HR Analytics: To a hammer, everything is a nail
Spoiler alert: I’m a skeptic.
HR Analytics: What is it?
HR Analytics is also known as People Analytics, Workforce Analytics or Talent Analytics. (I’m already nervous about this because the only group of people who think we can accurately quantify “talent” are statistical analysts. AND “People” by and large are unpredictable. “Analytics” is the consolidation of a lot of data to make predictions. …but people aren’t predictable…okay?…so I’m skeptical. Are you skeptical too?
Definition: CornerstoneOnDemand defines HR Analytics as gathering data on employee efficiency and using the data to make relevant decisions for improvement.
According to techopedia, HR Analytics is correlating business data and people data to improve business outcomes.
What data?… and where do they get it?
Employee Engagement is a key factor in HR Analytics. Now we have to figure out what is meant by Employee Engagement. And I’m not doubly skeptical because, once again, “people engagement” is also unpredictable.
The Business Dictionary tells us that, at the heart of employee engagement is the emotional connection an employee feels toward his or her employer .
Who believes that we can measure an “emotional connection?” Lots of people, to a point. What point. Well it’s fairly easy to differentiate between someone who hates their job and someone who loves their job. But HR Analytics takes this to a new level of quantification…down to the smallest detail.
Here are some considerations that “measure” employee engagement. My comments follow the bullets.
- Does the employee “feel” mentally stimulated?
- Is there trust and communication between employees and management?
- Do employees understand how their work contributes to the company performance?
- Is there growth potential?
- Does the employee have a high sense of pride about his or her association with the company?
Do you see a problem with this? Sure, there are indicators of these considerations, however, there is also an assumption (a HUGE assumption!) that this is what will motivate all employees to top performance. It also assumes that when employees answer these questions, there is consistency from one day to the next.
How do companies get this data?
Did you guess? It’s from the infamous Employee Engagement Survey that companies send to all their employees. I bet you remember some of these questions.
- On a scale of 1 to 10, how happy are you at work?
- Would you refer someone to work here?
- Do you feel valued at work?
- Do you feel the management team is transparent?
- If you were given a chance, would you reapply to your current job?
- How frequently do you receive recognition from your manager?
- With eyes closed, can you recite our organization’s values?
- Do you have fun at work?
(Compliments to Sabrina Son on her article.)
Here are similar questions in another style that also asks employees to rank their responses. These are from Decision-Wisehttps:
- It is easy to become absorbed in my job.
- I would recommend the company as a great place to work.
- Most days, I look forward to coming to work.
- I feel like I belong here.
- I feel challenged and stretched in my job in a way that results in personal growth.
And on and on. There are numerous styles of questions.
What’s my problem?
I’m sure you “feel” my skepticism, and perhaps I’m missing something, but it “seems” to me that for every question asked, it depends on the day. It depends on when the person is asked. It depends on what has happened in their recent communications with their management, their colleagues, and their family or close relationships.
Garbage in –> garbage out.
Nevertheless, the data from these responses is used to make business decisions.
This sounds more like a deal with the devil as Neil Patrick so clearly describes in his post: HR data and analytics drives profits but at what cost?
Why do we need HR Analytics?
Data technologists say we need HR Analytics because people can’t make good decisions. (Actually, one person said we need it because people make “stupid decisions.” (Yes, that’s a quote.)
The basic premise is that people are bias.
What’s “bias?”
Bias: a preference or an inclination, especially one that inhibits impartial judgment [emphasis mine].
To get rid of bias, the HR Analytics Process does the following:
- Identify the appropriate databases where information can be obtained.
- Collect the data.
- Cleanse the data. (This could take another blog to discuss…)
- Analyze the data.
- Aggregate the data to define meaningful outcomes.
Information is kept in various databases in a company. The information about attendance and absenteeism may be in one place. However, the distance employees live from the office and/or their commute time, may be in a different database. Then comes their personal history with the company (recognition, promotions, etc) along with their performance ratings. Finally, we have the all-important Employee Engagement Survey.
WOW…a lot of work goes into this process! All of this to remove bias so we can stop making “stupid” decisions.
Do we need or really want to get rid of bias?
Is “impartial judgment” always best? I’m not so sure. Here are a few case studies from HR Analytics “experts.” Let’s see if “bias” is removed, and whether it is helpful to do so, or not.
Case Studies:
Pilots leave the company:
Question: Why were so many pilots leaving a law enforcement agency?
Process: HR Analytics.
Situation: Pilots fell into one of two groups: those who were commercially trained prior to being hired and those who were hired and then trained on the job.
Conclusion: Following the detailed HR Analytics process, it was determined that the commercial pilots left the company and returned to their former employment in the commercial airline industry because the compensation was much higher.
Really? All of that work to figure that out? Was this over-engineered?
Bias check: If the agency, moving forward, dismissed any candidate that came from the commercial airline industry, aren’t they being biased? Or are they making a smart business decision?
Ethnic Prejudice on Résumés:
Question: Did ethnic names on résumés draw a bias from hiring professionals as they evaluated candidates?
Process: HR Analytics.
Situation: A stack or résumés included “white” names and names from other ethnicities was passed around to the selection committee.
