How to Validate a Synthetic Audience

Here is the test. Give the model a cohort whose outcome you already know, hide the outcome, ask it to predict, then publish how wrong it was. Repeat every quarter, because the model underneath you changes without telling you.

That is the whole protocol. Everything below is how to run it without being talked out of it.

I am not arguing against synthetic audiences. Hold one to the standard you would apply to any other instrument claiming to tell you about your customer. Almost nobody in this market applies that standard today.

What are you actually validating?

Not whether the output sounds like a physician. Whether it predicts what a physician does.

Those two get conflated constantly, and the conflation is where your money goes. A vendor demonstration shows you fluency. The simulated prescriber speaks in the right register, cites the right guideline, raises the objection you expected. It reads like insight.

Fluency is the easy part. Any competent model clears it. What you are buying is predictive value, and nothing in a demonstration tells you whether you have any.

Three questions separate the two.

Does it get the direction right, meaning does it rank options the way your market actually ranked them. Does it get the magnitude roughly right, meaning is the gap between options approximately the real gap. Does it get the surprises, meaning does it ever tell you something you did not already believe.

That third one matters more than it looks. A model that confirms your existing view every time is not validated. It is agreeable.

The surprises are where the money sits, and they tend to live in friction nobody articulates. Gartner found 62% of customer service channel transitions are high effort in industries that can watch the transition happen. A simulated customer never reports being worn down, because that is something people quit over rather than describe.

What does a proper validation look like?

Six steps, and you can run the whole thing in a quarter.

One. Pick a cohort with a known outcome. A brand launch you can see in full, a market where you have twelve months of behavior, an indication expansion that has already played out. You need truth that exists and that the model cannot have seen.

Two. Blind it properly. The vendor gets no outcome, no time period and nothing identifying the cohort, because a vendor already holding your historic data will flatter its own error rate.

Three. Write the prediction down before you look. In advance, in a document, with the specific numbers you expect the simulation to produce. This is the step everybody skips and it is the only thing standing between you and hindsight.

Four. Score it against behavior, not sentiment. Script to start conversion, time to therapy, persistence at six and twelve months, barrier resolution speed. Those are the measures set out in how to measure customer experience in pharma, and they are the measures a simulation should be answerable to.

Five. Publish the error rate internally. One number, visible and owned, rather than a fidelity score the vendor computed on its own work.

Six. Repeat quarterly. Models get retrained, replaced and tuned underneath you, and nobody sends a memo. An accuracy check at purchase describes a model that may no longer exist.

Which outcomes should you score against?

The ones where real people already voted with effort, money or time.

Pick your validation targets from behavior that cost somebody something. A prescription written, a prior authorization appealed or abandoned, a refill collected or not. Each of those has a price attached, which is what makes it worth predicting.

The appeal is a good place to start, because the burden is documented. In the American Medical Association's latest survey, 82% of physicians said prior authorization at least sometimes leads patients to abandon treatment. Ask the model which of your accounts will give up, then go and look.

There is plenty of this truth available. A 2026 JAMA study summarized by Johns Hopkins found insurer rejections reached 40.7% of initial brand name attempts in 2024. Of those rejected scripts, 48.4% were never followed by a fill of that drug or anything in its class within 90 days. Ask your vendor to predict a number in that family and then check it.

Be careful about how wide real variance is before you accept a tidy answer. An earlier review in the American Journal of Pharmacy Benefits reported that findings on primary nonadherence "range from as little as 2% of new prescriptions going unfilled to as many as 30%". A simulation that hands you a confident point estimate inside a range that wide has told you nothing, however precise the decimal looks.

What counts as good enough?

Better than the thing it replaces, measured without flattery. That is a lower bar than vendors claim and a higher one than most procurement applies.

Work out what the alternative actually is for the decision in front of you. Three cases, three different bars.

If the alternative is one brand manager's assumption formed in a meeting, your bar is low. Simulation beats assumption, and you should use it. Say so in the document.

If the alternative is real research you could have commissioned, your bar is high. The simulation has to get close enough that the difference does not change the decision, and you should expect it to fail that test on anything subtle.

If the alternative is a number you are going to report upward, there is no bar. A simulated output does not become a reported figure, ever, at any accuracy level. Use it to decide what to measure, never as the measure.

Why will your vendor resist this?

Because a published error rate is the one artifact that makes the category comparable, and nobody selling into an unmeasured market wants a measurement.

Expect four objections. Each has an answer.

"Our fidelity score is already 90 something." A score the vendor defines, computes and reports on its own work is marketing. Ask what it is a percentage of, and watch what happens.

"Synthetic research is directional, not predictive." That is a reasonable position. Hold them to it, then ask why the pricing, the dashboards and the sales deck are all built as though it were predictive.

