VideoWhat is Your Data IQLeveraging Causality for Data, Revenue Generation, & Your Business Strategy.

Introduction
0:00 [Music]
0:06 Thank you for joining us at Accelerate the Digital Velocity Summit and welcome to this session, “What’s Your Data IQ? Leveraging Causality for Data Revenue Generation and Your Business Strategy.”

0:18 I’m Dan Newman, Principal Consultant for Team Agility here at Agile Thought and your facilitator for this conversation.

0:24 It’s my pleasure to facilitate the conversation about causality and how to leverage causality to generate meaningful impact in the world.

0:32 If you have any questions as we go through our time together, ask them in the Q&A feature. Polls, chat, and community are disabled.

0:40 Question and answer are enabled during this session. As I’m just the facilitator, let me introduce the true stars of the session and experts on today’s topic.

0:51 I’m joined by Eric Sullivan, Senior Vice President of Innovation and Data Strategies from Innovalon; Sarap Alvarez, Vice President of Information Services at CMI; and Dr. Jerry Smith, Managing Director of Global Analytics and Data Science at Agile Thought.

1:08 Thank you all for participating in the session.

Working Definition of Causal Data

1:20 Eric: Sure, and good morning. First of all, as a disclaimer, I’ve spent the last 25 years in healthcare and analytics on the payer-provider side. I’m definitely considered a tactical data guy. I’m not a statistician or a machine learning coder.

1:47 Innovalon is one of the largest quality and risk analytics companies in the US for pharma, life sciences, payers, and providers. We leverage a lot of big data, and we focus on identifying the right interventions to improve quality.

2:13 Unlike randomized control trials, we deal with real-world observational data. We’re here to improve healthcare, and that means understanding underlying causes that impact interventions and outcomes.

2:38 Causal data isn’t easy to define. We’re dealing with humans. Social determinant variables play a huge role in why populations behave the way they do.

3:18 For example, our company Avalere studied dual-eligible populations—very complex Medicare/Medicaid groups—to understand drivers behind differences in quality outcomes like medication adherence and readmissions.

3:59 They found strong correlations after controlling for confounding variables, which ultimately helped influence CMS regulations.

4:07 Another example involved the University of Maryland studying how AI helps coders extract insights from medical records. They found reviewer circadian rhythms impacted AI effectiveness.

4:49 It’s not easy to find causality in observational data. There’s a spectrum between association and pure causality.

Understanding and Finding Causation

5:49 Sarap: Understanding and finding causation is extremely difficult. As humans and marketers, we see patterns and assume cause and effect without proving it.

6:26 Proving causality requires good experimental design, complex models, and common sense.

6:45 One example involved a pharmaceutical client launching a new ADHD drug for children. We wanted to understand what drove caregiver interest.

7:39 We assumed knowledge would drive interest, but we found that information-seeking behavior was the true causal driver.

8:27 Caregivers who relied on medical professionals showed optimism, while those relying on alternative blogs showed resistance—even though both felt knowledgeable.

9:12 This insight allowed the client to tailor messaging strategies based on information sources.

Why Causality Matters in Digital Transformation

10:30 Dr. Smith: From a cognitive psychology standpoint, these aren’t strange times—they’re predictable times. We want to change the world, not just observe it.

11:38 Companies invest heavily in technology but fail to change outcomes because they don’t understand causal relationships.

12:45 Data represents human activity. Once we understand causal data, we can change behavior, influence systems, and achieve desired outcomes.

14:55 When causal data is coupled with prediction models, it creates digital surrogates that allow optimization and prescription.

16:21 This creates a closed-loop system that bends the business curve and produces exponential growth.

Finding and Influencing Strategies

18:57 Sarap: Messaging strategies can influence behavior once causality is understood. Understanding competitive benchmarks is also critical.

21:23 Dr. Smith: Digital transformation teams often miss psychologists and sociologists, who help explain why people behave the way they do.

23:14 Eric: Telehealth adoption lagged because many seniors lacked device access—not because of willingness. Addressing that cause increased adoption.

Causal Modeling and Business Impact

34:12 Dr. Smith: Causal modeling creates asymmetric advantage. Companies using it grow exponentially while competitors grow linearly.

35:49 Human behavior changes, so models must be continuously refreshed.

Closing Thoughts

45:53 Dr. Smith: AI can now intentionally influence human behavior, which raises ethical responsibilities.

47:16 Sarap: Data science isn’t just about models—it’s about understanding business strategy and human behavior.

48:27 Eric: Start with practical problems, understand confounding factors, and build iterative learning cycles.

49:35 Thank you to all participants and attendees. We hope your Data IQ is higher than when we started.