SPEAKER:
Craig Ackerman, Partner, Alexander Group
KEY TAKEAWAYS:
- Nearly 80% of pharma companies call AI a revenue driver, but most deployments target internal workflows
- Autonomous AI outreach converted 4% of previously uncovered prescribers into active script writers
- Only 21% of companies are satisfied with incentive comp; that cohort outperforms on every measured indicator
- Quota design is the primary driver of commercial employee turnover, not culture or management
- Pilot architecture predicts adoption outcome: skunkworks experiments fail; scale-ready designs succeed
Every pharma commercial organization carries a long tail of prescribers who write enough scripts to register in the data but not enough to justify dedicated field coverage. The economics have never worked. A rep visit costs real money; a low-volume prescriber's incremental volume rarely justifies it. So the tail gets ignored, or it gets a quarterly email it doesn't open, or it gets lumped into a territory so large that meaningful engagement is structurally impossible. This isn't a new problem. It's a permanent feature of field-force-centric commercial models that the industry has accepted as unavoidable for decades.
Craig Ackerman, Partner at Alexander Group, came to argue that the assumption of unavoidability no longer holds, and that the industry's current AI investment pattern is systematically missing the opportunity that makes it false.
Pharma Is Investing in the Wrong AI Outcome
Alexander Group survey data shows that nearly 80% of pharma companies describe AI as a revenue driver. Ackerman acknowledged this with some irony, noting he "couldn't figure out" how AI was going to decrease anyone's revenue, which made the near-unanimity feel less like conviction than reflex. What the data doesn't show is a corresponding concentration of AI investment in external commercial applications. The deployments getting funded are internal: workflow automation, content generation, rep productivity tools, forecasting support. The customer-facing use cases, the ones that change how physicians actually experience the commercial organizations reaching them, remain largely theoretical in most portfolios.
"It's not just about efficiency," Ackerman argued. "That's us. That's our internal perspective. Our external perspective is this can be a massive customer experience lever that we can pull." That framing deserves more strategic weight than most organizations are giving it. The gap between where AI investment is concentrated and where its most consequential commercial value lies is the structural problem worth diagnosing.
The Long Tail Has Always Been the Business Case
What Ackerman's case study demonstrates is that AI changes the economics of the long tail entirely. "AI is not only about efficiency," he said. "It helps improve our customer experience because we can serve up better content to our customers...this is going to be a huge customer experience transformation." The case study deployed autonomous AI to engage a population of uncovered prescribers, physicians receiving essentially zero commercial attention, with targeted outreach designed to surface genuine interest before routing warm contacts to human reps.
The results were specific enough to model. "Of those customers that were reached out to, 8% were engaged. Doesn't sound like a lot, but when you consider that long tail of customers that most of us have, it's a lot of customers. And then of the 8%, half of them started writing scripts." Run that math against a realistic long-tail population and it converts quickly into a material revenue argument. Four percent of a previously uncovered segment generating new scripts represents revenue that simply didn't exist in the prior model.
Two things make this more than an interesting data point. First, the economics are structurally different from field-force coverage: autonomous AI engagement operates at near-zero marginal cost per contact, which means the long tail becomes commercially viable for the first time. Second, this is not an efficiency story. No headcount was reduced. No process was streamlined. A segment of the market that was generating zero revenue began generating revenue because the company changed how it engaged customers. That's a commercial model transformation, not an optimization.
The architecture matters: AI handles first contact and interest qualification; a human rep handles the relationship once engagement is confirmed. The destination isn't automated selling. It's AI-enabled human selling at a coverage scale that field economics alone could never support, deployed in service of a customer experience that previously didn't exist for that physician population.
Discover more on this topic at Pharma Customer Engagement USA 2026 (October 27-28, Philadelphia) - where commercial, marketing, medical, data and AI pioneers converge. Explore the agenda here.
The CRM Pattern Is Repeating Itself
The technology risk in AI adoption isn't the technology. It's the organization.
"I actually talked to a company this week in this industry, a pharma company. Their commercial leaders, the CEO and CFO, don't want to allow the company to use AI. They're too worried about compliance. I say 'Well, if you're worried about compliance now, you're probably going to have a whole host of other issues down the road.'"
Compliance concern is legitimate. Pharma's regulatory environment is genuinely complex, and AI-generated content carries real promotional risk. But organizations that respond to that complexity with categorical prohibition are making a specific strategic choice: they are ceding the period during which AI deployment competencies are being built, data infrastructure is being developed, and commercial teams are developing the workflows that will determine which organizations lead the next competitive cycle. The compliance risk they're managing today is bounded. The competitive risk they're accumulating is not.
