Data showing a person The mostpowerful AI and sentiment analysis tools can recognize hundreds ofhuman emotions (confusion, sarcasm) and themes (work-life balance,development. (Image: Shutterstock)

|

Employee benefits are complex and expensive, typically one ofthe top three expenses for employers. Yet,companies often make costly benefits decisions without meaningful data on what their employeesactually want or need. According to SHRM's 2018 Employee Benefits Report, more than onethird of respondents increased benefits offerings in the last year,likely without the right information at hand.

|

Increasing the benefits available to your employees may soundlike a smart move in general. After all, what person doesn't wantmore everything? But it's crucial that businesses consider whethertheir benefits offerings actually meet their employees' specific needs, and deliver the impactintended. Often, adding extra benefits or new plans to existingpackages leads to employee confusion and wasted organizationalresources, if the new options are underutilized or perceived asduplicative.

|

Related: How do millennials make benefits enrollmentdecisions?

|

The good news is that new technologies, including artificial intelligence (AI) and machinelearning, are primed to play a leading role in the future ofbenefits: advances in these areas will help HR and benefits leaderstruly understand, and efficiently meet, the needs of theirpeople.

|

Understanding your employees' needs

There's often a disconnect between employers and employees whenit comes to benefits offerings. While 78 percent of employers believe their employeesare satisfied with their current benefits offerings, only 52percent of employees report being satisfied! Through machinelearning, examining any organization's historical data, employeescan have real-time, automated recommendations into what plans wouldbest meet their needs at their current life stage. And when moreinformation is needed, AI-enabled chatbots could allow employees to asknatural-language questions directly of benefits providers, andinstantly receive targeted recommendations based on theconversation.

|

Another example is integrating AI with company surveys to betterunderstand the comments in open-ended questions, providingorganizations with instant, rich insight into employees' wants andneeds. Open-ended surveys (vs. simple multiple choice) allow formore meaningful responses, but traditional surveys can take monthsto administer, analyze, and act upon. AI-enabled survey solutions,by contrast, allow HR to immediately gather actionable intelligenceto make needed changes before they become problems.

|

The most powerful AI and sentiment analysis tools can recognizehundreds of human emotions (confusion, sarcasm) and themes(work-life balance, development), using the power of deep learningon millions of similar documents to classify and summarizeopen-ended comments instantly, and with greater accuracy than aperson could. The analysis of open-ended data provides specificinformation about how employees actually feel and can dramaticallyimprove the relationship between an organization, its managers, andits employees.

|

Sentiment analysis application

The HR team at Steel and Pipe Supply Companies Inc. (SPS) inManhattan, Kansas, is one example of an organization using thepower of AI and sentiment analysis to create meaningful change. SPSsought an intelligent survey solution to help engage theiremployees, and streamlining benefits was one of the first businessproblems they solved with sentiment analysis.

|

The team previously administered its Leadership and Engagementsurvey to 600 employees using a simple online survey platform, butthe process was cumbersome. More than half of SPS's employeesworked on a warehouse or manufacturing floor, so surveys had to becompleted on paper. Not only did this increase the chances ofaccidental errors, it also slowed down the process—fromsynthesizing the data to compiling reports and actions—to a crawl,making any actions taken after the survey significantly lesstimely, which eroded employee trust in management and HR.

|

However, after using a human capital management (HCM) platformwith AI and sentiment analysis as a core capability, SPS was ableto get a better sense of how its employees felt about certainbenefits offerings, where there were opportunities to shiftresources to better meet needs (not just “more!”), and ultimatelystreamline plans based on the preferences revealed by thesurvey—faster. By tasking AI with analyzing responses and reportingback on the tone and themes of open-ended responses, what used torequire six weeks of challenging work reviewing surveys by hand,collapsed to just a few days in total, with immediate insights, andactionable results. Previously, SPS had eight varying plansavailable for critical-illness benefits, which were difficult toexplain and implement on both front and back ends.

|

By giving intelligent surveys to employees using these existingbenefits, SPS understood what was actually important to them, andwhere similar or unnecessary options could be eliminated. Surveyresults were incredibly clear on preferences around two primaryplans, so the company was able to reduce eight plans down to two.Additionally, the cost savings on both the administrative andbenefits planning allowed SPS to launch SPS Thrive, a new HR-ledwellness initiative offering nutritional, physical, emotional, andfinancial support to employees—something also discovered onemployees' wish lists.

|

Humanizing AI for results

AI has great potential in the benefits space, but it should beused to assist, not replace, human analysis in helpingorganizations make better decisions. Understanding people startswith listening to what they have to say, and AI can help hear everyvoice and read between the lines to get at what matters most.

|

Ultimately, technology must find more natural ways of workingwith people in order to fulfill its potential as our quintessentialpartner. It needs to speak our language and accommodate all ourcomplexities—not the other way around. Asking survey respondents torank questions on scales of one to five is an example of humansevolving their speech to fit a computer's language. However,providing simple, open-ended surveys and using AI and sentimentanalysis to understand how employees feel about their experiencebrings technology to our “home court.”

|

The purpose of AI in the HR field should never be to replacepeople, but to make it easier for HR leaders and managers to makesmarter decisions with ultra-personalized, real-time insight. In HRand benefits management, that means providing them with the toolsto treat their people like people.


Armen Berjikly serves assenior director of growth strategy at Ultimate Software, where hisexpertise in human-computer interaction drives Ultimate'stransformative AI platform and direction.

Complete your profile to continue reading and get FREE access to BenefitsPRO, part of your ALM digital membership.

  • Critical BenefitsPRO information including cutting edge post-reform success strategies, access to educational webcasts and videos, resources from industry leaders, and informative Newsletters.
  • Exclusive discounts on ALM, BenefitsPRO magazine and BenefitsPRO.com events
  • Access to other award-winning ALM websites including ThinkAdvisor.com and Law.com
NOT FOR REPRINT

© 2024 ALM Global, LLC, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to [email protected]. For more information visit Asset & Logo Licensing.