"You cannot validate on patients." Correct, and that is the finding. It does not excuse the claim, it limits where the tool belongs. I have set out why the patient audience is the weakest ground in synthetic personas in pharma.

"Nobody else asks for this." True today. You are not buying what everybody else is buying, you are buying something that has to work.

A vendor who runs this test with you and shows you an unflattering number has told you more about their product than any demonstration could. Buy from that one.

What if validation is actually impossible?

Then you label it, bound it, and keep the real listening funded.

Patients are the hard case. In most markets you cannot speak to them, so there is no real voice available to score the simulated one against. The audience you most want to simulate is the one you can least check.

Notice why the pressure to skip the test keeps rising. Veeva Pulse data reported by BioSpace put HCP accessibility at 45%, down from 60% eighteen months earlier. As real access closes, simulated access gets easier to sell and harder to check.

Three rules hold in that situation. Any chart built on simulated respondents carries that fact in the title rather than a footnote. No unvalidated simulation touches a decision that cannot be reversed. Your spending on real listening does not fall because simulation got cheaper, which is the failure mode this whole category creates.

There is a cheaper source of real patient signal you already own, and almost nobody collects it. Your nurses, hub agents and patient support staff hear from patients daily, and that route is set out in voice of the frontline.

Who should own the revalidation?

Whoever reports the number, not whoever bought the tool.

Put the error rate in the same hands as the commercial scorecard. A research team that owns the vendor relationship has an interest in the tool looking good, which is not a character flaw, it is how incentives work.

Watch what happens when nobody independent holds the measure. Deloitte's 2025 research found that only 28% of HCPs believe pharma's engagement strategies meet their needs, against 82% of life sciences executives who say they are satisfied. Your industry is already capable of believing a comfortable number for years. A confident simulation makes that easier, not harder.

The reason a model cannot reach what actually drives the decision is in Predisposition. The three routes to real signal are in voice of the customer in pharma, and the argument all of it serves is the future of the pharma commercial model.

Validate it and simulation becomes a real instrument in your hands. Skip the test and you have bought a very articulate way of agreeing with yourself.

Key takeaways

Validation means a blind prediction against a cohort whose outcome you already know, with the error rate published and repeated quarterly.
You are validating predictive value, not fluency. Any competent model clears fluency and a demonstration tells you nothing.
Score against behavior that cost somebody something: conversion, time to therapy, persistence, barrier resolution speed.
A vendor defined fidelity score is marketing. An error rate against your own realized outcomes is evidence.
Where validation is impossible, as with patients, label the output in the chart title and never let it become a reported number.

Questions to ask before you sign

Which cohort will we validate against, and can you prove you have never seen its outcome?
What exactly is your fidelity score a percentage of?
Show me one engagement where the validation came back unflattering, and what you did about it.
Who inside my company will own the error rate, and will they be independent of this contract?
What decisions would you tell me not to make on your output?

About the author

Wayne Simmons is the founder of The Customer Excellence AGENCY and the author of The Customer Excellence Enterprise (Wiley, 2024). He is founding faculty of the MS in Customer Experience Management at Michigan State University's Broad College of Business. He led global customer excellence in Pfizer's first Chief Marketing Organization and in Bayer's Customer Powerhouse. Related reading: What is a synthetic persona?, Voice of the customer in pharma, three ways to hear and What is Predisposition?

< Older Post

Newer Post >

[PRACTICE NOTE] Your Call Plan Isn’t the Problem. Your Customer Context Is.

