AI is transforming how the pharma industry approaches drug development. On average, the industry invests $2.6 billion to bring a new treatment to market, taking 12 to 15 years with a 10% success rate. But pharma pioneers are leveraging AI to rethink traditional models. Sanofi, for example, is developing a “lab-in-a-loop” that uses AI agents to potentially compress elements of discovery work from years to weeks.
One implication of this progress is that pharma companies are using AI to unlock potential treatments for smaller and smaller patient populations, as in for rarer diseases and conditions that traditionally would not be a priority. Indeed, over the past decade or so, the percentage of FDA drug approvals that address rare conditions has risen from below 30% to above 50%.
This represents amazing progress and tremendous hope for the millions of Americans who suffer with rare conditions.
But this progress also raises a unique challenge for pharma companies. The smaller the patient population for a new treatment, the harder it becomes to reach that audience.
Traditional marketing has been designed, for the most part, to reach the largest audience for the lowest-possible cost, which, in many cases, makes economic sense. The logic works if you’re selling pizza or toothpaste. But the model breaks down if a drug treatment is relevant to an audience of just thousands rather than millions.
In pharma, increasingly, success will not be measured by reaching a mass audience. It will be about raising awareness among the right combination of patients and their health care providers –think of them as the original influencers – at the right moment in a patient’s journey and in the most appropriate environments – all while preserving privacy.
For example, when Ultragenyx launched Crysvita in 2013, it was the first treatment approved for X-linked hypophosphatemia, a very rare bone condition. The medicine addresses a significant unmet need and is highly effective. Initially, however, the treatment struggled in market, because the relevant patients and their health care providers were extremely hard to find. And so the company pivoted, focusing on disease education, free gene testing and payer assistance, helping successfully sustain the treatment in market with a more precision-based approach.
In other words, the challenge wasn’t whether the drug worked. It was the cost and complexity of identifying and reaching a small population of eligible patients and providers in the most effective way possible.
This example also highlights the core challenge that pharma marketers will increasingly face as AI speeds up and optimizes drug development: Precision medicine requires a precise approach to commercialization.
But just as AI is transforming drug development, it also has the potential to transform how we bring these treatments to market.
Programmatic advertising was an important first step. It gave marketers the ability to process enormous quantities of data and make decisions about where, when and to whom an advertisement should be delivered. But programmatic systems largely optimize existing advertising campaign execution.
The real opportunity is to apply AI to marketing and commercial decision-making itself rather than just to execution systems.
Agentic AI can synthesize vast amounts of health care, audience and market intelligence and turn it into actionable marketing and commercial insight. This information often sits in silos across teams and agency partners. Instead of teams spending weeks assembling data, analyzing spreadsheets and debating which assumptions to put into a forecast, AI agents can help marketers explore scenarios, identify opportunities, recommend strategies and continuously adapt as market conditions change.
And all of this can be done in simple, conversational language that requires little technical expertise.
Imagine being able to ask in plain language: Which patient populations are most likely to benefit from this treatment? What strategies would be most effective in reaching them? Where have comparable treatments succeeded or failed and why? Which health care providers are most likely to treat these patients? What market access dynamics could have an impact on adoption? Which engagement strategies are most likely to work and how should they evolve over time?
Rather than requiring teams to manually assemble answers from complex and disconnected systems, agentic AI can bring those signals together in easy-to-use interfaces, make recommendations and activate based on those recommendations. In this way, agentic AI is very much a copilot to human expertise, but it’s able to make sense of masses of disparate data that no human alone could comprehend.
The data underlying pharma marketing and commercialization is incredibly complex. It includes millions of signals and variables: claims, payer dynamics, sales representative activity, market events, audience behavior, health care data and media performance. But agentic AI can make sense of that complexity in real time, turning something that once seemed overwhelming into a very powerful and intuitive tool to bring and sustain treatments in market successfully.
In early tests using agentic AI tools, we have seen marketing planning and execution cycles for new treatments collapse by as much as 10x. These results are early, but they point toward something much bigger than greater marketing efficiency.
There are more than 10,000 known rare diseases, yet the vast majority still have no approved treatment. The scientific barriers to treating these diseases are beginning to fall, in some part due to advances in AI. But that’s not enough. For a treatment to succeed, it must reach the people who need it in a way that supports current and future treatment investments.
Agentic AI has the potential to dramatically reduce the cost and complexity of forecasting patient need and planning in order to execute successful marketing campaigns, reach the right audiences and continuously optimize engagement in a way that can make a great number of treatments more commercially viable.
But it’s not just about rare diseases and small biotech pioneers. Agentic AI can help pharma companies of all sizes launch and sustain treatments in market more successfully and effectively.
And it won’t just be pharma companies that benefit. If agentic systems can streamline marketing and commercial processes for treatments, we can commercialize them far more efficiently and improve treatment economics. All of which will drive more investment in treatments for conditions that have historically been too challenging or expensive to address.
If we can do that, we can help improve the quality of life for millions of Americans.
For more articles featuring Ian Colley, click here.
