October 11, 2017 | Marin Hitch
Lessons on Building Analytics Teams and Telling Stories with Data
Wharton Executive Education hosted its first-ever program dedicated entirely to customer analytics in September. The program, aptly named Customer Analytics, is designed to give participants an understanding of the best practices for data collection, the ability to forecast customer behavior, and the skills to translate data into meaningful results to leverage the revenue-producing power of analytics. As a part of the course, Wharton Professor and WCAI Faculty Co-Director Raghu Iyengar moderated a panel discussion featuring analytics leaders from Vanguard and URBN Inc.
As companies adapt to a new normal of data-driven business, it can be a challenge not only to hire the right people but also to figure out how to organize a data warehouse and implement an analytics strategy. To provide some real-world perspective, Kim Gallagher, Director of Marketing and Analytics at Anthropologie, and Douglass Stewart, Head of Client Analytics and Personalization at Vanguard came to campus to share their experiences and advice.
The sequence of steps taken by an organization when investing in analytics is crucial. It’s easy to become enamored with products that promise to be an all-in-one solution before taking a comprehensive look at the existing data. But without a clear understanding of current data assets, these products may not work the way they are intended to, causing more confusion and delays in progress. Further, it’s hard to organize a team of analytics professionals around messy data.
"What does analytics need to do for this business?"
A crucial first step is to understand what purpose the analytics solution will serve. Analytics functions within a finance company (e.g. quantitative trading) are different than those for a tech startup (e.g. developing recommendation algorithms). Organizations should have a clearly defined thesis that answers, “what does analytics need to do for this business?”
For a business at this stage in its analytics journey, Stewart advocated hiring someone who has worked with messy data suggesting, “get someone who has done it before so they can answer the technical questions.” While he doesn’t look for top-tier Ph.D.’s, he does look for people that have had titles like Data Engineer or who have demonstrated technical skills like SQL or Python. For URBN, they not only need to find good analytics talent but also candidates who feel comfortable in a culture that prioritizes the creative process. Though both companies have different applications of analytics both panelists agreed that finding someone who can tell a cohesive story using insights from their data is key; it’s often how you gain and keep the interest of stakeholders.
“Get someone who has done it before so they can answer the technical questions.”
Gallagher and Stewart also agreed on how to structure analytics teams within organizations. Ideally, customer analytics would live in the marketing department and would report to a central Chief Data Officer who would also oversee other dedicated analytics teams such as people or organizational analytics.
Regardless of business context, or the nature of an organization’s data, the clear take away from the discussion was that if a company is looking to survive and thrive today, organizing their data and data teams is a vital first step.