Using Big Data to find employees who will not leave

by on July 11, 2014 at 2:34 am in Data Source, Economics, Web/Tech | Permalink

This is for call center operators:

The results are surprising. Some are quirky: employees who are members of one or two social networks were found to stay in their job for longer than those who belonged to four or more social networks (Xerox recruitment drives at gaming conventions were subsequently cancelled). Some findings, however, were much more fundamental: prior work experience in a similar role was not found to be a predictor of success.

“It actually opens up doors for people who would never have gotten to interview based on their CV,” says Ms Morse. Some managers initially questioned why new recruits were appearing without any prior relevant experience. As time went on, attrition rates in some call centres fell by 20 per cent and managers no longer quibbled. “I don’t know why this works,” admits Ms Morse, “I just know it works.”

The rest of the Tim Smedley FT story is here, via Peter Sahui.

1 Just Another MR Commentor July 11, 2014 at 3:35 am

On a related note here’s a great post today by Noah Smith

2 ummm July 11, 2014 at 3:58 am

You cannot support the meritocracy and be opposed to affirmative action while also opposing high tech immigration. Limiting labor options, like racial quotas in schools, hurts competitiveness.

3 Just Another MR Commentor July 11, 2014 at 4:49 am

Exactly, I’ dissapointed on the lack of awareness coming from this blog on how the lack of immigration is damaging competitiveness. Several months ago some attention was drawn here but Tyler and Tabarrock have really dropped the ball since. Can’t have this kind of smart policy debate being championed by left-wing hacks like Noah Smith even though he’s dead right on this issue. Given the hostility of the left to expanding business’s labor options it’s very courages of Noah to post this hwoever.

4 ivar July 11, 2014 at 6:17 am

The reason they won’t support this rationally is because they won’t allow tradeoffs. These libertarians are absolutists on immigration and consider low skilled, even high crime, immigrants as equally desirable as high skilled immigrants. They ignore the externalities of the latter and don’t see how allowing in so many low end immigrants (especially with the explicit purpose of changing the patterns of voters) is abhorrent to the median American and damaging to the social contract. I bet a deal that the majority of the US would approve of would be to increase high skilled immigration in exchange for removing family reunification beyond minor children and spouses combined with rapid deportation of illegals with any criminal records at all. The fact that they insist on greater high skill AND low skill immigration means that they are absolutist and will not compromise.

5 Just Another MR Commentor July 11, 2014 at 6:56 am

All immigration can be useful – and moral. The best policy both economically and morally is open borders. I don’t think anyone could rationally argue othewise. And please “social contract” – I think you’re lost a lot of credibility pretending there is supposed to be suc a thing, and if there is why should it only apply to Americans.

6 Just Another MR Commentor July 11, 2014 at 7:02 am
7 Z July 11, 2014 at 7:10 am

Supply and demand are social constructs!

8 Just Another MR Commentor July 11, 2014 at 7:15 am

The demand for tech workers right now is basically infinite so additional immigrants can only be beneficial. It’s people like Alex Tabbarrock who really face challenges from immigration as he competes every damn day in a high-pressure global marketplace. Opening up more immigration would force him to work much harder to stay afloat but he bravely and selflessly advocates for this moral position in any case.

9 Michael July 11, 2014 at 10:23 am

I’d like to thank you for posting a link to that Open Borders thread, JAMRC. It may be one of the biggest embarrassments I’ve ever seen.

Sometimes I really worry that libertarianism is taking hold and that the libertarians are going to convince young people to vote against their own economic best interests just like the trickle-down conservatives succeeded to do in the early 80s. But then I read a post like that and realize, nah, not gonna happen–these libertarians are just too stupid.

10 Just Another MR Commentor July 11, 2014 at 11:05 am

UMMMMMMM I guess you just aren’t ever economically literate or are heartless. Open Borders is moral and smart policy. We want more talented immigrants in this country, not less. Just give me more tasty immigrants.

11 Ray Lopez July 11, 2014 at 3:39 am

But I thought management books taught your most useful employees are the ones that are most likely to leave? Big Data here will simply find the dullards that are ZMP workers within a corporation.

12 Just Another MR Commentor July 11, 2014 at 4:58 am

This is just for call centre operator jobs – by definiton most of these people are going to be dullards. When it comes to serious work big data won’t help us much, the smart, dynamic portion of the US population is already 100% utilized we need to look beyond our borders to find future talent.

13 Shane M July 11, 2014 at 6:52 pm

I’m reminded of a question on a big box store application – something like “I should always try to find a better way to do something. True or False.” The correct answer for the job is False. The big box store has already figured out the process – they’re looking for employees to execute the plan without deviation.

14 dirk July 11, 2014 at 3:41 am

I’m looking for a job. Hey, me, people! I’m a loyal MR reader. Does that mean nothing?

I’m almost smart and I’m capable of working hard if you give me a carrot. Really, just a carrot. Why does nobody hire me? I’ve been looking for a job for two years. Am I a ZMP worker? What do you think, Tyler?

