Description:

Our words reveal our personalities, values, needs, and biases—as well as those of our culture, time period, and other contexts. The way we linguistically encode our values and needs affects how listeners and readers perceive us, and how they react. By the same token, the better we can anticipate, model, and accommodate the people we’d like to reach with our language, the more likely we are to be able to convey our intended message to that audience. Moreover, the language that best expresses our values and targets our audiences differs by domain and changes over time.

These insights have direct applications in business; we can quantitatively measure the impact of contextually effective language on business outcomes. What would happen if a customer who calls into a call center were paired with an agent who complemented their communication style? What would happen if a hiring manager or recruiter writing a job posting or recruiting email could anticipate how their linguistic choices would affect who might respond? What if businesses could understand trends in linguistic usage and contextualize what patterns are prevalent in—or affect—which subdomains? In this talk, I’ll discuss these questions in the context of data products I’ve had the opportunity to work on that harness the linguistic representation of self, organization, and/or audience; contextualize usage in time and domain; and measurably affect business outcomes.