Introduction

SPEAKERS:

William McCully, Co-Founder & Director, Aithena
Hank Du, Co-Founder & Director, Aithena

KEY TAKEAWAYS:

·      Pharma content approval averages 15 days per asset with £1.5M annual affiliate costs

·      AI pilots demonstrate 90% total review time reduction saving approximately 75 hours per asset

·      Technical review achieves 61% time savings with linear scalability processing 100-200 claims

·      Other industries such as financial services have reduced compliance review time by 98% using AI rule-checking systems

·      Human signatories retain full accountability for AI-reviewed content, supported by a structured oversight framework

·
The pharmaceutical industry has a content crisis hiding in plain sight. "When I was having my breakfast this morning, I went on the top 10 pharma websites and looked at your product pages. Six out of 10 of them said down for maintenance or content coming soon," William McCully told the audience at Reuters Events: Customer Engagement Europe in London.

This isn't a technology problem, it's a compliance bottleneck.

"The average time it takes to get one piece of material out to customers is about 15 days. Half that time is in review, half that time is the back and forth re-edits," McCully said. The Medical Legal Regulatory review process was designed to ensure patient safety and regulatory compliance. But it's become the thing that stops commercial teams from moving fast.

The £1.5 Million Question

The economics are stark. "The average affiliate probably spends around about 1.5 million just on review time alone. That's not outsourcing to a third party. That's not the cost of re-edits with a creative agency, and it's certainly not the opportunity cost to teams spending too much of their time in review," McCully said.

Here's the kicker: two-thirds of PMCPA breaches come from preventable errors. Missing mandatory information, outdated data, unsubstantiated claims. The system is slow and imperfect.

Marketing spend is projected to grow 10% year-on-year over the next five years across Europe. AI content generation and omnichannel strategies are multiplying content requirements. Organizations already struggling with current volumes face an impossible choice: exponentially grow review teams or fundamentally redesign the process.

"We've seen some really positive results in our early pilots with an average 90% reduction in total review time. That's equivalent to roughly 75 hours or just over nine days time saved per asset," Hank Du reported.

But the time savings do not come from AI reading faster. They come from eliminating review idle time. Du explained further, “What we actually realised is that reviewers spend a relatively small amount of time in the MLR system doing the review itself. They spend much of their time on MLR-related activities, such as reading references or aligning with brand teams on brand messages.”

While all of that is happening, the content effectively sits in a queue waiting for someone to pick it up, and that queue time drives much of the latency. “That is not a problem for AI,” Du noted. “An AI review can begin immediately and can run continuously in the background, processing one piece of content after another without waiting for a human reviewer to become available.”

Where AI Actually Excels

Technical review is one of the most labour-intensive parts of MLR review. “It is a highly demanding, laborious process and costly; this stage is often outsourced to third parties,” Du said. Large language models (LLMs) can interpret complex scientific documents, identify factual statements and claims that require substantiation, and automatically link them to supporting evidences. “In our pilots, we see materials containing more than 100 or even 200 discrete claims, and the AI has been able to understand them, label them appropriately, and detect technical errors with an impressive degree of accuracy,” Du explained.

"AI is not simply picking up missing subscript or superscript in your text. It is actually understanding the information, applying the right references to the claim."

Compared with traditional human-only review, we estimated a 61% reduction in technical review time when using AI support. In addition to the raw speed improvement, the real advantage lies in how AI scales. “When you work on a large document, every additional page and every additional claim or reference adds disproportionate complexity for a human reviewer. For AI, the workload scales in a more linear way,” Du said.

Need five email variations instead of one? Human review complexity explodes with each additional claim, reference, and context dependency. AI processing time increases proportionally. Same accuracy, predictable scaling.

Financial services has already been down this road. Some organizations have cut compliance review time by 98% using AI to navigate rules and regulations. Same stringent requirements, massive efficiency gains.

Teaching Compliance, Not Just Catching Errors

Traditional MLR review works as a quality gate. Content comes in, reviewers find problems, content goes back for fixes. This concentrates compliance expertise with senior signatories while content creators learn through trial and error.

Explainable AI changes that dynamic. "When you think about compliance, you can think of it as an iceberg. There are surface level issues that are easier to detect, like prescribing information or job code. People without extensive compliance training can identify them, but we know those are not the ones that trip up the MLR review," Du said. "AI surfaces all, including nuanced issues, at the beginning, leveling that playing field, no matter what your experience is."

