In the past, pharmaceutical companies have invested heavily in Next Best Action (NBA) initiatives—rule engines, propensity models, dashboards, and AI-driven recommendations to improve field force effectiveness.
Yet the gap persists. Field teams still struggle with prioritisation, engagement sequencing, messaging, adapting to changing HCP behaviour, access constraints, and market dynamics. This is where a new paradigm, using Agentic AI, is beginning to reshape how the NBA is designed and executed in commercial pharma.
The Limits of Traditional NBA Approaches
Most current NBA implementations are built on static rules or a periodically refreshed model, single-objective optimisation (e.g. likelihood to prescribe), siloed decision logic across channels, and limited feedback loops from real-world execution
While these approaches generate recommendations, they often fail to answer the most important commercial questions, like Why this action? What should change if circumstances shift? How do we balance growth, access, experience, and compliance simultaneously?
What is Agentic AI in commercial pharma?
Agentic AI refers to goal-driven, autonomous AI systems that can:
- Interpret data from multiple sources, like CRM, Clinical documents, Sales, Population, HCP Profile, etc.
- Reason across business rules, objectives and constraints
- Decide on optimal actions
- Learn continuously from outcomes, field activity feedback, and notes.
When applied to NBA, Agentic AI shifts the focus to : “What should we do next, and why?”
This does not mean replacing human decision-making. Instead, it means creating an intelligent “decision companion” that works with commercial teams at scale to deliver high-quality output.
How Agentic AI Transforms Next Best Action
An Agentic NBA operates as a continuously improving system open to taking new rules and datasets.
At a high level, it:
- Understands context – HCP profile/behaviour, peer influence, patient dynamics, channel effectiveness, competition brands and organisational priorities.
- Reasons across objectives – sales impact, HCP experience, equity, compliance, and resource efficiency.
- Recommends and orchestrates actions – across field forces (MSL KAM, Rep), and channels
- Explains decisions – providing transparency and confidence to users with a why or with reference to the source of information.
- Learns continuously – from field feedback (CRM notes) and real-world outcomes
Use Case in Commercial Pharma
Here is a demo use case example we developed for the Kingdom of Saudi Arabia, and structured the outcome in three sections – Action, Support and Field tool.
The Saudi Arabian diabetes market represents a significant opportunity for SGLT-2
inhibitors like Empagliflozin, with growing prevalence rates (18% of adults) and a market projected to reach USD 3.4 billion by 2030. The analysis of ‘call notes’ and healthcare professional (HCP) data reveals specific opportunities to increase Empagliflozin adoption by targeting high-potential prescribers and leveraging the drug’s expanded indications beyond diabetes management. Agentic AI enables NBA scenarios that were previously difficult to operationalise:
Dr. Osamah Hamid, MD — King Fahd Specialist Hospital, Dammam
NBA (Treatment Protocol → Implementation Support):
Action: Request a focused 30-minute meeting to review his experience with Empagliflozin in his substantial diabetes practice (1,320 patients) and discuss the new Diabetic Centre protocol opportunities. Offer support for implementing an Empagliflozin pathway for CKD patients regardless of diabetes status based on EMPA-KIDNEY findings.
Support: Provide the Eastern Health Cluster with a Pharmacoeconomic analysis
demonstrating the cost-effectiveness of early SGLT-2 inhibitor initiation in reducing
hospitalisations and dialysis needs. Coordinate with the hospital pharmacist to ensure
consistent stock availability.
Field Tool: Share the newly developed SGLT-2 class effect comparison chart highlighting Empagliflozin’s differentiated renal outcomes and safety profile, particularly relevant for his diverse patient population
Security, Governance, Compliance, and Trust
Over the last few years, advances in technology have addressed security, governance, and compliance issues. In particular, the use of in-house data, such as CRM, sales, and patient-related data, is a greater concern for any Pharma company.
In the regulated commercial environments of the NBA:
- Humans remain in the loop
- Make sure all recommendations are explainable and auditable
- The technology platform selected is proven enough to maintain data security, possibly with an external audit.
- Guardrails are embedded to respect medical, legal, and compliance requirements
- Decisions are traceable back to data and reasoning paths, and, where necessary, ask for the reference.
This makes Agentic NBA not only more powerful but also more acceptable to commercial, IT, and compliance stakeholders.
Conclusion
The future of commercial analytics in pharma is not about more dashboards, models or reports. It’s about how quickly we can move from recommendation to implementation. And the answer lies in organisations’ willingness to embrace these Agentic solutions.