Wayne St. Amand: Thanks so much for sharing this story with the LTEN community today. Keenan, are you ready to start at the start?
Keenan Stare: Let's get into it. Before we do, I spent about eight years in the field as a sales specialist before moving into training, which is where I grew an affinity for this work. I also want to thank Wayne, the entire Quantified team, Tim and LTEN, and the cross-functional partners I worked with throughout this launch.
Wayne St. Amand: Let's start with the situation at Novartis. There's a bit of drama here, because the timeline changed significantly once you started seeing momentum with the FDA. What had you originally planned, and what changed?
Keenan Stare: We were preparing to expand one of our therapeutics into earlier lines of care. As our team stayed in close conversation with the FDA, we picked up on signals of strong momentum, and that consolidated our timeline. We ended up accelerating our launch training program by about three months.
That meant consolidating the traditional learning modalities we had designed, while still building in meaningful moments for reps to apply the new information. It also meant rethinking not what readiness was, but how we would keep readiness consistent across the entire field team. Historically, we had run certification in person, usually at a launch meeting timed to approval. This timeline shift demanded we be ready to deliver that same rigor virtually. That was doable, but it introduced a lot of codependencies and leaned on a much larger cross-functional team.
We started thinking about our experience with Quantified. We had used it before for knowledge assessments, not the simulator specifically, but some of our colleagues had piloted the simulator with really strong success. That felt like an opportunity to meet our accelerated timeline while also improving the model: greater consistency and objectivity in evaluations for our learners, better benchmarking and visibility for our sales leaders to inform coaching, and a lot more simplicity in training the cross-functional partners who had previously been part of the process.
Wayne St. Amand: That feels high stakes, on an accelerated timeline no less. What made the stakes so high here, and how did you approach it?
Keenan Stare: Part of it was the novelty of this therapy, moving the standard of care earlier in a space where providers already had years of confidence in how they treated patients. We had to get this right, both because our patients would not otherwise benefit, and because sales in our existing indication had plateaued. This represented a real opportunity to help more patients and recapture momentum for the brand.
Moving earlier in the treatment journey also changed the nature of the conversation. We had historically talked with customers about sequencing therapies. Here, the conversation shifted to intensification: how much therapy a given patient needs and deserves to get the outcome they are after. We needed training that directly reflected that strategic shift, to maintain the credibility and trust our sales force had built. We had also done internal research that identified four specific sales capabilities spanning the entire indication, and we needed to evaluate and benchmark reps against those to inform ongoing coaching from our first-line leaders.
Wayne St. Amand: Every organization defines readiness a little differently. How did you define it for this launch?
Keenan Stare: To be transparent, this is a step a lot of training programs get wrong, myself included earlier in my career. Competency alone was not going to be enough given what was at stake. Our goal was for reps to consistently translate new information into commercial impact in real customer dialogue, not just demonstrate knowledge.
Understanding the disease state and patient journey was not enough on its own. Reps needed to build compelling narratives about the unmet need in this patient population, even while the standard of care they were up against had strong clinical guidelines and years of physician confidence behind it. We relied on our sales capabilities to help reps surface that unmet need and ask the kind of provocative questions that get a customer to reconsider their own standard. On top of that, this patient population was far more heterogeneous than what reps were used to, which meant a much more nuanced patient identification approach. Our brand story did not change based on a given patient's profile, but how that patient experienced our product's benefit did, so reps needed real storytelling skill to translate key messages into real-world benefit, not just recite disease-state statistics.
Keenan Stare: Three things stood out to us. First, flexibility. We had seen a full spectrum of simulations built by our peers, covering both HCPs and non-prescribing stakeholders, applied to a range of business objectives. That gave us confidence we could build something relevant to our own strategic priorities.
Second, feedback. Evaluation was objective and consistent, aligned to our own selling model rather than any individual evaluator's discretion or bias, which made it equitable for everyone taking it. The data also surfaced real trends for us: we saw reps grow quickly in disease-state fluency with practice, while softer skills like storytelling sometimes stayed a bit lower, which gave our sales leaders a clear diagnostic for where to coach next.
