Using AI to Dig Deeper into Healthcare Strategy

If you're running strategy for a pharma brand, you're likely drowning in AI hype and AI fear mongering: it’s either going to unlock the answer to every question or take your job. And I’ll admit that I’ve fallen victim to some of that thinking, as well (thanks in part to my pal Harry Sharman, who I recommend if you want a more negative take.)

After a lot of experimentation and a bit of investment, I’ve come to believe that AI is primarily a force multiplier for people who want to ask difficult and interesting questions. As a healthcare strategist, I’ve had to accept that there are many hypotheses that I don’t have the time or resources to explore. AI allows me to chase down hunches, model unexpected solutions, and model competitive behavior. I’m not using it to replace or automate strategic thinking, but to process more information or generate more variations on a theme. At its best, AI gives strategists superhuman processing capacity to test more ideas, dig deeper into data, and explore strategic territory that would take months to cover manually. Below, I’m sharing five approaches you can implement tomorrow (especially with proprietary AI systems) that can help you do what you already do, but better.

1. Counter-Arguments: Stress-Testing Your Strategic Assumptions

Your best strategic ideas might be wrong. If they are, you'd rather find out now than after launch.

Strategic red teaming may have started as a way to test military strategy, but today it’s a critical business planning tool, and AI makes it dramatically more effective. AI can systematically explore multiple weak points in your strategy simultaneously, identifying vulnerabilities you hadn't considered.

The process: feed your strategic positioning, competitive assumptions, and market entry plans into a proprietary AI system and explicitly ask it to make the strongest possible case against your strategy. Not gentle feedback. The most devastating critique it can construct. Two of my favorite prompts are to ask it something like, “If we presented this strategy to Steve Jobs, how would he criticize it?” or “If Bill Bernbach was planning communications for [the leading competitor], how would he exploit weaknesses in our brand and our plan?”

This isn't about AI being smarter than you. It's about processing speed and exhaustiveness. Where a human red team might identify three or four major vulnerabilities in a day-long session, AI can systematically probe dozens of potential weaknesses in minutes. It can cross-reference your assumptions against competitor behaviors, market precedents, and scientific literature faster than any team could manually research.

What you'll get back is illuminating. The AI will identify logical flaws in your competitive assumptions, point out patient segments you've overlooked, and highlight scientific or market developments that could undermine your positioning. By asking AI to adopt an adversarial perspective, you surface unexpected vulnerabilities in benign scenarios, not just overtly malicious attacks.

The key is specificity. Don't ask "what's wrong with this strategy?" Instead: "Given our competitor's recent Phase 3 data and their established relationships with academic thought leaders, explain why our faster-follower approach will fail." The more pointed your prompt, the more useful the counter-argument. You're still the strategist deciding which critiques are valid (and a lot of them won’t be, so you get the added benefit of feeling like you aren’t being replaced) and how to respond. You've just compressed weeks of analysis into hours.

2. Scenario Planning: Building Probability-Weighted Models of Your Market's Future

Traditional scenario planning is constrained by human bandwidth. You can realistically war-game maybe three or four competitive scenarios before stakeholder patience runs out. That limitation forces you to pick the most obvious scenarios and thus limits the chance to identify a truly disruptive possibility.

AI eliminates that constraint. AI-powered competitive intelligence enables pharmaceutical companies to develop sophisticated predictions about competitor activities, including clinical trial outcomes, regulatory approval forecasts, and market entry timing estimates. For those focused on the marketing side of the business, you can generate dozens of scenarios of what the future competitive set might look like, and how those brands are likely to behave when they are competing with each other.

This is about computational horsepower, not replacing judgment. Start with your competitive set and known events: upcoming data readouts, major medical meetings, indication filings, patent expirations. Feed these into your AI system along with historical precedents and ask it to model outcomes from best-case to worst-case scenarios.

AI can rapidly iterate through dozens of permutations: What if Competitor A's data reads out positive but Competitor B files first? What if reimbursement takes six months longer than planned? It can model combinations and cascading effects that would take your team weeks to manually map. You can explore 30 scenarios in an afternoon, identify the three that actually matter, and then apply your team’s thinking to really understand what those scenarios mean for your brand.

The goal is to identify which early warning indicators should be monitored, which no-regrets moves make sense across multiple futures, and which contingent strategies should be prepared for rapid deployment. You're not predicting the future perfectly. You're building a comprehensive map of possible futures with probabilities attached, so you can recognize which one you're in as it unfolds. The AI gives you the processing power to be thorough. You bring the strategic insight to interpret what it means.

3. Tactical Planning: Automate the Predictable, Focus on the Novel

Every brand team faces this: 80% of your annual plan is evolutionary updates to last year's tactics, but it still consumes massive planning time. Meanwhile, truly innovative programs get squeezed.

