Please login to the form below

Not currently logged in

AI’s role in pharma marketing

by David Laros


The advancement of artificial intelligence (AI) and machine learning (ML) capabilities not only has application to pharma’s R&D divisions, it’s also being used to develop more efficient and more effective promotional campaigns. 

Through these techniques, brand teams are gaining more granular insights into their audiences’ needs and desires, so they can deliver greater value through their digital multichannel campaigns aimed at healthcare professionals (HCPs).

Let’s look at the opportunities to gain these insights, and examples of the role that advanced analytics can play for savvy marketers.

Gaining a deep understanding of HCP behaviours

For years, we’ve been engaging HCPs through our Univadis platform, a global site for clinical news and education. Through millions of digital interactions, we have gained insights into how different types of HCPs engage and the preferences that individual HCPs have for certain topics, types of content, formats and tactics. Through our strategic analytics suite, AptusAI, we’re able to generate those insights and apply them to the promotional campaigns we design on behalf of our clients.

For example, our in-house data scientists apply AI/ ML techniques to perform content searches based on client needs (eg, disease, brand, molecule), then analyse, classify, define and tag relevant content with consistency. They then analyse HCPs’ affinity with the different content topics and types and synthesise the results into relevant, straightforward HCP insights.

There are clear opportunities to increase the impact of customer engagement through personalised approaches. In fact, across dozens of recent campaigns powered by our strategic analytics techniques, we’ve measured a median Rx lift of 3.8%, with an average ROI of 3.7:1.

Using predictive analytics to increase engagement

Most pharma marketers already do some kind of audience segmentation based on publicly available data. Yet through the application of predictive analytics, there is an opportunity to better align promotional efforts with targeted HCPs’ specific areas of interest.

For example, across several promotional campaigns on Univadis, we analysed the impact of personalised engagement based
on an HCP’s inferred interests, as measured by the increase in the click-to-open rate (CTOR). This helped to identify new ways to maximise the impact of these campaigns. In fact, personalised customer approaches based on content affinity analysis have led to higher engagement; among primary care providers we typically observe an increase of between 25% to 50% in CTOR compared to more traditional campaigns, with even higher results among secondary care providers (for which we see about a 40% to 65% increase in CTOR).

Here’s an example of how insights like these can be applied to a specific campaign:

Our client had a compelling message to share about treatment options for its women’s health therapy, yet had limited visibility into the HCP population for whom this message would be most relevant and valuable.

The natural question was: ‘Which HCPs have an interest in women’s health topics?’ We were able to determine the answer to this question by looking into the proprietary data on Univadis, analysing which topics were viewed by relevant HCPs and then identifying new HCP targets for our client based on this analysis.

By offering custom content in the formats and topics that HCPs already use and appreciate, the client was able to significantly extend the reach of its audience and engage audience members in key branded messages that also help them test their knowledge on treating common symptoms that our client’s brand addresses. These digital tactics featuring custom content increased engagement more than 2.5x over baseline metrics.

Using machine learning for persona development

Clearly, understanding preferences and behaviours at the individual HCP level is important for driving engagement among target audiences. Such insights can also be effectively applied at scale, using machine learning techniques.

For example, we worked with a client to develop a series of HCP profiles for both general practitioners and specialists, driven by ML. These ‘personas’ are based on observational data gathered from Univadis.

The HCPs within these profiles differ in terms of disease area/treatment affinity, content type affinity ranking, device usage and projected affinity for a brand’s strategic imperatives. This presents an opportunity to leverage these insights to provide the type of content and messaging for which they have the greatest propensity to engage. For pharma clients that use email as part of their promotional campaigns, these personas can be used to create tailored ‘hooks’ for each target to drive engagement. It’s just one of the ways we’re working with our clients to optimise their promotional efforts and make every interaction work harder for their brand.

For years, pharma has been effectively using digital channels to engage HCPs in their messages. Now, with advanced analytics, we are using these interactions to bring insights to life – offering more value to our clients, and more value to the HCPs they serve.

David Laros is VP of Global HCP Analytics at Aptus Health

In association with


11th August 2019

From: Marketing


Subscribe to our email news alerts

Featured jobs


Add my company

Hamell is a full-service agency with a clear focus on delivering evidence-based, sustained behaviour change. So, whether you are looking...

Latest intelligence

Infographic: Therapy Watch Hyperlipidaemia
Our latest infographic offers an introduction to the real-world insights collected by our Hyperlipidaemia study....
Article: The rise of the healthcare influencer
In our latest article for Pharmaphorum Deep Dive, we explore the opportunities and benefits for healthcare companies in the online influencer space....
Top-10 Reasons Why You Should be Adding Asynchronous Virtual Touchpoints Between Your Real-time Meetings