"Explainable AI, when built correctly, can help the content creator not only see the error, but explain exactly why it is wrong, why it matters and how to fix it," Du said. Over time, content creators will internalise these principles and error rates in initial submissions should drop.

It also brings fundamental questions to the surface much earlier. Du recalled his own experience: “I remember working on a piece of content that had already been through several rounds of review and was close to approval when someone said to me, ‘Hang on a second, should this even be promotional?’ That is not the sort of conversation you want to be having when the content is about to go live.”

For global organisations, this democratises expertise across markets. "We can have global to local frameworks so a global team can create a whole series of campaigns, for example an omnichannel campaign, and now approve it for every local market at source," McCully said.

Who Signs Off?

The workforce question comes up in every implementation discussion. If AI automates the junior MLR review tasks that traditionally build expertise toward signatory roles, how do you develop future compliance leaders?

"Having been a signatory before, you earn your experience as a reviewer, then over time you become experienced enough to be a signatory and take on that responsibility to sign off materials," McCully acknowledged. Roles will likely shift toward orchestration - defining campaign strategy, managing AI oversight, making nuanced risk decisions.

But the accountability framework stays the same. "The accountability sits squarely with the person who's signing off the content. So whilst the AI can identify and pick up issues along the way, it is ultimately the person who's signing off the material that's accountable," Du said clearly.

"You should think of AI as a tool, but not as a replacement for human judgment. A well designed system with the right safeguards to serve up compliance risks for human intervention.

 The future vision goes beyond faster approval. "We're going to have agentic compliance workflows. You're going to have a compliance agent, a marketing agent, a regulatory agent, all working in tandem behind the scenes," McCully said. These workflows could automatically update content when SmPCs change or new clinical papers are published.

 "When we start to adopt AI content generation, we can couple that with AI compliance and now we can have compliant content at source. Again, this is a pace of marketing that we've never seen before," McCully said.

 "This isn't necessarily about speed of approval, this is about the speed of impact. And MLR review may not be the most glamorous topic, but it's the engine that drives content through to our stakeholders," McCully concluded.

 The pilot data suggests the technology is ready. The real question is whether organisations are ready to redesign decades-old processes around capabilities that promise both speed and quality at the same time.

To get you highlights of Pharma Customer Engagement EU 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.

"Historically we always wanted to be customer centric, but now we have the data and the operations to understand our customers, evolve to where they are, and let them tell us what they require and what they need,"

Canlas explained.

SPEAKERS:

Chiara Bolognini, Global Lead CX, Marketing and Sales, Roche
Kruti Popat, Director of Customer Experience, Boehringer Ingelheim
Yvette Venable, Head Global Market Access, Payer Partnerships and Policy, AstraZeneca
Jalilah Gibson, Associate Partner, Wavestone (Moderator)
KEY TAKEAWAYS:
• Fragmented touchpoints erode trust when customers experience pharma companies as different people across functions
• Early R&D engagement with payers and regulators prevents costly market access evidence gaps
• Patients consolidate all stakeholder perspectives, making patient experience the ultimate engagement quality test
• Human expert interpretation prevents data misattribution that misdirects strategy and wastes resources
• AI-powered responsive ecosystems enable customers to choose content rather than receive pushed messages

Imagine meeting a pharma company representative who speaks eloquently about patient outcomes, then encountering the same company through customer service where the tone, messaging, and approach feel completely unfamiliar. This jarring disconnect defines the current state of pharma customer engagement. "If you move for example, from sales to customer service, they feel lost our customer experience," Chiara Bolognini explained at Pharma Customer Engagement Europe 2025.

As Jalilah Gibson reinforces, "Our customers don't care which company we work for or which part of the function that we work for, but they do care when they get a very disjointed experience."

From Fragmented Touchpoints to Unified Ecosystems

The pharma industry's traditional organizational structure creates silos that customers experience as disconnected interactions. Sales, medical affairs, market access, and patient support each optimize their own engagement models without coordinating on how the company's identity is expressed across the patient journey. Trust, the foundation of prescribing influence and treatment adherence, erodes when customers cannot recognize consistent company presence.

"It's designing ecosystems where our customers can thrive, where we build the familiarity with them and basically expose them to best attributes of our identity," Bolognini emphasizes. This shift from isolated touchpoints to integrated ecosystems requires governance that operates above functional levels, with authority to transfer methodologies across commercial, medical, and patient partnership organizations.

