Dr. Wilson and Dr. Greenwood are two accomplished healthcare professionals associated with two very different health systems. They were born two decades apart and live in two different parts of the country. Dr. Wilson has a strong affinity toward learning about patient support programs and how they can improve lives. Data presented in the form of infographics resonates well with her. Dr. Greenwood, however, is more scientifically oriented and loves to consume granular clinical data and imagery (Figure 1). The one aspect they have in common is their patient centricity and their motivation to keep themselves educated on efficacious products and services for their patients. This leads them to interact with branded and unbranded communications from pharmaceutical companies, where they are expecting to have a very personalized, contextual and effortless experience.

 

Is pharma up to the challenge of delivering such experiences at the personalized level these doctors expect? Too often, the answer is no.

 

But content hyperpersonalization along the customer journey can help pharmaceutical brands cut through the clutter and deliver the remarkable experiences that Dr. Wilson and Dr. Greenwood want.

 

In the commercial model of the future, what will it take to realize the full potential of an intelligent content ecosystem? It requires deep understanding of customer preferences and prediction of microaffinities of customers across a range of content features including content context, aesthetics and tonality. And it requires the ability to dynamically curate a hyperpersonalized content experience across channels.

 

The transformation will require not only the smart use of data and analytics but also significant cross-functional organizational collaboration, process evolution and technological scale. While the success of an intelligent content ecosystem will rely on the transformation of the entire content supply chain, three key themes can help a program stand out.

 

1. Improving content planning through a smart data-driven strategy

 

Historically, market research and message mapping have informed content planning and creative development in the life sciences. Today, the process can be significantly augmented using a metadata-driven approach that leverages customers’ microaffinity for content features, improving customer impact and getting the most value for creative costs by producing content more likely to engage customers. In Figure 2, ZS research shows different content features can have varying degrees of impact on customer engagement.

Our analysis across a broad set of pharmaceutical brands indicates that aesthetics such as font size, number of images or the presence of infographics makes a meaningful difference in customer interactions. Similarly, the tonality of the content and even the placement of the call to action (CTA) button are also drivers of engagement. We can derive and learn customer preferences from these features and tag content using artificial intelligence (AI), create customer personas and use this intelligence to inform content development. Interestingly, ZS found that generative AI can deliver tags with similar accuracy as traditional natural language processing (NLP) and computer vision solutions with up to 50% effort.

 

For example, when the process was applied to a cardiovascular product, some customer personas preferred content that had an actionable tone versus an informative one. Similarly, different personas responded differently to features like number of images or number of CTA buttons. This data-driven approach can inform existing creative development processes. For instance, it can prompt creative agencies with new guidelines for modular content tailored to customer affinities—which can help manage the escalating creative costs of producing multiple content variants. Importantly, looking across multiple brands and therapeutic areas, ZS found that hyperpersonalized content that is dynamically curated based on customers’ affinity to content features can raise click-through rates between 20% and 35%.

 

2. Streamlining the medical-legal-regulatory process model

 

Intelligent content ecosystem at scale requires the evolution of medical-legal-regulatory (MLR) processes and enablement of a purposeful operating model between the omnichannel and regulatory affairs teams. The transformation hinges on process standardization, intelligent automation and an integrated governance model.

 

Process standardization: A good foundational step is designing and standardizing new MLR submission formats, such as dossiers that include pre-approved content modules, pre-approved dynamic content templates and examples of tactic illustrations. This process provides more context to reviewers and allows for a higher volume of tactics to be approved. The approach can be further enhanced using variation grids that showcase possible tactic variations, without the need to submit every variant for review and approval. Additionally, a systematic workflow that includes a targeted review process—such as routing scientific changes to pre-existing content only to medical teams, or graphical changes only to regulatory and editorial teams—generates additional efficiency. Based on ZS experience and research, a modular content approval process framework can generate 10 to 20 times the volume efficiency and reduce MLR cycle review time per package by more than 50%.

 

Intelligent automation: NLP can be used to extract sentences from new assets and matched up against a central repository of approved sentences for both concept discussion and content review meetings. The same process can be extended for accurate referencing of sentences by identifying the most similar assets. This scoring technique allows for multiple types of similarity checks and suggests whether an accelerated review pathway can be supported. For example, when this approach was applied to an immunology brand, ZS found 54% of sentences were repeats and did not need individual sentence review. As such, 33% of the total assets were largely similar and eligible to be routed through an accelerated tier approval pathway.

 

Governance model: An integrated operating model between marketing, omnichannel and regulatory groups with clear principles of accountability is a critical success factor. A well-functioning operating model supports the setup of agile cross-functional content collaboration teams and establishes parallel content co-creation sprints at scale. The accountability structure supports creating compliant content up front and encourages increased adoption of templates, business rules and process consistencies to enable efficient reviews. Establishing and monitoring a range of KPIs from foundational measures (such as MLR review cycle time per package) to economic measures (such as percentage bandwidth reduction of MLR operations) can also help companies track progress against their process improvement objectives.

 

3. Enabling personalization with a dynamic content deployment capability

 

The third key theme in realizing an intelligent content ecosystem in the industry focuses on an integrated process and technology framework to deliver dynamic and hyperpersonalized experiences across channels. Digitally advanced companies have developed a robust and integrated framework of solutions connecting data, decisions and activation to realize true omnichannel customer experience optimization.

 

Across the industry, a decision engine has started to play a key role in delivering personalized experience at scale. Most decision engines leverage AI to generate next best content recommendations to optimize customer engagements across channels. But these recommendations are at a composite asset level. In the world of the intelligent content ecosystem, decision engines need to be extended to make recommendations for the best available pre-approved asset variant with a modular content permutation that is predicted to drive the best engagement outcomes, as shown in Figure 3.

The recommended asset variant is then deployed in the channel as part of the overall engagement recommendation produced by the decision engine for a given individual customer. Based on ZS’s experience, brands that have piloted such decision engine-driven content hyperpersonalization processes have been able to generate an increase of 50% to 140% on click-through rates for their email tactics alone.

The industry is currently taking different approaches to modular content with varying degrees of sophistication, pace and outcomes. Some companies are taking a content infrastructure-first approach by setting up their digital asset management platforms and developing modular content. Some companies are taking a personalization-first approach by testing multiple content variants across different customer personas to better understand customer impact before building a full-scale intelligent content ecosystem.

 

No matter what approach you use, there are three no-regrets moves that companies can get started with:

  • Develop and apply an intelligent tagging taxonomy to better understand engagement drivers that can inform creative strategy
  • Develop an operating model that supports the use of modular content assembled via dynamic templates and associated business rules and align with MLR teams on a standardized approval process of asset variants
  • Pilot the deployment of dynamic content variants in existing customer journeys to learn and understand the hyperpersonalization opportunity

Customers like Dr. Wilson and Dr. Greenwood expect a dynamic experience, and content hyperpersonalization is helping elevate their experiences significantly.