The Challenges in Building a Strong Function Around People Analytics
By Arun Chidambaram and Michael Housman
February 26, 2016
Over the past few years, People Analytics has really found its way.
Whether they’ve called it “People Analytics,” “Human Capital Analytics,” or “Workforce Science,” the fundamental phenomenon is the same: HR practitioners are relying less on gut instinct and more on data to make decision about the workforce.
It’s a long-overdue phenomenon that takes its cues from the sort of data-driven approaches to that we’ve seen transform a variety of other industries: marketing, finance and consumer credit, insurance, and even major league baseball.
If one were to take their cues solely from the popular press, they might be tempted to assume that most large organizations are pretty far along in their People Analytics journey. But Josh Bersin recently wrote an article for Forbes based on a survey of 480 different organizations and found some surprising results: Only 4 percent of companies have achieved the capability to perform “predictive analytics” about their workforce and only 14 percent have done any significant “statistical analysis” of employee data at all.
For the remaining 82 percent of companies out there, the good news is that the fruit isn’t even low-hanging; it’s sitting on the ground.
Here are five (5) lessons they should bear in mind if they wish to jump-start their People Analytics function.
1. Don’t try to eat the elephant
There is no question that there is a ton going on in People Analytics: New tools are constantly emerging, new techniques for modeling the data, and new trends around what issues to be focusing on. Neuroscience, machine learning, advanced infographics and visual modeling are becoming a reality in HR.
Here’s the bottom line: it’s easy to get overwhelmed, so, it’s critically important to focus on the pieces that matter and ignore the rest.
Every single People Analytics function started small – often with descriptive reporting – and then built off of those early wins. The world of People Analytics abounds with cautionary tales of large companies that began down the path of large, expensive, and time-consuming data integrations that required years to complete, cost political capital, and brought the entire initiative to a grinding halt.
2. Hire right, not bright
Make sure to hire the right skills at the right time.
People Analytics is fundamentally inter-disciplinary in nature. You need I/O psychologists that can understand what makes your employees tick, you need economists and statisticians that can analyze the data, and you need experienced HR practitioners who can help guide you with the cultural transformation.
Hiring the right team is one of the most important steps in being successful and it’s important not to compromise when it comes to these skills: you cannot simply force a non-technical person to take statistics class and expect them to perform advanced research.
But equally as important as finding the right person is making sure that you do it in the right order. If you hire a statistician before you are ready for advanced data science, you’re going to end up with one unhappy camper.
3. Pick the right fruit
If this is the organization’s first foray into People Analytics, you’ll find that “low-hanging fruit” doesn’t quite characterize it; there are so many opportunities that they’re practically sitting on the ground.
A quick analysis into the sourcing channels that produce the best talent will showcase your skills. Engaging in a quick study of the impact of manager effectiveness on employee performance will produce some powerful findings.
So it’s important to brand your work efficiently, spend time picking the right projects, pick a tangible dependent variable accepted by all, and come up with an over-arching penetration plan. This isn’t a sprint, it’s a marathon. You need to ensure that each of these projects fit into a bigger plan and that this ultimately becomes a part of the organization’s DNA.
Remember that there are several distinct phases of the employee lifecycle: selection, development, plateau, and separation. Make sure that you devote time to thinking about all of them rather than focusing on just one.
4. Good ingredients make a high-quality meal
People Analytics really starts and ends with the data.
It’s tempting to focus on data quality as the most important piece of the pie but data has many facets: availability, agility, periodicity, coverage (e.g., response rates), objectivity, and repeatability. Understanding the different facets of data is very critical for a successful analytics launch.
People often don’t spend enough time thinking about the different aspects of their data and instead focus on a very narrow slice.
To use an analogy, it’s like making a pizza. You need to start with the dough but if there’s too much oil or the tomatoes aren’t fresh enough or if the oven temp is not set right, the end product will be ruined. Likewise, data needs to be combined in the right proportions from the HR systems, employee surveys, business systems like sales and manufacturing, benchmark outputs, Applicant Tracking Systems, and publicly-available data among others.
Each of these systems has inherent advantages and disadvantages that complement each other, and it’s important to understand them.
5. You need two hands to clap
People Analytics is as much about the people as it is about the analytics.
It’s tempting to gather a bunch of data in isolation and then present the results to HR with the expectation that they’ll act on them but those sorts of initiatives rarely work. HR isn’t just a stakeholder, they should be a partner.
Instead of just asking “does HR know me?” ask yourself “do I know HR?” This is a two-way street and you will inevitably find yourself working hard to engage HR.
None have cracked the nut
There are countless examples of People Analytics projects that unearthed fascinating insights but ended up sitting in a binder somewhere on someone’s shelf because the work had been done in isolation without involving or gathering input from HR.
In spite of the recent media attention devoted to this space, very few organizations have made it very far down this path and none have really cracked the nut. Each has their own challenges that they face as they try to build out a robust People Analytics function.
If you’re just starting to head down this path, don’t get intimidated. There are plenty of others in the same shoes as you.
But there are things you can do to help avoid mistakes and give yourself the highest probability of success. The fruit are there sitting on the ground, but before you just start running around and picking them up, give some thought to how you want to approach the task at hand.
You’ll find that an ounce of advance planning and strategic thinking is worth a pound of analytics.
To access the full-length article, click here.