July 16, 2018/Lauren Schivley/カテゴリーなし, Decision Support, Omnichannel

Q&A with Veeva’s Paul Shawah and Aktana’s David Ehrlich – Part 1

In this two-part series, Veeva Systems SVP of Commercial Strategy Paul Shawah and Aktana CEO David Ehrlich discuss how to leverage machine learning and AI to optimize go-to-market strategy — both today and in the years to come. Part one below focuses on the current state of machine learning in life sciences commercial processes, including applications, success factors, and the effect on the HCP experience.

What are the greatest opportunities for applying machine learning in commercial life sciences?

David Ehrlich (DE): You know the saying, “Half the money I spend on advertising is wasted; the trouble is I don’t know which half” (John Wanamaker, 1838-1922). With machine learning, life sciences companies can finally identify which half is ineffective and stop putting resources into it. Machine learning enables them to effectively analyze data and market dynamics to extract and deliver insights about what information an individual HCP (healthcare professional) wants, needs, and will make use of. Companies can use these insights to deliver the right information in the right way at the right time to the right person, boosting HCP engagement and care for patients.

The other major benefit of machine learning is efficiency. A lot of time and money is wasted on communicating with HCPs in ways that aren’t actually effective, such as holding in-person meetings with HCPs who prefer emails. By using machine learning to focus on preferred communications, we can drive tremendous cost out of the industry so that life sciences companies can focus on lowering drug prices or increasing investment on research and development.

Paul Shawah (PS): I agree that machine learning provides a great opportunity to boost efficiency. Life sciences companies invest a lot on field force and marketing promotions that aren’t optimized for HCPs’ needs. Their targeting is often based on the value of the customer (their potential of writing prescriptions and how much they’re currently writing) rather than what the customer actually values. Using machine learning to pinpoint what information an HCP values and how they want to get that information will free up investment, and enable life sciences companies to efficiently allocate resources in marketing and sales.

As you observe companies that are applying AI and machine learning, what do you think are the biggest success factors?

DE: The overriding success factor is truly integrating AI and machine learning across your organization and not isolating it in a technical department like IT or analytics. Many previous investments in data and business intelligence were ineffective because companies considered analytics a separate beast, confined to its own department. Although a lot of great insights were produced, by the time their proposals had worked their way through the decision-making process to lead to a change in strategy or execution, the findings were no longer relevant.

I would encourage companies to dive in and infuse AI and machine learning throughout the organization as much as possible. To do so, you must create a culture of innovation, appoint organizational leaders that pave the way to this new way of thinking, and plug machine learning insights into all operational processes and technologies to enable better decision-making.

PS:  What you’re describing reminds me of digital five to seven years ago. Digital was separate from both sales and marketing — it wasn’t integrated into brand planning or the tactical execution process. It was a separate team that executed digital and tried to work across all brand teams. Since then, we’ve seen much more success when digital is ingrained in marketing.

This is a common mistake with emerging technologies — they aren’t fully integrated into the organization until they’re proven, but it’s hard to prove their value until they’re fully integrated. AI is no different — it’s an essential technology that enables the business, not a separate playground.

DE: People know there’s value in AI, but there isn’t enough talent that knows how to extract that value. Machine learning is hard to do effectively and requires a lot of experience. Companies that are committed to trying it need to partner with AI and machine learning experts like Aktana, AI consulting firms, or academics. Of course, I’d be remiss if I didn’t say that Aktana offers access to all three AI expert perspectives (technology, consulting, and academic) through our own expertise, partner network, and advisory board.

What knowledge and skills do life sciences companies need internally to take advantage of machine learning?

DE: Most importantly, companies need to be clear on what they want to achieve from a business standpoint. This means knowing what problem they are trying to solve, who the stakeholders are, how they’ll measure success, and what the timeframe for getting there is. While it’s helpful to have a few people in-house to understand the work being done, knowing exactly what the company wants to achieve with machine learning is more important than having all the technical capabilities internally.

Another skill that we’re finding helpful for companies is the ability to set up control groups when deploying machine learning to understand what’s working and what’s not. Having someone in-house to set this up or help coordinate with a third-party is helpful for providing knowledge about the organizational culture and the business.

How does the HCP experience change when companies start using machine learning effectively?

PS: When done well, the HCP experience improves dramatically. Machine learning separates the signal from the noise. Top target HCPs may have hundreds, or even thousands, of touchpoints per year from life sciences companies, and the majority is noise, which is easily frustrating. Machine learning and AI filter that noise by identifying what information HCPs actually care about — what’s relevant to their patient population — and how and when they want to consume that information. Instead of receiving thousands of touchpoints, an HCP might just get ten relevant and timely communications. When this works effectively, HCPs will be better equipped to treat patients in the right way.

DE: We actually saw this change in a recent case study where we turned on an AI capability around emails sent to doctors. Originally, the company’s sales reps were sending emails every certain number of days, selecting them somewhat at random. When we turned on the AI capability, the number of emails sent dropped by 50 percent, but interestingly, the open rates increased by more than 50 percent. The AI filter, as Paul described it, constantly adjusts based on the HCP, her or his patient populations, and what’s happening in the marketplace to cut down on the noise.

Read part two of this series, where David Ehrlich and Paul Shawah share their insights about the future of machine learning in the life sciences commercial process.