Marketing Mix Modeling (MMM) has long been a valuable tool across industries for data-driven decision-making. Recently, it has gained renewed attention in healthcare and pharmaceuticals as a means of addressing critical questions such as:

  • What impact do our promotional activities have on prescription volumes?

  • Which promotional channels drive the highest impact, and which are nearing saturation with limited additional returns?

  • How can we optimise resource allocating to maiximise ROI and drive sales?

While MMM offers significant potential, implementing it in the pharmaceutical industry presents unique challenges. This article explores these challenges and offers insights for effectively leveraging MMM in a pharmaceutical context, drawing on Artefact’s recent experience with various pharma clients.

Key challenges in implementing MMM in pharma

Challenge 1: Limited data availability and insight granularity

Data availability is critical for any modelling effort, especially regarding sales data. The pharma industry’s regulatory requirements ensure the safe and effective provision of medicinal products to patients, but these regulations also impose strict limitations on data collection, storage, and usage, which can vary by region. Although weekly prescription data at the individual healthcare provider (HCP) or facility level is ideal, data is often restricted to monthly sales figures at the national or regional level, limiting analysis depth.

Pharma’s promotional channels are also unique. Due to strict regulations, promotional activities cannot involve mainstream media like TV or social media. Instead, companies rely on direct marketing to HCPs through sales representatives, so MMM primarily uses data from these interactions, including event attendance and webinar engagement, which are often less structured.

Moreover, pharma includes diverse roles that interact with HCPs, such as sales reps and Medical Science Liaisons (MSLs), each with distinct objectives. Sales reps comprise the commercial workforce, while MSLs serve as scientific experts and intermediaries between pharmaceutical companies and the medical community. As MSL roles are non-commercial, these interactions cannot be considered in MMM for compliance reasons.

Challenge 2: Strengthening the data foundation for scalability

A strong data foundation is essential for scaling MMM, yet the rise of digital channels and omnichannel strategies complicates the process. Pharma companies increasingly aim to optimise interactions with HCPs across multiple channels, adding complexity.
Pharma companies usually face challenges with fragmented data sources, particularly across new digital channels like webinars, emails, and offline events. Often, this data is stored in separate tables and managed manually, leading to formatting or accuracy errors. While CRM data was robust, non-automated data sources presented integration challenges.

Challenge 3: A hypothesis-driven approach to modeling

A common misconception is that MMM can precisely model an entire market. However, given data limitations, implementing MMM in pharma requires a hypothesis-driven approach aligned across brands and countries.

For example, one significant challenge is the lack of detailed spend data for each promotional activity. The exact cost of each promotional activity is often unavailable, so we have to approximate spending by considering assumptions on sales reps’ salaries and time allocation.

Estimating event costs is also complex. Event expenses include not only the event itself but also personnel costs, content creation, and sponsorships, requiring ad-hoc assumptions validated with the business to ensure accurate cost accounting.

Additionally, some companies expect MMM to track all competitor actions in detail, which is often unrealistic. Instead, we should focus on modelling broader competitive dynamics, educating companies on data quality limitations. This approach helps set realistic expectations while providing actionable insights.

Challenge 4: Promoting adoption and business alignment

Technical solutions alone are insufficient; cultivating a data-driven culture is crucial for MMM adoption. The pharmaceutical marketing funnel differs from other industries, as patients access products out of necessity rather than desire, which can lead to scepticism among business teams. Thus, it is crucial to demystify MMM from the 1st day and foster collaboration with internal Business and Data Science stakeholders. To achieve this effectively:

  • Drive adoption from day #1 by simplifying MMM’s mathematical aspects for business stakeholders.

  • Collaborate closely with pharma Data Scientists, ensuring shared understanding of the model’s assumptions, strengths, and limitations.

  • Promote transparency and humility through continuous feedback, co-creating insights with the business to align with company goals.

Establishing an MMM framework for pharma

While pharma-specific challenges complicate MMM implementation, they do not make it inaccessible. Our previous experience with pharma companies demonstrates that they can typically attribute a 7-20% contribution to marketing performance, consistent with observations in other industries.

In pharma specifically, MMM must strike a balance: leveraging data without compromising the specificity and granularity needed for actionable insights. With two years of monthly sales and promotional data and strong internal stakeholder support, Artefact can deliver an actionable MMM framework for your company and upskill internal teams, as we did with Ipsen*. Our approach prioritises openness and collaboration, ensuring that MMM becomes a practical tool rather than a black-box model. Success relies not only on technical expertise but also on effective change management and team training.

*For more details, please refer to the our webinar with IPSEN on “How to maximize your Marketing ROI in the Pharmaceutical sector with the Marketing Mix Modelling (MMM) approach”