CDO Matters Ep. 38 | It’s Crunch Time for CDOs with Mark Stouse
Nov 30, 2023
auto_awesome
Mark Stouse, Founder and CEO of Proof Analytics, discusses the frustration organizations face in realizing benefits from data management investments. He highlights the need for CDOs to prioritize specific business needs and outcomes. The podcast covers challenges of operationalizing analytics, the value of automation and targeted AI, establishing causality in predictive analytics, and the current state of data science functions.
CDOs should prioritize delivering insights and answers to the 20 key questions that business leaders care about, even if the confidence scores are in the 40-50% range.
Data analytics needs to move away from a shell game approach and focus on providing clear, pragmatic, and actionable answers to the business's pressing questions.
Deep dives
The Importance of Operationalizing Analytics for Business Success
Data leaders and Chief Data Officers (CDOs) need to focus on operationalizing analytics to help their organizations become truly data-driven. This involves overcoming challenges such as latency issues in data science, scalability, cost, and cultural differences between precision and pragmatism. The goal is to deliver value to the business now and earn the opportunity to generate further value in the future. CDOs should prioritize delivering insights and answers to the 20 key questions that business leaders care about, even if the confidence scores are in the 40-50% range. Data science teams should adopt a T-shaped approach, understanding the business and using business language to communicate analytical insights.
The Problem Between Data Science and Business People
There is often a disconnect and frustration between data science teams and business leaders. Data scientists tend to prioritize academic perspectives, over-focusing on complexity, perfection, and scalability. However, for business leaders, the focus is on making decisions that lead to incremental improvements and better outcomes every day. Business leaders are not interested in highly precise and caveated predictions. Instead, they value practical insights that help them make better decisions in a probabilistic world. Data scientists need to adjust their approach and speak the language of the business to bridge this gap.
The Need for Immediate Value and Lean Data Approaches
Organizations are often fixated on data perfection and big data management, which can delay delivering value from data analytics. However, the reality is that most companies do not possess big data stocks; their data represents time series and lean data. Data analytics should prioritize delivering immediate value with lean data approaches, such as regression analysis. By focusing on the business's key questions and delivering insights with greater than 50% confidence scores, data analytics teams can provide incremental improvements and demonstrate their value to the business.
The Shell Game and the Urgency for Delivering Value
CEOs and business leaders are becoming increasingly frustrated with data science departments that provide answers with numerous caveats and complexities, leaving them unsure of what actions to take. This frustration stems from the perceived disconnect between data science and delivering tangible value. Business leaders want actionable insights that help them make decisions and drive outcomes. Data analytics needs to move away from a shell game approach and focus on providing clear, pragmatic, and actionable answers to the business's pressing questions.
In this 38th episode of the CDO Matters Podcast, Malcolm interviews Mark Stouse, the Founder and CEO of Proof Analytics. An accomplished business leader with decades of experience in leading data science initiatives, Mark shares his perspectives on the growing frustration that many organizations are experiencing because of an inability to realize significant benefits from investments in data management and data science.
The risks of CDOs saying they aren’t ‘ready’ to provide business value from AI and data science are real and increasingly unwelcome from companies looking for more immediate returns from their data teams. This is producing an immediate need for CDOs to be unabashedly lean, with an unrelenting focus on a very limited number of specific business needs and outcomes.