The Fundamental Challenges of Employee Feedback Data

The organizational development world has long debated the best way to ask questions to get at the heart of an employee’s experience, in an effort to impact the success of their people and their business. Because quantitative data can be easily standardized and digested in high volumes, rating questions have largely dominated employee surveys and feedback forms.  Many companies recognize the value of quantitative feedback, focusing on new ways to display, process, and link it to employee and business outcomes.

When it comes to truly understanding what’s on your people’s minds, however, rating questions by definition provide a limited view because they can only measure their original intent. To gain deeper insights into the reasons behind a troubling rating, people teams, executives, and managers have to look to employee comments. Beyond open-ended questions on surveys, many companies also engage in on-going conversations with their employees through multiple channels like email, chat, and social networks. The proliferation of feedback — combined with the rich insights contained within individual comments themselves — has made it nearly impossible to process comments in the aggregate. Many people teams wind up analyzing only a small sample of comment data, or spend weeks manually coding comments one by one. As a result, important patterns, trends, and insights get lost — and human subjectivity is introduced into the process, biasing results.

Introducing Glint Narrative Intelligence

Today, we announced Glint Narrative Intelligence™, a next-generation natural language processing (NLP) engine that is designed to help organizations quickly make sense of qualitative feedback — while connecting this feedback to engagement and other outcomes.  Narrative Intelligence combines AI-for-HR™  technology with modern organizational development science to extract highly relevant topics and sentiment — as well as highly valuable contextual information — from each comment in real time. These insights are then aggregated and presented in an interactive format to enable a deep understanding of recurring themes, surrounding context, and how sentiment and importance around these evolve over time.

“Narrative Intelligence is exactly what I was looking for but didn’t know I needed. With this leading-edge technology applied to employee comments, understanding what our people are saying is easier and faster.”

— Ronan Cosme, Learning & Organizational Development, SS&C Advent

It all starts with our proprietary and cutting-edge machine learning algorithms (Glint AI-for-HR™), which have been trained on millions of responses across the entire employee experience. This enables Narrative Intelligence to infer sentiment and categorize underlying patterns in the data into topics based on Glint’s multi-level organizational development taxonomy, enabling a consistent, thematic understanding of employee feedback. Our algorithms and taxonomy continually learn and adapt as new data enters the system — and with emerging industry trends — resulting in industry-leading precision (>85%).

Next, we layer in context to make topics more concrete and actionable. For example, recognition is a commonly mentioned topic, but it can mean different things to different people. Depending on the organization or team, Recognition might be connected to a number of other topics, such as Manager Feedback, Company All-Hands meeting, Pay, or Bonus. The topic of Recognition alone is not informative, but knowing that when people talk about it, they are doing so in the context of Manager Feedback and sentiment has been trending negatively over time, helps you truly understand what your team means and gives you the basis for taking effective action on that feedback. We provide this context in an easy to understand visualization to save you the time and effort it takes to manually combine all the possible topic combinations in filters.

Depending on the size of your team or company, drilling into a connected topic set like Recognition and Manager Feedback may still result in an overwhelming volume of comments. Narrative Intelligence provides additional insights that summarize the comments in the form of keywords and phrases. This lets you hear how employees talk about topics and more in their own words and shows you the relative importance and sentiment of each word and phrase.

One of the key benefits of Narrative Intelligence is its ability to help you zoom in and zoom out to the appropriate level of insight. Once you’ve identified a set of connected topics or keywords to explore, you can drill into the individual comments associated with them to access the relevant insights you need. Most of our users are interested in hearing employee suggestions when deciding what actions to take. Narrative Intelligence automatically isolates prescriptive comments, allowing you to focus on actionable feedback without the work of going through every comment.

Each organization and team within it has a unique formula for people success. Narrative Intelligence allows you to explore employee feedback by specific populations and through the lens of outcomes like engagement, performance, or attrition risk. Now you can efficiently compare what high performers talk about versus low performers, how different departments think about Recognition using the topic visualization, and combine multiple filters to explore what’s working well — and what needs improvement — for a specific group such as women in engineering who have been at the company for more than five years.

Tremendous Progress to Date…and a Look Ahead

We are already seeing Narrative Intelligence have a big impact on the way our customers gather and interpret feedback. Leading companies across a number of industries are now identifying engagement themes in the comments that weren’t being measured before. As our customers find these unexpected points of interest, they are able to craft new items in subsequent pulses to further understand an area where employees express a high degree of interest. This allows our customers to keep pulses shorter, while still measuring what matters most to their people.

Narrative Intelligence and our machine learning team will continue to push the boundaries of what is possible in NLP. We’re currently conducting research into how various topics impact outcomes, in order to provide a unique topic outcome impact measure per population. Glint customers will also be excited to know that we will be providing integrations with other systems such as email, Slack, and Glassdoor to help them extract meaning from open-ended feedback through Narrative Intelligence. Finally, we are establishing a program to enable companies who are yet not on the Glint platform to experience the power of Narrative Intelligence for themselves. Starting today, they will be able to analyze responses from their existing engagement surveys and pulses, onboarding feedback, exit interviews, 180s, and other feedback channels using our Narrative Intelligence standalone offering.

With Narrative Intelligence, you now have a platform to encourage ongoing feedback, while still being able to make sense of the data and enable effective action. We have already been amazed by how our customers are using Narrative Intelligence to drive meaningful change in their organizations, and we are excited to empower them on their journeys with Glint’s people success platform.