Third, operational efficiency. Once we had approval, we had roughly a 24 to 48 hour window to get the team through certification, and our traditional model depended heavily on cross-functional partners who were tied up in their own launch priorities at that exact moment. This let individual reps spend the time they personally needed rather than follow a fixed schedule, and gave us a leadership dashboard to track practice and completion status in real time, which mattered for both support and compliance.
Wayne St. Amand: Was there anything that surprised you as you built the program?
Keenan Stare: How intuitive it was, even as a first-time user. I expected I would need to feed the model specific scripted quotes. Instead, the model took general, bulleted inputs, things like an avatar's beliefs, preferences, and concerns, and turned them into natural conversation. The build itself was lighter than I expected. What took real time was quality assurance: testing the scenario from different angles, since this was our first time working with this specific capability and we wanted to be sure it was right no matter how a rep approached it.
Audience question: In finding the unmet need, did you program in the questions reps needed to ask, or just the answers?
Keenan Stare: We do not feed the simulator anything we want the specialist to say. We coach the model on what matters to the avatar, their biases, preferences, and concerns, and it applies that conversationally. For our scenario specifically, we built in an avatar who was mildly resistant to the idea of an unmet need, or strongly attached to the existing standard of care, and let reps work to shift that. We also built a scenario that pushed reps to identify which patients were at the greatest risk using their own sales tools and strategy, since our patient population was broad and heterogeneous rather than defined by a few simple checkboxes.
Audience question: What happens if a learner says something noncompliant during the simulation?
Keenan Stare: We had full autonomy to build stringent criteria, including uploading our own promotional materials and prescribing information to inform the avatar. One of our own criteria, for example, flagged off-label questions so they would be triaged to the right cross-functional partner.
Wayne St. Amand: Every Quantified AI roleplay is built around your MLR-cleared library of material, so it is designed with compliance in mind from the start, listening for noncompliant language and flagging it so a rep can adjust in practice before it ever shows up with a real customer.
Keenan Stare: The framework was simple: role play, receive feedback, adjust your plan, repeat. We asked reps to practice at least three times before certifying. We did not gate that minimum, it was more of an iterative expectation.
After approval, we ran a few deliberate steps. First, a formal introduction to the platform and what we expected. Then a pre-call planning worksheet reps completed in advance of a workshop led by first-line leaders, where reps socialized their plans, got feedback from peers and leaders, and heard expectations reinforced directly from leadership. From there, reps had on-demand access to practice on their own, refining their approach based on feedback before certifying with an updated plan.
Wayne St. Amand: What held reps back from jumping in the first time?
Keenan Stare: I would split it into fear and skepticism. The fear was mostly about being recorded and watched. We addressed that early: Quantified does not record video of these sessions, it captures a transcript, which actually benefits the rep, since we can use it to correct a score if there is a discrepancy. The skepticism was broader, whether this would really reflect a real customer conversation. Both tended to fade once someone actually went in and practiced for the first time. On-demand access meant it was also just easier to hit a button when you personally felt ready, rather than wait for a scheduled session, and that helped create a safer space for reps to feel vulnerable and build confidence at their own pace.
Keenan Stare: I would break it into three periods. Pre-approval, once we saw early FDA momentum, we built our own business case for Quantified, brought it to cross-functional leadership, and stood up a small task force across training, marketing, and sales leadership to define the scenario, logistics, and which customer objections to prioritize.
At approval, that tiered rollout, the platform introduction, the pre-call planning, the workshop, gave reps what I would call a battle-tested plan before their first practice attempt. Post-approval, we opened the certification scenario for reps to complete at their own pace, and I was able to watch the dashboard update in real time to confirm everyone was meeting our requirements. That was a sharp contrast to how we used to do this, chasing down paper scorecards from a dozen or more rooms and evaluators.
