It’s Saturday night, you’re hungry and decide to order in. When the doorbell rings you find Tony Xu, CEO of DoorDash, holding an X-Large Pepperoni Pizza.
Why is the CEO of a $12.6B company delivering pizzas? He is keeping his finger to the pulse of the customer. He is seeing first hand how customers are interacting with his product.
This event, unknown to most, is recorded in a data warehouse along with millions of other DoorDash orders. On the other end, a team of DoorDash Data Scientists are mining it for insights.
“Send coupons for Domino's pizza to teenagers on Saturday nights”
“Improving our food recommendation engine will increase purchases rates by +20%”
Nearly all technology companies have similar teams -- and they continue to recruit highly paid experts to read and analyze data. Why? Valuable insights are hard to find.
Worse yet, the life of a data analyst is hard. Trying to mine data in today’s dynamic world of user behavior not easy.
Data tables, aka spreadsheets, are the primary tool of a data scientist. However, spreadsheets can be limiting. They are good at describing numbers, but bad at describing humans.
Not only that, but spreadsheets are a comfortable place for a data scientist. The rows line up, the data doesn’t talk back, and rarely does a spreadsheet ask them to talk in front of their peers.
It’s paramount that data analysts step away and understand who the customer is on the other side of the data. You do this through building Customer Empathy -- a data scientist’s ability to feel the customer.
Customer Empathy helps you understand the motivations of your customer, what state of mind they are in when using your product and where your product fits into their lives. Building empathy allows analysts to fill in the blanks between data points. It’s crucial to remove data roadblocks, or spark inspiration for new analysis.
How do your teams build customer empathy? Simple: Get away from the keyboard and go talk with the human behind your data.
Customer panels in the office for a fireside chat
Data Scientists joining sales call or customer success engagements
Analysts working a marketing booth
Sit-ins on user research sessions
Helping with support tickets
Even data scientists occasionally doing an alternative job with your product (Like Tony Xu delivering pizzas once a month)
Look for opportunities for your team to start talking to customers. Be grateful when a customer is willing to speak up. For my data teams I shoot for 8 hours of customer interaction a month per person. This one day a month multiplies their ability to do analysis and speak the language of the customer.
If Tony Xu can deliver pizzas, data scientists can stay calm and get to know the person on the other side of the spreadsheet.