Billy Beane’s data-driven approach to baseball recruiting made the Oakland A’s a winning franchise, inspired a Brad Pitt movie — and might soon change the way long-term care operators hire staff.
At least, that’s what I’m thinking after reading Don Peck’s “They’re Watching You at Work,” in the December issue of The Atlantic.
Peck investigated how companies are utilizing data to make hiring decisions. Organizations are doing this in a variety of ways, he reported. Some are using specially designed video games that capture “several megabytes” of data about a player/prospective hire, including the order of actions, decision-making time and tendencies to hesitate. These translate into “a high-resolution portrait of your psyche and intellect,” according to proponents — and the games proved remarkably accurate in tests run by oil company Royal Dutch Shell.
Other companies are turning to startups such as Gild, which specializes in matching computer coders with employers. Gild’s sophisticated algorithms trawl the Internet to catch a huge amount of data about coders, looking at everything from actual written code to advice coders give on websites. It even evaluates coders’ Internet browsing habits (turns out great coders tend to all be devoted to a particular Japanese manga website, for reasons that remain obscure).
At first, it seemed to me that these new, data-driven hiring strategies might work for huge businesses like Shell or hot labor markets like tech, where job candidates are plentiful and there’s a need to winnow down using fine-grained variables. Then I read this:
Because the algorithmic assessment of workers’ potential is so new, not much hard data yet exist demonstrating its effectiveness. The arena in which it has been best proved, and where it is most widespread, is hourly work.
The hourly work in question is — so far — mostly at big-box stores and call centers, according to Peck. These jobs have been the proving ground for data-driven hiring because of a few factors: They exist in huge numbers, they turn over quickly, and workers’ success can be clearly measured. So, hiring managers have been under pressure to fill vacancies, and they have turned to using data in hopes of making better hires, faster. They then have been able to measure how successful their new recruits have been. And the results have been good.
Applicants at Xerox, for example, now take an online assessment. Their responses are evaluated by an algorithm, which assigns them a color: green means they’re a best bet, red means hiring managers should look elsewhere, yellow means take a closer look. Using this method, the company’s attrition rate quickly fell 20%, and promotions are becoming more common, according to Peck.
There’s no doubt about the rising demand for LTC professionals, including hourly workers such as certified nursing assistants. It’s also no secret that turnover is an ongoing challenge in this sector. As for how easily success can be measured in this realm, that might be trickier, given the intangible people skills that great caregivers possess.
But just ask anyone who works in a skilled nursing facility these days about quality measurements and you’re likely to hear about intense projects to improve documentation of everything from pressure ulcer incidence to hospital readmission rates to customer satisfaction. The point is, there are metrics that reflect staff performance, and operators are already on a quest to capture as much data as possible.
Technology vendors certainly recognize this trend, and they’re developing products to help operators’ efforts. With a punch of an iPad button or the swipe of a card, it’s becoming remarkably easy to capture and record information about residents — and about workers. And this is where the data-driven future might look less like a utopia and more like an Orwellian nightmare. Here’s what Peck writes about specialized badges created at the Massachusetts Institute of Technology, meant to be worn on the job by workers:
The badges capture all sorts of information about formal and informal conversations; their length; the tone of voice and gestures of the people involved; how much those people talk, listen, and interrupt; the degree to which they demonstrate empathy and extroversion; and more. Each badge generates about 100 data points a minute.
The innovators who created and have tested the badges say that they’ve been able to draw striking conclusions, and have even identified the “data signature” of great leaders. These are people who, among other habits, have “brief but energetic conversations” while they “circulate actively,” and who are equally good listeners as talkers.
Are we entering an era when CNAs will be hired by algorithm, and administrators will be evaluated through the data captured on their lapel badge?
The promise of data-driven hiring is great. Clearly, so are the privacy pitfalls. Weighing the risks against the rewards, Peck described himself as “cautiously optimistic” about the future, and said this is the consensus among the psychologists and economists he spoke with in his reporting. He noted that he did not consult philosophers.
I wish he had talked to some long-term care workers. They might not be philosophers, but they’ve certainly given a lot of deep thought to these conundrums. From video cameras in rooms to monitoring devices on residents to floors that can track movements, they’re being confronted with new technology that is constantly testing boundaries, forcing them to weigh residents’ dignity against their safety. It appears that LTC professionals will only have to think even more deeply about these issues — as they apply not to their residents, but to themselves.