March 15, 2026

Why healthcare professionals now judge pharmaceutical engagement against the best experiences in their lives, and what that means for the future of commercial leadership. When commercial performance falters, the reflex inside many pharmaceutical organizations is to adjust the machinery of field execution. Leaders revisit call plans, recalibrate targeting models, and increase the volume of activity in the hope that more precision or more frequency will restore momentum. For decades this system has been treated as the central instrument of commercial performance, determining which physicians are prioritized, how frequently representatives engage, and how resources are deployed across territories. Yet the growing gap between commercial effort and commercial impact suggests a deeper issue. T he problem is rarely the design of the call plan itself. It is the context in which healthcare professionals now operate. Physicians are navigating increasingly complex clinical, administrative, and informational environments, and that evolving reality now shapes prescribing behavior far more than the cadence of promotional interactions. What many organizations are experiencing is a widening Customer Context Gap. Commercial systems were designed for a time when prescribing decisions could be influenced primarily through promotional interaction and product information. Today physicians operate inside a far more complex reality shaped by administrative burden, reimbursement constraints, digital information overload, and growing expectations for seamless support across the entire care journey. In this environment the physician’s decision is influenced not only by clinical evidence but also by how easily a therapy fits into the practical realities of care delivery. When commercial models remain anchored in promotional activity while the customer’s context has fundamentally changed, even the most disciplined call plan struggles to deliver the outcomes it was designed to produce. Closing this gap requires a different way of thinking about commercial performance. The question is no longer how to optimize promotional activity but how to align the organization around the real journeys through which physicians help patients receive therapy. Prescribing decisions unfold within complex sequences of clinical evaluation, reimbursement navigation, patient readiness, and ongoing support. When commercial strategy is designed around these journeys rather than isolated interactions, the role of the field force begins to evolve. Representatives are no longer positioned primarily as messengers of information but as partners in removing barriers that slow care. Organizations that recognize this shift begin redesigning their commercial systems accordingly, aligning field engagement, digital support, access programs, and patient services around the same goal: helping healthcare professionals help patients move from clinical intent to successful treatment. From Promotional Activity to Customer Journeys The pharmaceutical industry has historically organized commercial activity around the moment of promotion. Call plans, targeting models, and message sequencing were designed to influence prescribing behavior primarily through informational engagement with healthcare professionals. While this model brought structure and scale to commercial operations, it reflects an earlier era in which the path from clinical awareness to prescribing action was comparatively linear. Today the journey is far more complex. Physicians must navigate an intricate landscape of clinical evidence, treatment guidelines, payer requirements, prior authorization processes, patient affordability concerns, and adherence challenges. Prescribing a therapy is no longer a single decision point. It is the beginning of a chain of events that determines whether a patient ultimately receives and remains on treatment. This is why the commercial conversation must expand beyond the traditional moment of prescription to encompass three interconnected journeys. The first is the Path-to-Prescribe , where scientific evidence, clinical education, and confidence in the therapy shape the physician’s willingness to recommend treatment. The second is the Path-to-Fulfill, where access, affordability, patient readiness, and operational support determine whether that recommendation ultimately becomes therapy in the patient’s hands. The third is the Path-to-Adhere , where ongoing patient support, monitoring, and engagement determine whether patients remain on therapy long enough to realize the intended clinical benefit. Science drives the Path to Prescribe, where evidence, clinical education, and confidence in the therapy shape the physician’s willingness to recommend treatment. Experience shapes the Path to Fulfill, where access, affordability, and patient readiness determine whether that recommendation becomes therapy in the patient’s hands. Sustained outcomes depend on the Path to Adhere, where ongoing support, monitoring, and engagement ensure patients remain on therapy long enough to realize its intended clinical benefit. When commercial organizations focus almost exclusively on the first while leaving the latter journeys fragmented and burdened, a significant portion of therapeutic value is lost between intention and impact. In many therapeutic areas, the result appears in the persistent gap between prescriptions written, prescriptions filled, and therapies sustained—gaps that reflect not a failure of science but a failure of system design. Recognizing these three journeys shifts the unit of focus from promotional activity to the real-world pathways through which care is delivered. It reframes the role of the field force, the purpose of digital engagement, and the design of patient support programs around a single objective: reducing the friction that stands between clinical intent, treatment initiation, and sustained patient outcomes. Customer Context Is the New Commercial Variable For much of the pharmaceutical industry’s history, commercial performance was largely explained by a familiar set of variables. Product efficacy, clinical differentiation, promotional reach, and sales force execution determined the trajectory of most brands. When performance lagged, leaders adjusted those levers by refining segmentation, optimizing targeting, and recalibrating call plans. Today those traditional levers still matter, but they no longer explain commercial outcomes on their own. A far more powerful variable has entered the equation: customer context. HCPs now operate within an environment defined not only by clinical complexity and administrative burden but also by rising expectations shaped by their experiences outside healthcare. Physicians are also consumers. In their personal lives they interact daily with companies such as Apple, Amazon, Tesla, and Netflix that anticipate their needs, remove friction, and simplify complex processes through thoughtful design. These experiences quietly reset the benchmark for competence, responsiveness, and respect for their time. When those same physicians step into their clinical roles, they do not shed those expectations. They carry them with them. The contrast between the seamless orchestration of their consumer experiences and the fragmented systems surrounding many healthcare interactions becomes difficult to ignore. What once felt acceptable now feels unnecessarily burdensome. This dynamic represents the Consumer-Grade Imperative. Healthcare professionals