Our ca­reer

Finweek English Edition - - MANAGEMENT - You’ve ob­vi­ously had some very pub­lic ex­pe­ri­ence with the fact that even when the data is good and the model is good, peo­ple can push back a lot for var­i­ous rea­sons, le­git­i­mate and oth­er­wise. Any ad­vice for once you’re in that po­si­tion, and you have a sea

lytic tool kit to pair with ex­per­tise?

I think you want to in­te­grate it as much as pos­si­ble. That means that they’re go­ing to have some busi­ness skills, too, right? And learn that presenting their work is im­por­tant. But you need it to be in­te­grated into the fab­ric of the or­gan­i­sa­tion.

You’ve seen this shift in base­ball teams, for ex­am­ple, where it used to be that you’d hire an an­a­lyst to check that box and have them com­part­men­talise. That doesn’t ac­com­plish much at all.

If you can’t present your ideas to at least a mod­estly larger au­di­ence, then it’s not go­ing to do you very much good. Ein­stein sup­pos­edly said that I don’t trust any physics the­ory that can’t be ex­plained to a 10-year-old. A lot of times the in­tuit ions be­hind t hings aren’t re­ally al l t hat com­pli­cated. In Money­ball, that on-base per­cent­age is bet­ter than bat­ting av­er­age looks like “Okay, well, the goal is to score runs. The first step in scor­ing runs is get­ting on base, so let’s have a statis­tic that mea­sures get­ting on base in­stead of just one type of get­ting on base.” Not that hard a bat­tle to fight.

Now, if you feel like you’re ex­press­ing your­self and get­ting the gist of some­thing and you’re still not be­ing lis­tened to, then maybe it’s time to change ca­reers. It is the case [that] peo­ple who have an­a­lytic tal­ent are very much in de­mand right now across a lot of fields, so peo­ple can af­ford to be picky to an ex­tent.

Don’t take a job where you feel bored. If it’s chal­leng­ing, you feel like you’re grow­ing, you have good in­ter­nal de­bates, that’s fine. Some fric­tion can be healthy. But if you feel like you’re not be­ing lis­tened to, then you’re go­ing just want to slit your wrists af­ter too much longer. It’s time to move on.

A lot of times when data isn’t very re­li­able, in­tu­ition isn’t very re­li­able ei­ther. The prob­lem is peo­ple see it as an ei­ther/or, when it some­times is both or nei­ther, as well. The ques­tion should be how good is a model rel­a­tive to our spit­ball, gut-feel ap­proach? And also, how much do we know about this prob­lem. There are some is­sues where you just don’t have a good an­swer and you have to hedge your risks as a busi­ness and not pre­tend that you’re more cer­tain than you re­ally are.

A lot of pri­vate busi­nesses are very re­luc­tant to deal with un­cer­tainty in their out­look. The man­ager doesn’t want to seem like he’s not sure what he’s do­ing. And the con­sul­tant or the an­a­lyst wants to pro­vide in­for­ma­tion to make the man­ager feel more con­fi­dent. That’s quite prob­lem­atic be­cause a lot of prob­lems that are on the fron­tier of busi­ness, on the fron­tier of sci­ence, [are] by def­i­ni­tion fairly chal­leng­ing ones that no one else has solved.

That’s where hav­ing a more hum­ble at­ti­tude about what you can ac­com­plish and what you can’t is im­por­tant. Just be­cause a model is not go­ing to be very pre­cise or ac­cu­rate doesn’t mean that there­fore you should trust your gut in­stinct af­ter a cou­ple of whiskeys and as­sume it’s go­ing to be very much bet­ter. Wal­ter Frick is an as­so­ci­ate ed­i­tor at Har­vard Busi­ness Re­view.

© 2013 Har­vard Busi­ness School Pub­lish­ing Corp. © The New York Times 2013.

Newspapers in English

Newspapers from South Africa

© PressReader. All rights reserved.