For Kruti Popat, the strategic imperative is clear. "Delivering a seamless customer experience is not nice to have. It's really important strategic imperative. It's a really great differentiator," she notes, particularly in complex environments like the UK's NHS where multiple stakeholders each have distinct needs yet expect coherent company engagement.

The ecosystem approach extends beyond commercial interactions to the earliest stages of development. Yvette Venable describes how AstraZeneca's "Transforming Healthcare" framework orients all teams toward patient outcomes and health system resilience. "It's crucial that we're talking to patients, to payers and policymakers, and also to clinicians very, very early on in R&D in order to make sure we can change those paradigms over time," Venable explains.

This early engagement prevents the costly disconnect between regulatory approval and market access that plagues many launches when evidence generation fails to address payer requirements.

<add orange lines for page break and this text>

Discover more on this topic at Pharma 2026 (22-24 April, Barcelona) - the global collaborative network for leading pharma innovators. Visit the website here.

The Patient as Universal Stakeholder

Patients aren't just another customer segment; they represent the consolidation point of all stakeholder perspectives. Healthcare providers treating patients today become patients themselves tomorrow. Policymakers making reimbursement decisions will eventually navigate the healthcare system as caregivers or patients.

"The patient is the consolidation of all our customers. Each of us in the company, so experiencing the perspective of the patient and the customer experience is fundamental," Bolognini emphasizes. This perspective shift has profound implications for how companies design information architecture and support systems across disease awareness, diagnosis, clinical trial participation, treatment initiation, and long-term management phases.

Yet the industry still struggles to move beyond transactional engagement models. "We as an industry do so much better today than we ever have on patient engagement, but at the same time, we're still not doing it well enough," Venable observes. The gap lies in viewing patient engagement through the lens of extraction rather than true partnership with mutual benefit.

Structural issues within organizations, where patient advocacy teams, patient support programs, and commercial patient engagement operate independently, create the same fragmentation patients experience externally.

The solution lies in persona-based approaches that maintain message consistency while adapting delivery. "It's understanding what is important to these personas, what's the value, what language resonates for them. You're still using the same content thread, you're not amending the content," Popat explains. For HCPs with high empathy toward patient journeys, the same clinical evidence is delivered with emphasis on patient experience rather than purely clinical endpoints.

Bridging the Evidence Gap Through Multi-Stakeholder Dialogue

A critical challenge in pharma development is the misalignment between regulatory requirements and payer value recognition. Clinical trials are designed using validated endpoints that satisfy regulatory approval, but health technology assessment bodies and payers often require different evidence, particularly hard outcomes rather than intermediary endpoints.

Venable shared a powerful example from AstraZeneca's experience with a chronic condition therapy. The regulatory-approved intermediary disease progression endpoint wasn't acceptable to payers demanding solid heart outcomes. However, extending trials to capture those outcomes would take so long that withholding new treatments from patients would be unethical.

"We got the regulators, the payers, clinicians and patients in a room together to have a conversation about how meaningful these endpoints were," Venable recounts. The resulting dialogue revealed striking disconnects, with clinicians sometimes "literally head in hands going, how do you not understand how important this is to clinical practice?"

While the immediate outcome didn't change payer positions on endpoint acceptance, it initiated the long-term process of building shared understanding across stakeholder groups that often operate in isolation from each other. This example illustrates why early multi-stakeholder engagement in R&D is essential.

When market access teams, medical affairs, and R&D collaborate from early development stages, they can shape trial designs that satisfy both regulatory approval and value demonstration needs. The alternative creates costly delays and restricts patient access to innovations.

Insights, AI, and the Human Interpretation Imperative

The pharma industry is investing heavily in data analytics, real-world evidence, and AI-powered insights to understand customer behavior and optimize engagement. Yet without human expert interpretation, these investments can generate misleading conclusions that misdirect strategy and waste resources.

Venable illustrated this risk with a striking anecdote about real-world evidence analysis. Data showed consistent patient treatment drop-offs at six months, which analysts attributed to intolerable side effects. "You bring a patient expert in and they look at that data and they go, well yeah, you need refrigeration and cold chain and everybody goes on holiday in August," she explained.

The operational reality was invisible in the data itself but immediately obvious to patients living that experience. As Gibson reflects, "Too often we've sat in data point and go, why is that happening without asking actual customers." This gap between data observation and customer validation leads companies to address the wrong problems.

