Beyond Engagement: Measurement for a New Era of Work
The realities of work in 2023 have altered the types of insights and intelligence that organizations and HR professionals need about their employees, employee wellbeing, and the employee experience. With today’s work-from-home and hybrid work environments, coupled with ongoing diversity, equity and inclusion (DEI) initiatives, organizations and HR professionals require more nuanced information than ever before, and the ability to access these insights has significant downstream implications for the organization.
Engagement still matters, but there's more
Engagement has been used as a key workforce health indicator for decades. Research by HBR, Gallup and others show that “70% of respondents rank employee engagement as very important to achieving overall organizational success.” Research shows that companies with highly engaged employees deliver higher Earnings Per Share (EPS), lower Cost of Goods Sold, (CGS) stronger revenues, fewer safety incidents, less theft, lower turnover, and less absenteeism. The list of benefits goes on and on. But today, nuanced insights are also needed into the effectiveness of remote and hybrid work strategies, the wellbeing of employees, and the effectiveness of DEI initiatives.
One significant challenge though, is that it's difficult for employers to survey employees about wellbeing, stress, burnout, and a multitude of sensitive but critically important topics that have significant implications on absenteeism, and team and organizational performance. There is, however, a novel way to solve this problem.
New insights from existing unstructured data
It is possible to gain insights into workforce wellbeing and DEI experiences without asking employees -- and the open-end responses you’ve already collected from your existing engagement surveys are the surprising source of this information. Yes, the responses you’ve previously collected from questions you’ve already surveyed -- even if they weren't questions about wellbeing or DEI-related matters -- still contain an abundance of signals about workforce wellbeing and their DEI experiences. We'll explain how this works, but first a quick explaination of the unique natural language processing technique that will be used:
Stop words and why they matter
Traditional natural language processing techniques disregard almost 50% of the language contained in your surveys. It's part of the data cleansing process, and it's because traditional natural language processing techniques ignore “stop words.” Why?
Stop words are prepositions and pronouns -they're ignored because they don’t have any purpose in topic modelling or sentiment analysis - because prepositions and pronouns provide no context. However, by analyzing language from a psychological perspective by looking patterns in stop word use, psychological signals emerge about the people who use them. For example, people who are stressed use stop words very differently than people who are not stressed.
Language-psychology-based text analytics methods focus on the psychological signals that emerge from the ways that people use stop words, and make it possible to understand levels of stress, plus a wide range of other important psychological signals from people’s language.
Insights into wellbeing and DEI initiatives without asking
By aggregating open-end responses by departments, teams, demographics, etc., and analyzing them using psychology-based text analytics, it is possible to uncover a wealth of new signals from the data you've already collected – including insights about the wellbeing of teams and departments, stress levels, indicators or burnout, insights into the success of diversity, equity and inclusion initiatives -- highly-actionable information that isn't available from traditional sentiment and topic-based text analysis of engagement survey open-ends.
This means that it's possible to extract specific insights into team, department, group or organizational wellbeing and DEI-related matters without asking questions about wellbeing or DEI, and without changing the questions you ask -- simply by analyzing the open-end language data you’ve already collected.
The good news is that it's surprisingly easy to do. Check out this simple how-to-guide that explains how to measure team wellbeing from the survey text data you've already collected. For data scientists, the guide also includes R code so you can generate new insights right away.