Using natural language interfaces, marketers can instruct AI to identify relevant content modules, apply them to templates, and produce tactics for immediate review, dramatically reducing production time. Early implementations show potential to improve speed to market by 50% and increase content delivery volume by 25% to 40%.

This is purely about processing capacity. With a proprietary AI system, upload last year's tactical plan, recent market research, and updated guidance. Then ask for specific updates: "Revise the congress strategy based on the new NCCN guidelines," or "Update the KOL engagement plan given recent investigator meeting feedback."

The AI isn't thinking strategically. It's doing the mechanical work: cross-referencing your old plan against new information, identifying what needs updating, maintaining formatting and structure. Work that would take a junior strategist three days happens in three minutes. You review, refine, and approve. What you get back are draft updates that maintain strategic continuity while incorporating new information, requiring review and refinement but making the “block and tackle” part of planning much easier.

The real power isn't just speed. It's freeing you and your team to think through novel programs that traditional planning processes often sacrifice to deadline pressure. You're not getting worse tactics done faster. You're spending your brain power where it actually matters.

4. Conversation Modeling: Pre-Market Testing Without the Market

You've developed three brand narratives, but you aren’t confident you have a clear winner after market research is done. Do you role out materials based on what the research team is calling the winner? If you do, and the results aren’t what you hoped, you've spent months and significant budget on the wrong approach, and you have to decide whether to stick with something that is underperforming or go through the whole expensive and time consuming process again.

AI-powered sales training simulators use large language models custom-trained to mimic specific segments of healthcare providers and present common objections that reps would encounter. The same technology works in reverse: test message strategies before launch without burning real HCP relationships.

This is about iteration speed. Create AI agents representing different physician segments: the academic skeptic, the community prescriber focused on convenience, the cost-conscious VA provider. Run multiple versions of your message strategy through each persona and see what objections emerge.

These platforms have analyzed over 15 billion data points from human conversations, enabling them to directionally predict real-world responses. You can test 50 story variants in a few days, identify the most promising few, and move to live market research with much stronger hypotheses.

The AI isn't making creative decisions. You're still choosing the messages and deciding which responses indicate success. But you're testing them against synthesized physician perspectives at a speed that would be impossible with human role-play. You're not replacing qualitative research. You're making it exponentially more efficient by strengthening your input before you commit research budget.

5. Insight Mining: Turn Research Archives into Strategic Assets

Most pharma brands sit on years of market research. All of it cost serious money. Almost none of it is systematically revisited because manually re-analyzing old reports (let alone the dozens of transcripts and surveys that they are based on) would take weeks.

AI is now facilitating insight mining of past research to inform future marketing decisions more efficiently, with generative AI able to summarize interview transcripts in seconds, extract quotes, conduct text analytics, and generate summary insights. Novartis used generative AI to analyze HCP interview transcripts, and reported that they cut data ingestion time by 600% and time-to-insights by 90%.

This is pure information processing. Upload your research transcripts into a proprietary AI system with an ambitious prompt: "Based on these transcripts, identify 50 potential strategic insights, with supporting quotes." Not five insights. Fifty.

This forces the AI to surface non-obvious patterns: themes that appeared in multiple interviews but were never explicitly connected, contradictions between physician segments, gaps between what your research explored and what it revealed. It can process and cross-reference thousands of pages of transcripts in minutes. You bring the strategic lens to evaluate which patterns actually matter.

Then get specific: "Based on this research, what communications tactics would likely be most effective with community oncologists?" Or: "Compare interviews from 2023 with this recent research. How have physician attitudes shifted?"

The AI gives you pattern recognition at scale that no human could manually achieve. The real power emerges with longitudinal analysis. You can identify how the competitive landscape has evolved, which physician concerns have intensified, and where your previous strategy assumptions may no longer hold. You're still doing the strategic thinking, but you’re being given leads to pursue rather than digging through raw data and hoping you find the pattern buried within. It’s the difference between working with the complete picture versus the fragments you could manually process.

The Optimist’s Case

All these use cases are united by a philosophy: AI shouldn't and can’t replace strategic thinking and human discernment. It should give strategists the processing power to think more deeply and test more rigorously than humanly possible. Challenge your assumptions across dozens of parameters. Create multiple models of your future market. Automate tactical updates to spend more time on big ideas. Test message variants at scale before you talk to customers. Find the patterns in old research that you otherwise wouldn’t have time to revisit.

It is a pessimistic time in our industry: a lot of talented marketers are struggling to find work, and no one I talk to feels very confident they know what their job will look like in a few years even if they have one. But the pharma industry has always been about exploring the cutting edge of science, and we should use that spirit and leverage this amazing technology in ways that make us better.