is still essential to have a human ultimately filter the results and decisions reached by AI. If the output tells you to give 100 samples to a physician that normally receives just 10, human reasoning should be applied.
Technology providers must closely monitor outputs, and platforms should include a feedback loop for users so adjustments can be made continuously. As life sciences users grow more accustomed to interacting with AI solutions, reps or marketers can dismiss a suggestion, act on it, or ignore it now and then act on it later. Tools should have the ability to continue to learn and improve over time – which will create trust between user and technology, between man and machine.