Wayne St. Amand: Now for the fun slide. What did you see once this actually ran?
Keenan Stare: Like anything new, we saw some healthy skepticism at first, since this was the team's first time with this specific capability. That changed quickly once people started engaging with the platform. Every rep passed certification within our defined timeline, and a lot of people went well past what we asked for: we saw an average of 4.5 practice sessions before certifying, with some reps stopping at three and others practicing as many as twelve times. Across just two days, the team logged roughly 150 hours of simulation time, in sessions that ran about twelve minutes on average. Scores moved too, practice attempts averaged 4.1 out of 5, certification attempts averaged 4.5, which tracks with reps sharpening their approach each time they went back in.
We also heard from reps directly, people telling us how real the experience felt, and others worried the practice environment was going to go away because they wanted to keep using it to try different approaches. Overall the program was really well received, and it is still very early, but we are launching with great pace. Because of that reception, we are now expanding our Quantified library to new customer segments and adding non-prescribing stakeholders to address some of the non-clinical factors our team needs to execute on.
Wayne St. Amand: What is your advice for other teams trying to get reps to make that first attempt?
Keenan Stare: Lean on your own experts to build a relevant scenario, and bring in the field's perspective where you can. And be proactive about setting aside real time for testing and quality assurance. It is rare that anyone in training is working on only one thing, but since this was the team's first time with this capability, and we were some of the only people with visibility into the final approved materials, we had to protect that time intentionally. Otherwise quality can slip when priorities compete.
Audience question: How did you get your AI governance committee comfortable approving the use of your product information inside an AI-powered simulator?
Keenan Stare: I was not directly part of that conversation, our operations team led it, but my sense is that this is a strong risk-mitigation tool. How narrow or broad you build your criteria is entirely up to you, and you are informing the model, not the other way around.
Wayne St. Amand: We have a team internally that specializes in exactly this conversation. The specifics vary by company and policy, but we have worked with governance committees across life sciences and pharma specifically around compliance, privacy, and data standards, and we are glad to help any team working through those questions.
Audience question: How often did discrepancies show up between a rep's self-assessment and the simulator's score?
Keenan Stare: We did not have reps self-assess beforehand, but we consistently saw scores trend upward with practice, which tracked with the 4.1 to 4.5 movement I mentioned. Anecdotally, we also heard a lot of positive, specific feedback from reps about how the flexibility helped them try different engagement strategies before landing on their approach.
Audience question: What was the impact on your training team, more work hours, fewer, or about the same?
Keenan Stare: The investment shifted rather than disappeared. It is heavier on the front end, developing the scenario and doing quality assurance, and much lighter on the back end, since everything reports into one dashboard instead of a manual collation process. Overall it saved me an enormous amount of time, and I am now working on an agent that can help build these kinds of scenarios even faster for future launches, whether they end up delivered in person or through a simulator.
Audience question: What criteria or competency framework did you use to evaluate reps, and what tool did you use to calculate scores?
Keenan Stare: We aligned the criteria to our own organizational sales model, a handful of key pillars we developed with our skills training partners into a standardized framework. For scoring, Quantified's dashboard calculates everything, nationally, regionally, and individually, so leaders and trainers can see trends at whatever level they need.
Wayne St. Amand: Keenan, thank you for sharing your story. You were constrained by a tight, shifting timeline, and still found a way to make this launch successful.
Keenan Stare: Thanks for having me, and for being a fundamental part of making that happen.
Keenan Stare is a Product and Disease Training Manager at Novartis, designing and delivering innovative learning. His work focuses on Radioligand Therapies and translating complex clinical and matrixed business strategies into engaging, practical development programs, with an emphasis on experiential learning & sales capability development. Keenan is passionate about creating learner-centered approaches that combine meaningful practice, coaching, and technology to accelerate field readiness and performance. Outside of work, Keenan enjoys exploring new experiences with his wife and son.
