All human beings have feelings. It may seem obvious, but organizations are composed of humans that need to have their feelings recognized and understood. That’s where empathy comes in!
Generally speaking, empathy is the ability to understand the feelings of others. Since communicating empathy is paramount to maintaining healthy interpersonal relationships, it should be no surprise that it’s equally important in the workplace. Studies accordingly show that empathy is positively correlated with job performance: particularly, managers who demonstrate more empathy are perceived to be better performers by their bosses.
Measuring Empathy
Given these insights, we began a push to incorporate empathy into our product. We specifically trained a natural language processing model (adopted by research at University of Washington) that makes inferences about the user’s expressed empathy based on their digital communications. It measures empathy in three categories:
- Communication of understanding. The user demonstrates that they meaningfully interpret feelings and experiences of who they’re speaking to, e.g. “I definitely see what you mean.”
- Exploring further. The user asks follow up questions and potentially touches on unstated feelings e.g. “What different approaches have you tried?”
- Expressing compassion and warmth. The user shows emotion at the appropriate times, depending on the context of the conversation e.g. “I’m sorry you’re feeling like that.”
How We Cultivate Empathy
Leveraging our new empathy model will allow us to make impactful updates to our product, such as:
- Providing users real-time analytics on how their empathy breaks down by category e.g. how much they communicate understanding vs. how much they express compassion
- Giving users periodic nudges to commend when they express empathy to promote positive reinforcement for that behavior
- Creating a tool to allow users to understand how to be empathetic in digital conversation by using our AI to rephrase their responses
We strongly believe that these product features will empower our users – namely, make them more self-aware and encourage them to communicate with more empathy in the workplace.
Raj is a Machine Learning Engineer at Cultivate. While he loves all things machine learning, he is particularly fascinated by its applications in modeling human behavior. He accordingly loves using his engineering and quantitative skills to make impactful AI products.
