CNA Insurance Moves Rapidly from Data Foundation to Analytic Products on Google Cloud

About CNA

CNA is one of the largest U.S. commercial property and casualty insurance companies. The Chicago-based company provides a broad range of insurance products and services for businesses and professionals in the United States, Canada, and Europe.

Industries: Financial Services & Insurance

Location: United States

Products: Google Cloud, BigQuery, Vertex AI, Looker

Tell us your challenge. We're here to help.

About Accenture

Accenture is one of the world’s largest professional services companies, with specialties in digital transformation, cloud services, data management optimization, and IT modernization.

CNA worked with Google Cloud to build a strong data-analytics foundation by consolidating global data, using automation to accelerate time to market for machine learning models, and establishing a path for developing analytic products to enable faster decision making.

CNA is one of the largest U.S. commercial property and casualty insurance companies. Backed by more than 120 years of experience, CNA provides a broad range of standard and specialized insurance products and services for businesses and professionals in the United States, Canada, and Europe.

During the past several years CNA has experienced substantial growth, with data providing a strong focal point across the company as executive management, business unit leaders, and independent brokers looked for more insights to help create and sell new offerings. The company invested in machine-learning (ML) capabilities to improve their predictive insights, but the process of creating and deploying ML models was a challenge.

To increase its already strong market position and competitiveness, CNA decided to work with Google Cloud and Accenture, a Google Cloud Premier Partner. The goal was to build a far more efficient analytic lifecycle, consolidate data sources across the globe, and provide a single pane of glass for decision makers while maintaining rigorous data security and management practices.

Migrating and consolidating global data on Google BigQuery 2x faster than expected

Over the years, CNA had compiled hundreds of data sources covering policies, claims, and more across domestic and international business units. This created business challenges for the company that were further complicated by a mix of legacy applications and systems used to store and manage data.

CNA’s first action was to consolidate its segmented data by building a global data lake on Google Cloud BigQuery, a highly scalable and fully managed cloud data warehouse.

The CNA team evaluated two approaches to migrating the data to GCP. The faster, easier path would have been a lift-and-shift of its existing data to Google Cloud. However, the benefits of moving a 20-year-old data warehouse to the cloud would have minimal benefits.

To achieve a greater impact on business outcomes while building a scalable data foundation for the future, CNA decided to establish new data pipelines by “true sourcing” data in real time. Initially, the team estimated that this migration approach could take up to three years. However, working with Google Cloud engineers and Accenture, CNA engineers built reusable data-migration patterns and leveraged automation to compress the migration timelines by nearly 50 percent.

"Whether it was raw or curated data, the Google Cloud team and BigQuery really helped us consolidate and leverage the horsepower of Google’s data cloud to stitch our data together into a global-360 view," says Santosh Bardwaj, Senior Vice President and Global Chief Data and Analytics Officer at CNA. "Every few minutes, new real-time data feeds land in BigQuery. This is a radical improvement from the daily and weekly data loads that CNA did previously."

As part of the data migration, CNA also implemented proper data governance and metadata practices at the onset of the project. "With solid data governance and a comprehensive data glossary, we are on the way to reducing our reliance on tribal knowledge and replacing it with a self-service, knowledge repository," Bardwaj says.

“Whether it was raw or curated data, the Google Cloud team and BigQuery really helped us consolidate and leverage the horsepower of Google’s data cloud to stitch our data together into a global-360 view. Every few minutes, new real-time data feeds land in BigQuery. This is a radical improvement from the daily and weekly data loads that CNA did previously."

—Santosh Bardwaj, Senior Vice President and Global Chief Data and Analytics Officer, CNA

Building the CNA Model Factory in record time with Vertex AI

The second element of CNA’s work with Google Cloud was improving its time to market to deploy insights using machine learning models.

"We were developing high-end ML models, but the deployment was taking many months or longer, and it was very manual and expensive," Bardwaj says. "It slowed our time to market, and was stifling a lot of ideas because people didn’t want to commit so much time to the process."

CNA decided to turn to Google Cloud’s broad expertise in machine learning, analytics, and big data. "Data analytics and machine learning are big Google Cloud strengths," Bardwaj notes, "and we wanted to capitalize on that."

In just 10 weeks, CNA Data science engineers worked with the Google Cloud team and Accenture to develop a prototype for the new CNA Model Factory built on Google Vertex AI, which enables companies to build, deploy, and scale ML models quickly with both pre-trained and custom tooling.

Using this comprehensive and automated ML platform, CNA was able to deploy multiple forecasting models in hours, a process that previously took many days or longer. The company expects that the Model Factory will dramatically accelerate its use of machine learning across many business operations. By modeling forecasts faster, CNA has a more up-to-date picture of key performance indicators such as revenue streams and cash reserves.

"We really showcased the scale and ability to accelerate both ML model development and deployment. We are partnering with Google to continue maturing the Model factory into an end-to-end assembly line."

—Santosh Bardwaj, Senior Vice President and Global Chief Data and Analytics Officer, CNA

Driving business value with analytic products

The rapid build out of CNA’s data foundation on Google Cloud is helping CNA pivot to the next phase of its transformation, which includes providing value-added analytic products for business units. A long term goal is to give CNA executives, underwriters, and brokers fast and easy access to clear, accurate data for better decision making.

"We want to create a single pane of glass as the universal source of truth to replace hundreds of local and segmented reports," Bardwaj says. "The goal is to build a set of robust, easily consumable sources of analytic insights and metrics to address critical business queries like, ‘‘What happened?’, ‘Why did it happen?,’’ and ‘What will happen next?’ in a single view."

Using Looker, CNA’s analytic product teams are working closely with business product owners to launch a global 360-degree view of KPIs across the organization.

"Whether it’s the executive leadership team, functional heads, or branch managers, we want everyone to be able to access the same views and the same data to make decisions," Bardwaj says. "Using Looker dashboards, we can adopt a product-centric model where we work iteratively and develop insights quickly based on a continuous feedback cycle."