In conversations about AI, industry leaders say they’re mostly intrigued by the technology’s potential and how AI is evolving to meet their needs. But questions and cautions abound, and many pharma innovators are feeling analysis paralysis. Should they forge into unknown territory or wait and see? Will the answer be precision AI-powered tools, or will it be mega-projects years in the making?
At this inflection point, we thought it would be helpful to take a step back, look at the bigger picture, and answer one key question:
How did we get here?
Understanding the current landscape
Pharma is faced with ever-accumulating amounts of data. On the surface, this might seem like a good thing. Still, the life science industry is divided on the matter: research on data for medical affairs shows that while about 40% of medical affairs leaders say they have too much data, about the same number say they don’t have enough.
The issue causes problems downstream, too. Pharma teams must analyze more information, exacerbating demands on individuals who have other critical work. Reporting also takes time – and these intensely manual processes mean reporting happens less often than is ideal. Worse, when reporting does occur, it may be too late to take effective action.
AI offers a potential solution, but pharmaceutical companies still struggle with how to apply it most effectively. Some companies still forbid its use, others insist they will develop their own solutions, and a third group is experimenting with AI for tasks like summarizing lengthy transcripts or examining the sentiment of conversations and other unstructured data.
But even those leading the charge to harness the power of AI face obstacles, including cost and resource constraints that limit large-scale investments, confusion and fear around how AI could compromise information security, and organizational structures and processes that may not support agile experimentation and rapid iteration.
Amid these challenges, it’s useful to turn back the clock and look at the advancements that led us to this moment – and perhaps understand where AI in pharma is headed.
2015: the pre-AI generation
Many pharma teams were on a level playing field not so long ago. The discipline of insights management was nascent, and organizations relied on one-to-one meetings via field medical teams or in-person advisory boards to gather information. Analysis and reporting were mostly manual, and data management was done using spreadsheets and costly data integrations. Related: look back at how we saw the future of insights management.
“This is the idea [of taking] field data, put it into spreadsheets, and comb through them,” says Within3 CEO Lance Hill. The problem with this approach? “It’s slow. You can’t keep up with the data volume for any sort of large organization.”
2016-2018: dashboard AI
As technology evolved, business intelligence tools tackled the problem of having too much data. The resulting dashboards helped pharma companies see their data more visually, and natural language processing enabled faster trend analysis.
While these capabilities are still used today, they require considerable data integration investment and require humans to spend time configuring, analyzing, and communicating the answers they glean from the dashboards.
In other words, they stop short of strategically suggesting where to go next. “This approach…doesn’t really encompass the whole process,” explains Hill. “It doesn’t automate or improve any of the gathering of data and observations. And it certainly doesn’t help you improve communication or impact.”
2019-today: generative AI and integrated solutions
In 2024, we understand much more about how generative AI impacts insights generation. Now, AI does the analysis that humans were previously tasked with, and humans can interact with AI in plain language. “It eliminates the heavy lifting so organizations can get value faster,” says Hill.
Pharma teams get direct answers and recommended actions versus only views of data. This is the fastest and most impactful approach yet, and it’s no wonder the industry is eager to leverage it.
In the real world
The most effective way to understand how AI can impact your insights management process is to start with a pilot project, evaluate success, and scale up. In one of our client examples, a newly appointed head of medical affairs was charged with understanding why adoption of the company’s flagship product had been slower than expected and creating a strategy to build awareness and usage.
The head of medical affairs assessed the situation and determined the following:
- Manual reporting processes were cumbersome, resource-intensive, and potentially subject to human bias
- Reporting format varied, with no ability to see cumulative data or track quantitative trends
- Medical affairs was not consistently viewed as impacting the wider organization
By implementing an AI-supported insights management process, the team could consistently produce reports that showed trends over time. Key insights and recommendations for action enabled the team to make faster decisions and identify opportunities to increase their product’s awareness among prescribing physicians by more than 20%.
Looking ahead: trends and opportunities
AI addresses key challenges faced by the pharmaceutical industry, including the pace of data accumulation, rapidly changing technology, data silos, and the complexity of data analysis in large global organizations. As companies decide how to move forward with AI, it’s important not to lose sight of the ground gained by starting small, achieving results, and applying lessons learned.
Understand more about how medical affairs leaders use AI in an on-demand webinar with Impatient Health.