The solution requires building patient expert consultation into data interpretation processes, not just data collection phases. Roche has developed sophisticated infrastructure to consolidate insights while enabling actionable application. "We have actually created a wonderful insights platform called Before Insights Intelligence, linking to our R&D excellence program to basically identify commercial ability of our molecules," Bolognini describes.

The platform uses large language model interfaces, allowing marketers to request insights for specific strategic needs rather than analyzing dashboards themselves.

Responsive Ecosystems and the AI Transformation

The convergence of these themes points toward a fundamental transformation in pharma customer engagement. The future belongs to responsive ecosystems where customers choose their own content through AI-powered interfaces rather than receiving pushed messages through predetermined channels. Large language models connected to validated company knowledge bases can provide instant, personalized responses while maintaining narrative consistency across stakeholder groups.

This technological capability must be balanced with human insight. Emotional journey research, patient expert validation of data interpretations, and persona-based delivery ensure that efficiency doesn't come at the cost of empathy. "Delivering a seamless customer experience empowers accelerated decision making and therefore ultimately patient outcomes," Popat notes.

The companies that will differentiate themselves are those that establish governance above functional silos, co-create narratives across medical affairs and commercial teams, and engage multiple stakeholders from early R&D through market access. Technology enables scale, but human expertise ensures relevance and trust.

The pharma industry's ultimate customer, the patient, navigates a complex ecosystem involving diagnosis, treatment decisions, adherence challenges, and disease management. When companies design experiences that support patients through these transitions with consistent identity, validated insights, and responsive support, they build the trust that influences not just individual treatment choices but entire standards of care. The shift from fragmented touchpoints to unified ecosystems isn't merely operational improvement. It's the foundation for delivering meaningful health outcomes at scale while strengthening relationships across every stakeholder group that influences patient access to innovation.

To get you highlights of Pharma Customer Engagement EU 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.

"It's the combination of the patient and the HCP journey. As the patient is living with their disease for two years, three years, the information that physician needs is very different than the patient that walks in and says, 'I have these symptoms and I don't know what's wrong with me.'"

explained Dalton.

Wall highlighted the importance of considering the entire patient journey, not just the initial prescription: "One of the things I think we don't talk enough about is that it isn't just that first script. Part of the journey is also when you stay on something, and your physician helps you throughout the entire journey."

Cross-Functional Collaboration and Implementation

Successfully implementing omnichannel strategies requires breaking down traditional silos between departments. Canlas emphasized the importance of cross-functional teams that include "the entire commercial customer engagement, commercial operations team, the brand team, medical, legal, regulatory, and privacy."

The panel agreed that involving compliance teams early in the process is essential. "Having very good conversations and transparent conversations with your compliance colleagues are just key," said Canlas. "Leaning in, bringing them in that level of inclusion earlier in any experiment that you're looking to do, any sort of journey mapping exercise, is so key."

Measuring Success Beyond Activities

When asked about the most impactful KPIs for omnichannel marketing, the panel emphasized business outcomes over activity metrics. "MBRX [market basket Rx]," Canlas stated firmly, "because that is the culmination of all of the efforts. That means that physician and that patient have made a choiceful decision to start therapy."

Wall cautioned against metrics that focus solely on output: "Often I hear my friends and colleagues talk about how much they did, how many assets they created, how many channels they're on... if at the end of the year those people write in their self-input, 'I created a lot and I did more,' then we're measuring the wrong things."

Key Takeaways

1) Patient-centricity drives effective omnichannel: "It's the combination of the patient and the HCP journey... You start with the patient, identify then who are the HCPs in that moment," said Canlas.

2) Content strategy requires fundamental transformation: "Instead of that claim, chart, and table, we need to think about five claims, three different ways to display that data, and four different charts," explained Dalton.

3) Cross-functional collaboration is essential: "It's bringing the entire commercial customer engagement, commercial operations team, the brand team, medical, legal, regulatory, and privacy together," noted Canlas.

4) Measure business outcomes, not activities: "MBRX is ultimately the most customer-centric KPI," stated Canlas, emphasizing the importance of measuring actual patient impact.

5) Consider the complete patient journey: "Part of the journey is also when you stay on something. And your physician helps you throughout the entire journey," Wall emphasized, highlighting the need to support patients beyond initial prescription.

Text

Discover more on this topic at Pharma 2026 (22-24 April, Barcelona) - the global collaborative network for leading pharma innovators. Visit the website here.