Trends can be dangerous to follow, but they can also be important to latch onto so business leaders — and their organizations — don’t get left behind. These conflicting statements are exactly why my colleagues and I track trends and report on the good, the bad and the ugly. Here are four predictions we have for human resources analytics in 2018.
- Continuous Listening Can’t Continue
Having your finger on the pulse of what employees think sounds great in theory. Why survey only once or twice a year if you can survey employees all the time? Measuring more often is not a strategy. Also, do you really think your employees want to take more surveys? As such, I predict that this approach won’t last, especially if limited action is taken on the results. Companies that continue with pulse surveys will quickly see the damage.
On the flip side, these companies will learn to turn their focus to the right approach — one that uses surveys to impact business objectives and measures the results. The bad news is that HR and the survey process will be collateral damage of companies abusing the survey process.
Continuous listening can’t last. A client of mine insisted on conducting monthly random pulse surveys, which went against my firm’s recommendations. Their response rates plummeted, and the value of the random pulse surveys greatly diminished. The client met the need of the organization to fill in a box on their monthly scorecard, but it greatly limited what the results told them and what they could do with the data. Now, the client is moving back to a specific survey strategy with a census survey event accompanied by a few targeted pulse surveys with a directed objective (e.g., low-performing leaders).
- The Circle of Life, Employee Life
Many organizations monitor the employee at different times during the their tenure but keep the data in silos (and honestly, don’t ever really do much with said data). I predict a shift: organizations will move toward measuring the entire employee life cycle — from pre-hire to exit. The opportunity exists to build a cohesive measurement strategy to assess the different phases of the life cycle together. This requires an integration of assessment content and a strategy for how to use and harvest the intelligence from this data.
For example, organizations could measure the employee life cycle to reduce voluntary turnover, an activity that provides a real return on investment to the organization. Each phase of the life cycle can provide unique insights and value to the organization regarding voluntary turnover. Several survey firms are already moving in this direction.
Proceed with caution, though, because this approach must be done correctly. Done poorly, it will just be more data with limited value.
- HR Analytics Are Here to Stay
Most HR professionals have acquiesced that predictive analytics is not going away. The majority of midsized to large organizations are trying to invest in and build substantial HR analytics capabilities. As such, these organizations are working through several issues as they build out capabilities:
Internal or External: Organizations will continue to grapple with the “buy” or “build” approach. Many are doing both, working with external partners and building internal capabilities. A good partner with this approach will focus a significant amount of time and resources to training internal resources.
The Struggle Is Real: From data warehousing and data integration to data reporting, basic analytics and predictive analytics, there is a ton to consider when integrating HR analytics into the department’s projects. This will be difficult for HR departments to prioritize.
Technology Traps: You’ve heard about dozens of new technologies coming to the market that claim to leverage analytics, machine learning, algorithms, etc. I predict that several organizations will fall into the shiny-new-object trap and purchase a technology that doesn’t provide any value to the company. Said companies will be hesitant to invest in other (potentially valuable) HR technologies in the future.
Track Attack: Many organizations will continue to set out to “do” HR analytics at their organization, but end up solely creating tracking dashboards. This is simply not analytics. The CEO will be less than impressed by a dashboard when they ask for the results of HR analytics investments.
The organizations that journey down the long, complicated path of leveraging the power of HR analytics will have the potential to benefit greatly. Those without a strategy will not be so lucky.
- Machine Learning and Artificial Intelligence in HR: The Time Is Not Now
HR is probably years — and maybe even decades — from real machine learning and AI drastically impacting day-to-day operations, so I predict that it will be just hype for a long time due to huge barriers. The first hurdle is that predicting human behavior and/or performance is difficult and complex. Letting a computer make these HR decisions, such as which candidate to hire, has danger written all over it. Just trying to put employees into specific buckets (e.g., race, age, gender) might have absolutely nothing to do with their performance, intent to turn over, etc.
The other huge hurdle is that AI doesn’t apply context well. Don’t underestimate the power of human judgment, and consider how hard it is to replicate that with AI (at least so far). For example, AI was used to judge a beauty contest. A robot panel judged faces based on algorithms that evaluated the “criteria linked to perception of human beauty and health.” The results were considered racist, as all the winners were white. Imagine making an employment decision using a similar approach. Insert lawsuit and damaged company reputation. Adverse impact and discrimination can happen when algorithms point to conclusions based on data alone, without critical context.
Shane Douthitt is managing partner and co-founder at SMD. To comment, email firstname.lastname@example.org.