Conclusion: It was found that if “white” people reviewed the stack of resumes, they favored “white” names. It was found that people from other ethnicities did the same.
Bias check: Obviously, there are times when candidates are overlooked for the wrong reasons and less capable candidates are selected and candidates that are more qualified are dismissed. However, are there times when a less qualified candidate should be selected?
Example: Would we expect the National Black MBA Association (NBMBAA) to choose a non-black candidate as the Director of Marketing because his or her MBA was from a more prestigious university? Or perhaps the non-black candidate had 12 years of experience as compared to 8 years by a black candidate.
If the selection excluded people they thought were white (and looked it on LinkedIn?), aren’t they being biased?
Can we escape Bias? Should we? Do we want to?
Is it fair to say that there are consequences to making the decisions “by the numbers” when it comes to people?
Did you know that HR Analytics doesn’t belong in HR?
Uh huh. A particular company made the case that for the Analytics process to be interpreted accurately, the “interpreters” should be trained in analytics. And HR people are fluffy, shoot-from-the-hip types and not capable of this kind of analysis. …of course that company also offers courses in HR Analytics for …wait for it… those fluffy, shoot-from-the-hip HR professionals.
I’m still skeptical…and now I’m even cynical… and concerned
When I started this blog I was definitely skeptical. Now that I’ve spent a few hours with the material and wrestled with it, paced up and down in my living room, and walked four miles with my pooch, I’m cynical as well.
Skeptical: doubtful.
Cynical: distrusting the motives.
Concerned: disturbed, troubled.
Skeptical: first of all, people are fickle, inconsistent, and unpredictable. If a company is committed to finding the right candidate, then they will need to spend some time to actually get to know their top candidates.
(Oh dear, wouldn’t that be costly? As costly as implementing an HR Analytics program?)
Cynical: Am I cynical in thinking that companies really want their employees to enjoy coming to work? (Aw…that’s sweet.)
OR
Is it that the leadership wants higher revenue and this is one way they think they can get it?
Am I cynical in thinking that this approach is disingenuous? Does motive matter? Is it possible that some people, perhaps many, hear the right words, (e.g. “Morgan, your work here is valued!”), and if the words feel disingenuous, or manipulative, that Morgan will not perform at peak.
Concerned: The marketing glitz surrounding this movement is like a loose tiger in a playground of toddlers. The marketing glitz is compelling companies to make many or most decisions based on data-analysis, predictive analytics, machine learning and Artificial Intelligence. At its core, it is inherently bias! The toddlers are smiling and saying, “Here kitty-kitty.”
NOTE: I use the term “toddlers” because machine learning and Artificial Intelligence are in their infancy and therefore, so are its users. Should I have used the term “infants?”
This does sounds more like a deal with the devil. Check out Neil Patricks blog, here it is again: HR data and analytics drives profits but at what cost?
…and remember…
To a hammer, everything is a nail.
To an Analytics engineer, everything can be solved by analytics.
What do you think?
HR Analytics: articles of interest
Engage Employees By Actually Listening To Them by James E Smith
HR Analytics should not be located in HR …wrong skill sets there.
https://mortenkamp.com/2014/11/11/5-reasons-hr-analytics-should-not-be-located-in-hr/
Engagement Survey Blogs:
https://www.tinypulse.com/blog/sk-employee-engagement-survey-questions
https://blog.cultureamp.com/employee-engagement-survey-questions#your_survey
Misconceptions about Employee Engagement Surveys
http://www.workshopexercises.com/Engagement_misconceptions.htm
Perhaps someday this will make sense, for now, it’s marketing hype and HR grasping for straws. Here are a few things to consider. Very few people in HR are classically trained psychologists yet they are attempting to alter human nature. The drop off of experienced managers, people have managed diverse business groups long enough to have become experienced managers, below the CHRO is significant, yet these are the people someone is expecting to train and coach managers throughout the enterprise. The real problem with engagement surveys’ abysmal results has nothing whatsoever to do with the data analyzed or otherwise.
Management has been ignoring perfectly good valid employee data for decades for no other reason than management convenience. Beyond HR deciding whether more art is needed on the walls, different color paint, the menu in the break room, a change in benefits etc. management will not just wake up one morning and decide to more aggressively act upon the data because it came from a software program.
Here’s what a single 9-word survey question generated with a different approach. The CEO was the one asking the question, a third-party reporting to the CEO was collecting the data and management, those folks who make up the culture, politics, and silos, was challenged to justify NOT taking action. Blockers who did not comply, even officers, were presented with an opportunity to explain at the CEO’s staff meeting. Only one did that in twenty years.
The next major difference was that everytime a decision was made it was announced to the employees within two days. Here’s where the EE/EX industry gets it all wrong trying to be everything to everybody. With the approach described below, HR had no involvement and if only one person mentioned an issue, that was enough, it was on the fix list.
The employees experienced management making decisions nearly every day, no HR committee taking weeks or months to take action. Every major announcement raised the participation level, one announcement alone driving a one week increase of 3900. Without execution, the data is the least impactful aspect of engagement.
At this point, AI should be considered a weapon of math destruction. https://chiefexecutive.net/employee-engagement-ceos-actually-listening/