Report: 73% of ML decision makers fear headwinds could hinder further ML investment

Capital One’s new research commissioned by Forrester Consulting reveals the top challenges, concerns and opportunities businesses face when they use machine learning (ML) to improve business performance across the enterprise.

At a time when organizations are increasingly investing in and prioritizing the deployment of ML, Capital One’s research shows that a majority of data management decision-makers face significant operational barriers that can hinder the deployment of ML , including transparency, traceability and explainability of data flows (73%) and breaking data silos between internal departments (41%).

“Companies see tremendous potential in applying machine learning, but are facing headwinds in their data,” said Dave Kang, SVP and chief of data insights at Capital One. “This can keep companies from seeing actionable insights, and perversely shy away from adopting and operationalizing ML solutions in the first place.”

Obstacles to machine learning data

Another major obstacle for data managers: breaking data silos. More than half (57%) believe that internal silos between data scientists and practitioners hinder ML implementations, and 38% say data silos across the organization and external data partners pose a challenge to ML maturity.

Other top challenges include:

• Working with large, diverse, cluttered datasets (36%)
• Difficulty translating academic models into deployable products (39%)
• Reduce risk of artificial intelligence (AI) (38%)

Despite these concerns, the data also shows that ML adoption continues to grow, with nearly 70% of executives planning to increase the use of ML in their organizations. Key priorities for ML implementation over the next three years include automated anomaly detection (40%), automatically receiving transparent application and infrastructure updates (39%) and meeting new regulations and privacy requirements for responsible and ethical AI ( 39%).

Believing in the promise of ML

The research shows that data management decision makers believe in the promise of AI/ML to grow their business, but to continue developing their ML applications, decision makers must overcome the silos between people and processes.

They also need to find better ways to turn academic models into deployable products to better illustrate ROI to executives. By leveraging partners with first-hand experience and remaining relentlessly focused on the business promise of ML, decision makers can prove key results of operationalizing ML such as efficiency, productivity, and improved customer experience (CX) for executive leadership.

Methodology

Commissioned by Capital One by Forrester Consulting, 150 data management decision makers in North America were surveyed about their organization’s ML goals, challenges and plans to operationalize ML.

Read the full report by Capital One and Forrester.