Promote the Wrong Person? Don’t Blame it on Data or Software (Why People Decisions Must Be Owned by People)
December 3rd, 2014
Written by: Elizabeth Richards
I recently ran across a wonderful article on Google’s quest to develop a mathematical formula to decide which engineers get promoted—and how that effort failed.
Google, not surprisingly, has a People Analytics group whose original mission was that, “All people decisions at Google should be based on data and analytics.” Among other initiatives, the group developed an equation to help management determine which software engineers to promote. As Google executive Prasad Setty says in the article, “Initially…we wanted analytics to spit out people decisions.”
Given Google’s analytical, engineering-oriented DNA and its massive hiring needs, it makes sense that the company would try to find a more efficient, data-driven and accurate way to assess engineers’ performance.
But as I read the piece, I felt a bit queasy about the concept of completely data-driven promotions. Not because TPO is a consulting firm that relies on the very human elements of experience and judgment. We very much agree with and practice a data-driven approach—equipping decision-makers with metrics and tools where it makes sense. But ultimately, we believe that people-related decisions can’t be made by formulas alone. TPO’s position is that, when provided with the right information, it is people who make the best people decisions.
So given that philosophy, you can understand why I began vigorously nodding my head up and down upon reading this sentences:
“(Google’s engineers) didn’t want to hide behind a black box, they wanted to own the decisions they made, and they didn’t want to use a model. Not everything can be distilled down to an algorithm: even for Google’s engineers, automation has its limits. Today, The People Analytics group exists to arm decision makers with better information, not to replace them with algorithms.”
Now, this does not mean decision-makers should devalue data—on the contrary, we find that most decisions are made with too little of it. Too often, performance management tends to consist of checking some boxes on several forms, writing a pro forma paragraph about the employee’s performance over the past year and then—ultimately–making a decision whether to promote based primarily on how well-liked the employee is.
The right performance management approach is not just about forms—it is a process, supported by tools, to accurately identify high and low performers towards an organization’s larger business objectives. Who is ready to move up? Who’s almost ready? And who should be moved out?
As Google discovered, it is about putting the right information in the hands of decision-makers. Sound simple? Well then why does a recent Society for Human Resource Management (SHRM) Foundation Report state that:
“No more than 30 percent of those surveyed reported that their performance management system effectively establishes goals, provides feedback and actually improves performance.”
See our previous blog post on this topic here.
So it’s clear that many organizations need to change the way they view performance management. It’s not data. It’s not forms. It’s not software. It’s not just human judgment. It is the combination of those elements.
But if you’re still struggling with performance management, take heart—even Google, one of the world’s most advanced and smartest companies, got it wrong the first time.
Leave a Reply