We're Celebrating 10 Days of SHRM! Today's Gift: $15 to Starbucks w/ a SHRM professional membership. Promo code 10DAYSBUCKS.
Training, policies and tools to help HR prevent and respond to harassment claims.
Is your employee handbook keeping up with the changing world of work? With SHRM's Employee Handbook Builder get peace of mind that your handbook is up-to-date.
Develop your HR competencies and knowledge in-person in 12 U.S. cities or virtually.
#SHRM18 will expand your perspective – on your organization, on your career, and on the way you approach HR. Join us in Chicago June 17-20, 2018
Tech advancement gauges how employees feel about work so employers can address their concerns
Organizations have generated unprecedented amounts of employee feedback through weekly or monthly pulse surveys, annual engagement surveys, and internal social networks and collaboration platforms.
But many still struggle with how to efficiently comb through that mountain of information to identify actionable insights leaders can use to improve employee engagement and retention.
Some companies are now turning to artificial intelligence (AI) tools to conduct sentiment analysis on employee feedback, gauge how employees feel and address their concerns.
While text analysis of survey responses isn't new, the emergence of smarter algorithms enables faster and more precise search and categorization of unstructured data, such as open-ended comments, said Alan Lepofsky, vice president and principal analyst with Constellation Research, a technology research firm in Silicon Valley.
Lepofsky, author of the recent report Why Artificial Intelligence Will Power the Future of Work, said vendors have made advances in sentiment analysis technology. And companies like Google and Microsoft have showcased new tools for text, voice and facial analysis at their user conferences.
The use of these tools extends beyond HR into areas like social media and the political realm. For example, a reporter for National Public Radio recently used sentiment analysis software to analyze the tweets of President Donald Trump to identify positive and negative sentiments.
It wasn't long ago that text analysis was limited to simple word count and topic extraction, but today's tools feature more complex and nuanced natural language processing (NLP), according to a 2017 report from Forrester Research, Artificial Intelligence Technologies and Solutions: It's Time to Put on Your Training Wheels. These AI-enhanced applications turn the unstructured data from engagement surveys into structured data that's easily analyzed. Key phrases and categories that exist in the text help gauge employee sentiment and add a richness and depth to behavioral analysis.
Companies have reaped the benefits from this improved functionality, Lepofsky said. "Are employees upset that you haven't upgraded your laptops in five years, that they can no longer work from home or that you've taken away free snacks in the cafeteria?" he said. "Advanced sentiment analysis can more quickly and easily identify those patterns and themes with less need for human intervention."
When IBM Corp. used its own sentiment analysis tools to analyze feedback from its employee communities and surveys, one dominant theme that emerged was dissatisfaction with computer choices in the workplace, Lepofsky said.
"As a result, IBM soon began making Mac computers available as a choice for employees, and many thousands took advantage of the offer," he said, a move that wasn't cost-prohibitive for IBM because of an earlier partnership with Apple.
'Providing context for the words that employees use is the holy grail of text analysis.'
Bob Schultz, general manager of IBM Talent Management Solutions in San Francisco, said that sentiment analysis software is part of IBM's talent insights product suite. "It sifts through employee survey data to find patterns and insights that are actionable for leaders," Schultz said. "Engagement scores go up when employees see that companies are actually doing something with their feedback."
Lepofsky said more advanced forms of sentiment analysis are industry-specific, and that means the software is trained to interpret and categorize the vernacular used in different fields. These tools also can be combined with other data to predict sentiment trends, such as a negative sentiment that's likely to increase.
"The words or phrases that a nurse, an engineer or a financial analyst use can have different meaning depending on the industry context, and newer employees also can have different sentiments and mindsets than veteran employees," Lepofsky said. "Providing context for the words that employees use is the holy grail of text analysis."
The Choice Is Yours
AI-enhanced sentiment analysis software is available in several employee engagement platforms and in the form of stand-alone tools. Glint, for example, is a platform with a natural language processing engine that synthesizes employees' open-ended feedback to identify core themes and feelings—even some issues not addressed by survey questions.
The application is designed to surface perceived strengths and weaknesses of organizations that survey designers may not think to ask about, and can aggregate those many open-ended responses into a visual map of key topics. Glint's "AI-for-HR" technology promises to guide leaders to the right course of action based on this interpreted feedback.
Employee engagement platform CultureAmp employs smarter algorithms to give users more actionable insights right from platform data or to conduct deeper dives into other employee surveys. TINYPulse, another engagement platform, uses advanced sentiment analysis tools and machine learning to adapt to specific organizational cultures and highlight key feedback trends for clients.
Sentiment Analysis of Job Candidates
New sentiment analysis tools have emerged that scrutinize job candidates' video interviews to use voice or facial analysis to assess honesty or personalities. While many of these applications have promise, AI analysts believe it's still too early for the technology. In the case of facial analysis, for example, accurate interpretation is often dependent on the uncertain quality of laptop cameras used by candidates to record interviews.
Elaine Orler, founder and CEO of Talent Function, a San Diego-based talent acquisition technology consultancy, has seen demos of vendor systems that analyze candidate voices for pitch, tone and other characteristics, as well as some systems that analyze candidate faces for tics or other indicators that might suggest dishonesty. But Orler said the most useful tool she's seen to date converts audio from candidates' recorded video interviews into easily searched text.
"Conversion of voice to searchable text can be highly valuable to recruiters," she said. "It allows you to go back in after video interviews to quickly find, for example, which candidates said they had C++ programming experience, then be taken right to the spot in the interview where that experience was talked about."
Dave Zielinski is a freelance business journalist in Minneapolis.
You have successfully saved this page as a bookmark.
Please confirm that you want to proceed with deleting bookmark.
You have successfully removed bookmark.
Please log in as a SHRM member before saving bookmarks.
Your session has expired. Please log in again before saving bookmarks.
Please purchase a SHRM membership before saving bookmarks.
An error has occurred
Recommended for you
Refer a Friend to SHRM
SHRM’s HR Vendor Directory contains over 3,200 companies