Reuters Events | Pharma

Thought leadership and innovation for the Pharmaceutical Industry

The omnichannel era presents new opportunities for pharma to more effectively engage with its customers by identifying next best actions (NBAs) driven by more relevant and timely data insights.

By Joseph Constance on Oct 9, 2023

Pharma has far more extensive data on the behavior of their healthcare providers (HCPs) and patients, which should allow them to personalize their engagement with them and impact their behavior.

But it’s a challenge to sift through all of the data to separate what is useful and actionable…from what is useless. It’s important to discern what information to prioritize, and which tools and platforms can help differentiate the signal from the noise.

Bayer Pharma’s work involving NBAs began a few years ago as a pilot effort, says Florian Ibe, Head of Customer Engagement Operations at Bayer Pharma. The company has generally developed two types of NBAs. One is a rules-based NBA focused, for example, on obtaining HCP consent. The other is more advanced and involves the use of AI algorithms to identify, for example, prescription behavior or interest in certain topics around brands, disease areas, or patient groups.

Buy, then build

Bayer initially worked with a vendor’s solution, into which the scientists fed data and leveraged the vendor’s algorithms and analysis. Bayer wanted to scale up the system beyond the pilot stage, but determined it would be expensive to do so.

“So we then started to build our own model with an external vendor, using our own algorithm, analytics, a different data platform, and other components. We were able to scale it and go beyond the initial pilot.

“We chose to complement what the vendor had, only taking parts of their offer because that was financially more attractive to us. We also decided that this was a core capability that we wanted to build out. So we hired experts, data scientists, and data engineers who would help build those algorithms and ensure that we had the right analytics in place while also leading the external vendor to implement some of their capabilities.  We worked with the vendor to implement the program on a data platform,” explains Ibe.  

Currently, Bayer Pharma is developing algorithms and analytics use cases for NBAs. The company has established a global group, which includes data scientists and analytics translators, who communicate with marketing and who work with the scientists to write the algorithms that prompt NBAs.

“The group communicates with our marketing and brand teams so they can learn the nature of the business in each region, including barriers and drivers. The group turns that information into NBAs, which are then validated and implemented with the local teams,” says Ibe.

Identifying relevant behavior

At the same time, Bayer asks the local teams to use HCP input from the field and identify behavior that is relevant for them. The central group then will help teams in the field to reconfigure and refine NBAs from other markets and make them appropriate for their market. The teams will determine how well the actions have been implemented and accepted in the field to measure acceptance and adoption. The results are correlated to prescription data sales growth to determine their impact on business overall, according to Ibe. The process involves an ongoing exchange between the experts and the field teams.  

One example of a successful NBA, Ibe explains, involves the propensity to prescribe based on previous behavior. “Depending on a physician’s activity and behavior, an algorithm is able to identify a likelihood to prescribe, and a suggestion or next best action for the rep that engages with that customer -- what they should be talking about to convert the interest of the physician into real prescriptions.”

The type of NBA recommended will differ, depending on whether a physician is interested in certain types of information, such as a drug’s efficacy information, its safety profile, or a patient support program, according to Ibe.

“We’ll see a difference in sales or growth for those reps that adopt more NBAs than others. The specific algorithms actually work if we address the doctors with the right messages or topics,” he adds.

Ibe sees three types of challenges to successfully developing NBAs. The first involves accessing and collecting the data. The second involves linking it to individual physicians, which can be done through a website, for example, if a physician identifies himself through a registration process, or through an email that can be tied back to a physician.

The third challenge involves sifting through the data and determining what is relevant and would drive an NBA and customer engagement, and what is useless. The process entails extensive trial and error testing, prototyping, and running different algorithms, he says.

Eliminate corporate silos

“The key to getting teams to work together is to eliminate (corporate) silos and better understand our customers’ needs. We have to have a written strategy in place and work in harmony, and agree on objectives and priorities. Everyone must have a specific responsibility,” says Mutlu Gunenc, Director, Digital Strategy and Innovation, at Beigene.

Many pharma companies are in a position to develop and offer a variety of digital marketing approaches for engaging with their customers. These go hand-in-hand with more traditional marketing approaches.  

“You need an open mind and must consider all angles when reviewing the data so you learn as much as possible about the provider and the patient. You ask yourselves, ‘What meaningful story can we generate?’ This has to be done consistently,” he adds. “Every time an HCP engages a channel or platform, you may receive new personal information about them. Teams have to be constantly ready to collect and assess that new data.”

A strong AI platform would seamlessly combine with different analytics, sources of data, and marketing data to enable companies to best manage their strategic engagement. Such platforms should be both easy to use and scalable, he indicates. “We must have a strong foundation in data analytics. We have millions of data points, and we need to interpret those data in a meaningful way,” he explains.

For example, when analyzing data and planning an NBA for customer engagement, a pharma marketing team may have a medical educational program that it wants to release to HCPs. By analyzing available data, the group may determine that it’s best to release the program on a Sunday morning when HCPs are more available to review it than during a weekday when they are busy seeing patients, according to Gunenc.

The Covid effect

Covid-19 continues to have a long-lasting effect on the behaviors of HCPs and patients. Being exposed to telemedicine and meeting remotely enabled them to receive information more quickly and in different ways than through traditional one-on-one meetings, and they still prefer those options by using social media and other channels, says Gunenc.

When trying to successfully engage with customers, Gunenc cautions that companies should not dictate how they believe their customers should behave. For example, a pharma sales team may prefer to have face-to-face meetings with HCPs, but HCPs may prefer to engage digitally, through a variety of channels, in creative ways.

“If you engage with them their way, they’ll be less resistant to your message. And if their ways change, you have to change your approach and remain aligned with them. Their needs are not static. We have to adopt to give them the content the way they want it, whether it’s in small bites on social media or with more information in emails or on our website,” Gunenc notes.

To reach patients, based on knowing their needs, pharma can engage HCPs and intermediary organizations to funnel disease and treatment information to patients for informational purposes, so that the companies do not appear to be trying to sell medicines directly to patients, according to Gunenc.

Challenges remain

But one challenge pharma teams face involves how different regions and governments, such as in Europe, have privacy laws that restrict the data available from HCPs and patients that can be accessed. Then data has to be collected anonymously, and trajectory analysis must be used to assess the HCP’s or patient’s needs. “We don't see this as an obstacle but as an opportunity to be respectful and trustworthy in our interactions with all relevant stakeholders,” says Gunenc.

Another challenge is the restricted physician access that has been instituted by some hospitals, which has forced companies to engage digitally, says Gunenc. “We need to find as many ways as possible to reach out and engage with HCPs, delivering the information they are looking for the way they want it.”