So far, 2017 seems to be the year of the HCM technology startup. Before the end of 2016, the sector had hit an annual all-time high in deal counts, according to CBInsights, with nearly $2 billion invested. From what we can see, the issues these new companies are trying to address are all over the map, ranging from employee engagement to candidate screening, recruitment marketing, mobile recruiting and background checks.
Many new-generation products are based on some kind of analytics, machine learning or artificial intelligence platform. Some are building better mousetraps while others pursue truly innovative solutions. We’d wager many a founder’s exit strategy is acquisition by a bigger player, and more than a few observers predict that we’ll see quite a shakeout over the next year or so as some startups are purchased and others fail.
All that aside, we’re wondering how many of these HR-focused businesses have included the rise of cross-functional analytics in their plans. More and more, we hear data scientists express the view that metrics need to be drawn from across the organization if they’re to have real meaning. If HR tech startups haven’t considered that, it’s the ERP-rooted companies involved in HCM technology—brand names like Oracle, SAP, IBM and firms that have partnered with them—who are in the best position to provide the most meaningful analytics.
Demolishing Silos for Real
Analytics can help an organization determine whether it’s approaching recruiting, training and retention in a cost-effective way, but not if it relies on HR numbers alone. That’s why data scientists regularly complain about the limitations of siloed data.
In HR, the demand for powerful analytics has been driven by the conviction that, properly put to use, data analysis can measure and improve a number of performance measures related to the workforce, from quality of hire to productivity. That’s not a new idea in business thinking—sales, marketing, manufacturing, customer service and other functions have all been using data-based management for years. But in the world of human capital management, that line of thought led to several unfortunate conclusions, including the belief that everything from learning results to engagement could be boiled down to numbers if only HR would adopt the right technology, embrace data science and act on what it learned.
Beginning around 2013, a veritable industry developed around the idea that HR practitioners were hopelessly out of step with the rest of the organization when it came to data science. As a function, HR stood accused of dragging its feet, either because it was locked into process-driven approaches or unwilling to boil workforce performance into cold, hard numbers.
Pick your summation: HR hadn’t climbed aboard the train, or it hadn’t drunk the Kool-Aid. Whichever works.
Can People Analytics Live Up to its Promise?
In truth, HR executives, at least at large companies, work hard to translate employee performance into the kinds of numbers senior business leaders can understand. As we’ve previously reported, investors want public companies to disclose more workforce data as part of their regulatory filings. And while Bloomberg BNA reports 70 percent of HR leaders said they have “full” or “substantial” involvement in making key business decisions, it also noted that, “While HR has secured a seat at the corporate table at many, if not most, organizations, its strategic role still tends to be collaborative or supportive. Employers are largely disinclined to give the HR department exclusive control over organizational development (25 percent) or succession planning (11 percent).”
If HR’s going to take on a truly strategic role, anyone thinking about incorporating data science into human capital management should realize that several old dynamics are still with us.
- First, although HR sees itself as being heavily involved in strategic decisions, its role remains limited, to the “collaborative or supportive,” as Bloomberg noted. To draw an analogy: The Department of Agriculture sits at the same table as the Departments of State or Defense. In most matters, who do you think has the stronger voice?
- Most businesses put workforce management into the hands of someone to whom HR is a sideline, either because the company’s not large enough or its leadership doesn’t see the value of investing in HR specialists.
- Most business don’t have enough data from which to draw meaningful conclusions. Consequential analytics require a lot of data. A company with fewer than 50 employees simply isn’t generating enough information to be telling in most HR-related scenarios.
- Finally, most employers still want to invest as little as possible into their workforce. Witness the consistent expansion of contingent workers. Who wants to bet that all of today’s talk about “engagement” will get awfully quiet when the economy begins its next downward cycle?
Describing much of this in a June 2017 online article for the Harvard Business Review, the Wharton School’s Peter Cappelli expressed skepticism about whether people analytics can live up to its hype. “The questions that really matter have been under investigation longer than most other business topics,” he wrote. “What determines a good hire, for example, has been studied in almost the same way since WWI. So the idea of bringing in exploratory techniques like machine learning to analyze HR data in an attempt to come up with some big insight we didn’t already know is pretty close to zero.”
As an example, Cappelli cites Google’s numerous and ambitious efforts to analyze its workforce data. The company’s “Project Oxygen,” for instance, set out to determine the characteristics of a good manager. “Most of the conclusions from that very intensive exercise were ones that research discovered decades ago and which could have been found in textbooks,” Cappelli believes. “That doesn’t mean it’s not a worthwhile exercise to test how those standard assumptions of management play out in our own organizations, but expecting to find big and new insights is simply a bad bet.”
Analytics’ Value May Be Hidden In…
To our reading, Cappelli’s posing a simple question few people ask about people analytics: What’s the point? For analytics, the answer may lie in putting more emphasis on its use as an operational rather than a research tool. For years, HR has sought a strategic role in the business, and it may be that applying analytics to the traits and behaviors of the workforce is the way to get there.
For example, consider Microsoft’s new Workforce Analytics product. It compiles aggregate metadata generated by Office 365 Enterprise users and puts it to work generating insights into productivity. It might calculate how many meetings workers attend each week, how much time they spend with email and the number of connections they have both inside and outside the company. Overlay such information with sales data, for example, and you can look for patterns that define the working habits of successful people in a variety of roles.
They key word there is “overlay.” Once and for all, organizations need to accept that people analytics by itself is of limited value. To measure a hire’s true cost, an employee’s true value or a learning program’s true return on investment, data from HR must be melded with metrics from other functions.
Cappelli writes that “HR should be analyzing relationships among the data,” such as how hiring criteria relate to actual performance. If hiring decisions are based on certain skill sets, years of experience and talent assessments, the performance of new hires must be examined in the context of their results while doing the actual job.
But more often than not, HR develops that perception without looking at data from the employee’s own department. In every organization we’ve ever spoken to on the subject, no mechanism exists to link an employee’s efficiency and productivity with their compensation, benefits and other costs. A UX designer who resents feedback, even if they hide it well, can slow projects down, thus impacting sales and customer satisfaction. The design projects they’ve worked on in previous positions may not translate well to their current assignments. And, the hiring process for this particular person may have lasted six months, incurring both internal and external costs all the while. So, was the hire successful? No one can know until they’ve looked at data from every step in a role’s lifecycle, from the cost of writing the job description to the revenue/expense ratio involved with compensation and other overhead costs that apply to that individual.
It’s the big vendors—the ones whose technology touches multiple departments—who are in the best position to figure this out.
Image Copyright: rawpixel / 123RF Stock Photo