Ama­zon’s data-driven ap­proach more com­mon


Ama­zon isn’t the only com­pany that is us­ing data on em­ploy­ees to im­prove pro­duc­tiv­ity.

A New York Times ar­ti­cle over the week­end por­trayed Ama­zon’s work cul­ture as “bruis­ing” and “Dar­winian” in part be­cause of the way it uses data to man­age its staff. The ar­ti­cle de­picted a work cul­ture where staffers are un­der con­stant pres­sure to de­liver strong re­sults on a wide va­ri­ety of de­tailed met­rics the com­pany mon­i­tors in real time — such as what gets aban­doned in peo­ples shop­ping cards and what videos peo­ple stream — and en­cour­aged to re­port praise or crit­i­cism about col­leagues to man­age­ment to add to more data about work­ers per­for­mance. The story led to an out­cry on so­cial media.

Ama­zon’s CEO Jeff Be­zos said in a memo to staff on Mon­day that the ar­ti­cle doesn’t ac­cu­rately de­scribe the com­pany cul­ture he knows. But ex­perts say the kind of data-driven staff man­age­ment Ama­zon uses is set to be­come more com­mon as tech­nol­ogy con- tin­ues to trans­form the Amer­i­can work­place.

“Ev­ery com­pany is some­where in process to­ward us­ing data to get a bet­ter han­dle on who their top per­form­ers are and to un­der­stand where peo­ple stand,” said John Chal­lenger, CEO of out­place­ment con­sul­tancy Chal­lenger, Gray & Christ­mas, Inc.

Com­pa­nies, both large and small, have been mov­ing away from tra­di­tional hu­man re­sources re­views that rely on an­nual per­for­mance eval­u­a­tions. They’re mov­ing to­ward a more data-driven ap­proach with more fre­quent feed­back, check- ins, and other met­rics.

Con­sult­ing firms Ac­cen­ture and Deloitte both said this year they would re­vamp their per­for­mance re­view pro­cesses, for ex­am­ple, adopt­ing a more data driven ap­proach that in­cludes more fre­quent rat­ings by man­agers and other in­ter­nal feed­back and data that can be ag­gre­gated and an­a­lyzed to pro­vide a bet­ter por­trait of per­for­mance than a sin­gle rat­ing. In an es­say in the Har­vard Busi­ness Re­view, Deloitte said the new ap­proach uses “the tech­nol­ogy to go from a small data ver­sion of our peo­ple to a big data ver­sion of them.”

Tech com­pa­nies have been even speed­ier in ap­ply­ing data an­a­lyt­ics to staffing. Google, for ex­am­ple, uses data to fig­ure out how to put to­gether op­ti­mal-sized teams for projects and fig­ure out what makes ef­fec­tive lead­ers.

Paul Hamerman, a For­rester an­a­lyst who fo­cuses on hu­man re­sources man­age­ment and fi­nan­cial ap­pli­ca­tions, says the fu­ture may look more like what Glint Inc., based in Red­wood City, Calif., is of­fer­ing clients. The com­pany, with clients in­clud­ing mu­sic-stream­ing site Pan­dora and mar­ket­ing au­to­ma­tion com­pany Mar­keto, sends em­ploy­ees what it calls “pulses,” or short sur­veys about how they are feel­ing and how they feel about their job.

Glint CEO Jim Bar­nett said the sur­veys let ex­ec­u­tives see how the health of their em­ploy­ees and com­pany are far­ing in real time, in the same speed with which they might be able to check sales re­sults or mar­ket­ing im­pres­sions. Since the “pulses” to com­pany em­ploy­ees re­cur more fre­quently than tra­di­tional re­views. And their data can be ag­gre­gated to give a clearer pic­ture of how em­ploy­ees are far­ing over­all.

A Faster World

“The old men­tal­ity was once a year we would check in with an an­nual sur­vey, have an an­nual re­view, set goals,” said Bar­nett. “What we’ve learned is the world to­day moves much faster than that.”

One of Glint’s clients, Mar­keto, was able to use the data gleaned from the Pulses to see that women in one depart­ment were rank­ing their work/life bal­ance sub­stan­tially lower than ex­pected. The com­pany found a staffing short­age in that area and in­creased staff.

“What they were able to do was to go in and in­crease the staffing be­fore they had sig­nif­i­cant at­tri­tion,” Bar­nett said. “The beauty of sys­tems like this is you’re able to link ac­tions to out­comes.”

The down­side to a data-driven ap­proach is it can seem “Big Brother-ish” to staffers. But Glint said the sur­veys that the com­pany sends out have an 80 to 85 per­cent re­sponse rate. “Em­ploy­ees tend to be will­ing to share,” Bar­nett said.

Another draw­back: Re­ly­ing strictly on num­bers can lead to the per­cep­tion of a cold-hearted work­place. “It’s easy to get so hung up on sta­tis­tics that you miss the value of what that in­di­vid­ual brings to the ta­ble in terms of per­son­al­ity, con­nec­tiv­ity and those in­tan­gi­ble pieces,” said David Lewis, CEO of HR out­sourc­ing and con­sult­ing firm Oper­a­tionsInc in Nor­walk, Conn.

That can lead to a dys­func­tional work­place. “If ev­ery­body is mis­er­able about what they are do­ing at work that bleeds over,” said Jay Stark­man, CEO of En­gage PEO.

But in gen­eral, per­for­mance track­ing makes sense, he said: “Com­pa­nies and jobs should not be fam­ily. You can’t fire crazy Un­cle Jim. Fam­i­lies aren’t gauged by per­for­mance.”

Michael Diste­fano, chief mar­ket­ing of­fi­cer at ex­ec­u­tive re­cruit­ment firm Korn Ferry in Los An­ge­les, likens the data-driven work­place ap­proach to how data has taken over fit­ness train­ing. Rather than hav­ing your fit­ness level mea­sured at a checkup or visit to the doc­tor, it’s now avail­able in real time to ev­ery­one with a fit­ness tracker like Fit­Bit or Ap­ple Watch.

“I en­vi­sion a world where we’ve all got wear­ables on and you send a pulse sur­vey to a few hun­dred em­ploy­ees a day ask­ing about dif­fer­ent lev­els of en­gage­ment: ‘Here are four emoti­cons, which one best de­picts how you feel about your job?,’” he said. “That gets trans­mit­ted to a dash­board, and the first thing a CEO does in the morn­ing is grab his tablet, plug into the dash­board, see how peo­ple are feel­ing and make a few calls ac­cord­ingly.”


In this June 16, 2014 file photo, Ama­zon CEO Jeff Be­zos walks on­stage for the launch of the new Ama­zon Fire Phone in Seat­tle, Washington.

Newspapers in English

Newspapers from Taiwan

© PressReader. All rights reserved.