Every morning as I get into work, I check our Slack KPIs channel to see how many of our Crisis Counselors (CCs) churned yesterday (28 days without being being active on our platform). On days this number is above average, our data team typically gets asked, “What happened?” A quick look at our data could tell us the average tenure of these CCs, the cohorts they trained in, and their past activity are all strong predictors of churn.
These are useful insights, but if we recommended a set of actions to reduce churn solely based on these outputs, we’d miss part of the story. We’d miss that our CCs sometimes stop volunteering because they’re not sure if they’re having an impact with their texters. Or, they have a life event (like a wedding, school, new job) that impacts their ability to volunteer consistently. If we focused exclusively on the quantitative data on our volunteers, we’d miss a central component of understanding our users. When you don’t understand your users, there’s a high chance you take the wrong actions based on your data.
So, how do you better understand your users? Pick up the phone and give them a call.
“I had two imminent risk texters my last shift. Both of them stop responding after a bit. It was heartbreaking not knowing if I had helped them through their crisis.” – Level 1 CC
At Crisis Text Line, we’re experts in empathy. We teach empathy to our volunteer Crisis Counselors (CCs) to prepare them for conversations with people in crisis. We foster empathy to grow rapidly while providing a high-quality, consistent service. And, we embrace empathy internally to manage vicarious trauma among staff and volunteers.
In my 3+ years at Crisis Text Line, I’ve received a crash course in empathy. It’s transformed the way I operate as a Data Scientist.
In school, we learn that data science relies on math, statistics, and linear algebra. These disciplines speak the language of numbers and variables. We use them to train our models, indicate the significance of our tests, and operationalize business metrics. There’s a certainty to hard numbers that’s comforting and actionable. Strong data scientists acknowledge this strength but also embrace that behind every hard number is a user, is a behavior, is a feeling. Hard numbers are just one key element in crafting a larger data-informed narrative. An empathetic touch enables us to extract the other elements of this narrative and bring data to life.
At Crisis Text Line, empathetic data science means starting with our users. We take every opportunity to jump on the phone and interview our counselors to better understand their motivations and their pain points. We do this throughout a CC’s lifecycle and our Chief Data Scientist interviews every single CC who hits 2,000 conversations. Collecting qualitative feedback allows us to put ourselves in the shoes of our users to bring hard numbers, model outputs, and analyses to life.
And this is exactly what the data science world today needs.
This process of leading with empathy creates a north star to guide your work. It means that even when you encounter noisy data, you can recenter your efforts in the ground truth of your users. This north star can be understood by anyone in your organization – product, sales, tech, C-level – because it came from your users (when possible, actual quotes from your users are most powerful since they retain an authentic voice). With this sense of shared understanding, it’s easier to build alignment and work cross-functionally to address user problems.
Have you ever found a data insight you knew to be true, but no one acted on? It’s easy to blame the other person – they just wouldn’t listen. But if you want to sell an idea, truth is only half of it. It’s up to you to weave a narrative that sells your truth, and the best way to get there is empathy. Next time you embark on a data investigation, the fundamental question to ask yourself is: “what do our users say about this question/pain point?”
Embrace empathy & happy collecting!