In January, I had the pleasure of attending the Chief Analytics Officer Forum in New York City. The event featured talented leaders from a wide range of industries and companies. There were a number of common themes that emerged, and it was exciting for me to see that Trupanion is aligned with many of these best practices. I’ve highlighted some of the themes below that are useful for any person or organization building out an analytics team.
1. It’s important for an organization to understand the difference between descriptive and prescriptive information.
Our analytics department was born from our business intelligence department, which seemed to be a very typical progression among the other leaders I met. Business Intelligence implies the timely delivery of good metrics which inform business decisions and performance. That is an essential first step toward creating a data-driven business. However, that isn’t good enough anymore. Businesses who are leading the charge are not only reporting on their metrics but also using advanced techniques to find opportunities in their data to improve their operations in customer acquisition, customer retention, and expense reductions. Too often, business leaders don’t understand the difference between reporting and analysis.
2. It’s important to have a centralized analytics team.
At Trupanion, early on, most of the data needs were run through a single person, your humble author. This was a great benefit as I, and later the business intelligence team, was deeply integrated into the operations of the business and maintain data integrity and consistency between departments. The downside, of course, is those resources became a bottleneck as the department’s resources couldn’t grow as fast as the company’s appetite for data. To resolve that tension, we emphasize educating the rest of the organization on how to acquire and interpret the information relevant to their operations. As our company has grown, we have also improved our ability across the organization to interpret and react to data. While building that skill set across the organization, we maintain a centralized team which supervises those activities, maintains data consistency, and provides a more advanced toolbox for studying data when it’s called for. Many of the companies at the CAO Forum had a similar structure.
3. It’s critical to focus on the questions being asked.
Too often “Big Data,” “Business Intelligence,” and other popular jargon used for describing data and analytics, are focused on finding a lot of answers that don’t necessarily respond to a business question. A company building out an analytics function must have the leader of that team embedded in the management of the company, and that leader must be focused on guiding the team to ask good questions, and provide the guidance for how to answer them. Ultimately, the Chief Analytics Officer’s job is to help the business make good data-backed decisions. Understanding, and often guiding, the decision-making process is essential for understanding if the analytics team is adding value.
A company building out an analytics function must have the leader of that team embedded in the management of the company, and that leader must be focused on guiding the team to ask good questions.
4. Create a culture of testing.
A common theme across many of the companies was they had embedded testing across their organizations. At the most extreme, companies shared that essentially nothing gets done without a documented test with approved methodology. This is a powerful cultural shift for many companies, but has the potential to significantly increase performance. Many tests shared at the conference, and in my own experience, have counterintuitive results, which reinforces the need to be constantly questioning and testing. At Trupanion, we are continuously testing, particularly in new acquisition channels, and the analytics team is critical to establishing those tests and evaluating the results.
5. Produce tangible results.
It was repeated by many that analytics teams must be able to demonstrate tangible value-add. We’re in the business of 3x+ returns on our efforts, not marginal wins. For the vast majority of businesses, the opportunity for adding value through the use of analytics, so analytics team leaders need to take responsibility for making sure their resources are deployed in the most effective way possible. This becomes especially challenging because operational departments tend to demand Business Intelligence and descriptive metrics, which distracts from working on the big projects that can provide significant value to the organization. We maintain this discipline by setting aside committed time for big projects, and continuing to encourage self-service around the organization which frees the team’s resources.
Many companies and individuals are trying to build out analytics and data teams within their organizations, and it’s important to hear what other companies are doing and understand best practices. Every company is a bit different, but you can benefit from hearing about the challenges and successes of others.
By: TJ Houk
TJ Houk is VP Analytics at Trupanion. Trupanion offers medical insurance for cats and dogs in the United States, Canada, and Puerto Rico, and integrates data throughout the company. TJ Houk has a background in finance and actuarial science, and has grown the Analytics team from a single person to 10 cross-discipline professionals as Trupanion has grown from upstart to category leader in medical insurance for cats and dogs. The Analytics Team is responsible for Business Intelligence, Actuarial Services, and Financial Planning & Analysis, and supports all departments across the company. TJ is passionate about building strong, analytical teams, and integrating insights from those teams across organizations.