The historical parallel Ackerman invokes should be uncomfortable for anyone who has watched pharma's relationship with CRM platforms. "The only people who made any money off of installing Siebel, because they were really the first, were the consultants," he observed. "So we don't want AI to be the same. So we want to make sure we can adopt." The pharma industry has a documented pattern of investing in commercial technology platforms and systematically underinvesting in the organizational change required to realize their value. CRM implementations became data graveyards. Multi-channel engagement platforms never achieved the integration their architects designed. The technology worked. The adoption didn't.
AI is a higher-stakes version of the same challenge, and the early deployment patterns suggest the industry is making the same category error. The corrective is architectural. "Build the pilot like you're going to scale this. Don't build a pilot as a skunkworks side project, because if you do that, it will fail. If you build it to scale, or build the pilot at scale, much higher degree of success." A pilot running in an isolated sandbox with no defined integration pathway to commercial operations, no change management plan, and no executive accountability structure isn't a pilot. It's an alibi, one that allows an organization to claim AI investment while insulating the core commercial model from the disruption that investment is supposed to create.
The 21% Benchmark Reveals Structural Misalignment
Alexander Group's incentive compensation data produces a finding that reads, on the surface, like a satisfaction problem. Ninety percent of pharma companies change their comp plans every year. Only 21% report being satisfied with the result. "Only 21% of companies or people are happy with their incentive compensation plan," Ackerman noted. "The 21-percenters, they outperform in all key indicators."
The chronic redesign cycle is the tell. Organizations changing their comp plans annually aren't engaged in strategic refinement; they're in a continuous correction loop, adjusting mechanics because the underlying architecture isn't producing the behavior the commercial model requires. That's a structural misalignment problem wearing a compensation problem's clothes.
This has direct implications for AI adoption that extend beyond what Ackerman explicitly draws out. If a commercial organization cannot design quotas and incentive structures that align rep behavior to current strategic objectives, layering AI tools onto that misaligned system produces noise, not performance. AI-generated insights about prescriber behavior are only actionable if reps are incentivized to act on them in the ways the model anticipates. AI-optimized call routing only improves outcomes if the incentive structure rewards the behaviors the routing recommends. Incentive comp alignment is a precondition for AI-augmented commercial model performance, and the 21% benchmark is a quick read on whether an organization has met it.
The turnover dimension makes the cost of misalignment concrete. "If you have an employee turnover problem in your commercial organization, look directly at your quotas. That is the number one driver of employee turnover, because they don't make their target income, so they'll go somewhere else where they get paid." Organizations treating retention as a culture or management problem while leaving quota design unexamined are addressing a downstream symptom.
Lifecycle Stage Determines What AI Can Do for You Now
"Even in today's day and age of more complexity, half the commercial resources are still sales reps, still traditional sales reps," Ackerman observed, with genuine surprise at the channel's durability. Decades of diversification, digital investment, and commercial model evolution have not dislodged the human rep as the dominant resource in pharma commercial organizations. That's not evidence of stagnation. It's evidence that the channel works, and that physicians' preference for it is durable.
Ackerman grounds this in behavioral constants rather than economics or regulation. The industry won't see the disappearance of sales reps or other human commercial resources, he argued, "because why? We're humans. We want human-to-human interaction." The hybrid model, AI enabling and amplifying human commercial resources rather than displacing them, isn't a transitional compromise on the way to full automation. It's the architecturally stable destination, because the physician preference that makes human reps effective isn't going away.
What this means strategically is that first-mover advantage in AI-enabled commercial models accrues to organizations that build the human-AI architecture, not to those that automate most aggressively. The long-tail case study is instructive precisely because it preserves the human relationship at the point of highest value, the confirmed-interest conversation, while using AI to make that conversation possible at a scale the field force alone could never reach.
There's a sequencing implication the industry hasn't fully confronted. A company that has already built role specialization, integrated data infrastructure, and multi-channel engagement capabilities can deploy AI against that foundation and amplify it. A company still working through basic commercial model architecture, territory design, role definition, channel integration, cannot skip to AI-augmented selling by purchasing the right platform. The commercial foundation has to exist before AI can build on it. Lifecycle stage doesn't just describe where an organization is; it prescribes what AI can actually do for it right now, and what prerequisites have to be addressed first.
The compliance-cautious, the pilot-tinkering, and the efficiency-obsessed are all, in different ways, deferring the same decision. And the window for first-mover advantage in AI-enabled customer experience closes in exactly the way first-mover advantages always do—when enough competitors have built what you delayed building.
"You have to start investing in AI now. If you wait until it's figured out, you're already behind."
To get you highlights of Pharma USA 2026 faster, we are using generative AI technology to summarise the transcripts of the sessions. If you have any feedback about the summary, please contact lucy.fisher@thomsonreuters.com.
Discover more on this topic at Pharma Customer Engagement USA 2026 (October 27-28, Philadelphia) - where commercial, marketing, medical, data and AI pioneers converge. Explore the agenda here.