increasingly evaluate pharmaceutical engagement not against other pharmaceutical companies but against the best experiences they encounter anywhere in their lives. In this environment even a clinically superior therapy can struggle if the surrounding system makes it difficult to initiate treatment, navigate reimbursement, or support patient adherence. Customer context therefore becomes the new commercial variable. It determines whether scientific differentiation translates into practical adoption. It shapes whether prescribing intent becomes therapy initiation and whether therapy initiation becomes sustained patient outcomes. Call plans were designed to manage activity. Customer context requires organizations to manage journeys. The Field Force in the Era of Customer Context Recognizing customer context as the defining commercial variable inevitably reshapes how the role of the field force is understood. For decades the pharmaceutical sales representative has been positioned primarily as the carrier of scientific information. Call plans optimized the frequency and sequencing of these interactions to ensure that physicians received consistent messaging. That role does not disappear, but the environment surrounding it has changed profoundly. Physicians today are navigating administrative burden, payer complexity, digital information overload, and increasing time pressure. In this environment they are not simply seeking more information. They are seeking clarity, simplicity, and support that helps them navigate the complexity surrounding treatment decisions. This shift transforms the representative from a messenger of information into something far more valuable: a partner in removing friction from the care journey. Conversations move beyond repeating clinical claims toward understanding the practical barriers that physicians and their teams face as they attempt to initiate and sustain therapy for patients. The most effective field forces are therefore supported by commercial systems designed around journeys rather than activities. Representatives are equipped not only with scientific messaging but with the insight and coordination required to address obstacles across prescribing, reimbursement, and patient support. Field engagement becomes a catalyst for problem solving rather than simply a vehicle for promotion. From Call Plans to Customer-Aligned Commercial Systems If customer context has become the defining commercial variable, then the systems designed to support the field must evolve accordingly. The traditional call plan was built to manage activity. It provided structure for how frequently physicians were engaged, how territories were covered, and how resources were deployed. Yet activity alone does not determine whether therapies ultimately reach patients. What determines impact is whether the commercial system surrounding the physician reduces or increases the burden of delivering care. A customer-aligned commercial system begins with the journeys through which physicians help patients move from diagnosis to treatment and beyond. Marketing clarifies the scientific story. Sales provides trusted relationships and real-time understanding of physician needs. Access teams simplify reimbursement pathways. Patient support programs reduce administrative burden. Digital engagement reinforces and extends human interaction. The result is a commercial system that operates less like disconnected functions and more like an integrated network designed to help physicians help patients. This is the essence of Customer Excellence. It aligns the entire commercial enterprise around the real-world context in which care is delivered. The problem was never the call plan. The problem was the context. Key Takeaways Commercial performance in pharma organizations has traditionally been managed through field execution mechanics, yet the effectiveness of those mechanics increasingly depends on how well they reflect the real-world context in which physicians operate. Customer context has become the most pivotal commercial variable as administrative burden, payer complexity, and consumer-grade expectations reshape how prescribing decisions are made. HCPs now evaluate pharmaceutical engagement against the best experiences they encounter anywhere in their lives, raising the standard for clarity, responsiveness, and ease. Optimizing promotional activity alone is no longer sufficient. Commercial success depends on reducing friction across the journeys physicians navigate as they move patients from diagnosis to treatment. Customer Excellence represents the structural response, aligning marketing, sales, access, digital engagement, and patient support around the real journeys of care delivery . Diagnostic Questions to Consider Are we optimizing the activity of our field force, or designing commercial systems that support the real journeys physicians navigate to help patients receive therapy? How well do we understand the administrative, reimbursement, and operational barriers physicians encounter after they decide to prescribe a therapy? Do our commercial systems reduce the burden placed on physicians and their staff , or unintentionally add to the complexity of care delivery? Are we benchmarking our engagement against other pharma companies , or against the best experiences physicians encounter in their lives as consumers? Have our investments in digital platforms simplified the physician’s experience, or multiplied the number of disconnected interactions they must manage? Are we still managing performance through activity metrics alone , or beginning to understand the context that ultimately determines whether therapies reach patients? Closing Reflection The pharma and life sciences industry has spent decades refining the mechanics of field execution. Call plans, segmentation models, and targeting systems brought structure and discipline to commercial organizations. Yet the environment surrounding physicians has evolved far more rapidly than the systems built to support them. Healthcare professionals now operate in a world defined by consumer-grade expectations for clarity, responsiveness, and ease. When the experience of engaging with a pharmaceutical company fails to reflect those expectations, the contrast becomes impossible to ignore. Organizations that recognize this shift will redesign their commercial systems around the realities of modern care delivery. They will move beyond managing activity and toward understanding the context in which physicians help patients receive treatment. In doing so they will close the gap between scientific innovation and real-world impact. Your breakthrough science deserves experiences worthy of it. Together, we turn customer excellence into real-world impact. About the Author Wayne Simmons is a hands-on commercial excellence architect and founder of The Customer Excellence Agency, where he partners with pharmaceutical and life sciences leaders to turn customer-centric ambition into durable commercial advantage. He previously served as Global Customer Excellence Lead within Pfizer’s Chief Marketing Organization and has held leadership roles with Bayer Pharmaceuticals and The Ritz-Carlton Leadership Center. Wayne writes The Customer-Centric Marketer newsletter and is the author of The Customer Excellence Enterprise: A Playbook for Creating Customers for Life. The Customer Excellence Agency: Advancing the Pursuit of Excellence in Service of Science.

Tan building with a hanging sign against a clear blue sky