Anybody else reading this think I have a chance at getting a job? I spent a decade in the oil industry then quit for 2 years to try to sell my screenplays in Hollywood to no avail. Now I’m desperate & have been without income for 4 years. My professional experience is technical sales in the oil services industry. I also made a lot of money in the stock market when I shorted it in 2008 before the crash, although economists tell me I was just lucky.

15 Just Another MR Commentor July 11, 2014 at 3:43 am

If you are involuntarily unemployed by definition you are ZMP

16 dirk July 11, 2014 at 3:53 am

I quit my oil industry job in 2010 to pursue screenwriting. I started looking for another job in late 2012. What does that make me? I suppose I became “involuntarily unemployed” the moment I started looking for a new job.

17 dirk July 11, 2014 at 4:11 am

I’m a two-year-old “unwrapped cracker”, aren’t I?

18 Just Another MR Commentor July 11, 2014 at 4:50 am

If you had positive MP you would be employed, that’s how markets work.

19 dirk July 11, 2014 at 4:59 am

My MP has been measured at -.0023.

What can I do to change that? My wisdom has been measured at + 2.34. That’s not worth anything?

20 andrew' July 11, 2014 at 5:26 am

I can’t tell if you guys are kidding around. Not that I have any actionable input. Getting a two year programming degree worked gangbusters for a family member. Then get any job at a relatively dynamic place you can switch jobs quickly.

21 dirk July 11, 2014 at 5:58 am

I’m not kidding around. I’m looking for honest input.

22 andrew' July 11, 2014 at 6:59 am

Film distribution?

23 Keith July 11, 2014 at 9:16 am

My own theory is there will eventually be two kinds of jobs available in this world: IT and Healthcare. If you are introverted you will work as some kind of programmer, web designer, sys admin etc. Extroverts will gravitate towards healthcare.

Decide what you are, take some classes and get your foot in the door. Good luck.

24 Just Another MR Commentor July 11, 2014 at 9:23 am

This theory makes basically no sense, but on the other hand I use it as justification to advocate for massive amounts of new immigrants.

25 Andrew' July 11, 2014 at 1:40 pm


By doing screenwriting, isn’t that the most competitive highest barrier to entry part of the most competitive highest barrier to entry industry? Until you sell one it is a nice hobby. So, how can you make money while keeping the fire burning? Write for a sit-com or something?

26 Rahul July 11, 2014 at 3:49 am

Most of the paeans of these smart hiring algorithms are sung by employees of the firms peddling such software. I’d take that with a big pinch of salt.

I’ve rarely seen a third party assessment or even a fairly long time horizon study of the retention impact of such hiring strategies.

27 Just Another MR Commentor July 11, 2014 at 3:52 am

Nonsense, big data WORKS

28 andrew' July 11, 2014 at 5:16 am

Oh, I’ll get right on that. Just after I get the president to admit he lied about everything he said about the NSA and reverse the worst government abuse in us history.

Along with all the other stuff I can control. Like after 15 years of pointing out how obvious marijuana legalization is we have one sort of state dipping their toe while wringing their hand.

Yeah, we agree republocrats should do no-brainers.

29 andrew' July 11, 2014 at 5:18 am

Meant for the skilled immigration disappointment comment.

As for big data, measuring whether you have two or three Facebook type accounts us what passes for big data?

30 RobertW July 11, 2014 at 7:17 am

Because Algorithms! :/

31 Andrew' July 11, 2014 at 1:37 pm

I have no doubt big data will work just like the hydrogen economy will work. And that isn’t a joke. Hydrogen is coming, but it’s going to be as slow as Christmas.

But I want to start a list of all the fads that come and go with a lot of promise and not much deliver.

32 Dan July 11, 2014 at 5:59 am

Mostly, this sounds like the psychological assessment tests that have been around for 50+ years, tarted up with Social!Big!Data! I don’t deny it works, but it sounds like a small twist on an established practice.

And this sounds like the insight you could more easily get by asking your employees where to post job ads:

“When we analysed the data of the top performers in that job family, we found out that they all hung out at a very unique, niche social media site”

33 Bill July 11, 2014 at 11:24 am

You’re correct. What is different is the greater use of social network analysis, ways to measure engagement, etc.

What this means is not exactly what the post had in mind, however.

For example, one of the assessment questions might be: “Who do you turn to if you have a question about X subject. Or, who would you trust with the following problem.” These questions end up revealing who is important in the network, but it also reveals isolates, those who are less connected. You can view this two ways: get rid of the isolate, or, try to get the isolate involved in group activities or connected to a person who is. Management may not have time for engagement activities, and the isolate may experience job performance problems (not knowing what to do, not understanding norms), or may leave because because of being isolated.

Now, go back to the initial post and ask yourself this question: Is the post measuring what IS, or is it trying to determine what management should change. Answer: It measures what IS–how well connected persons are to groups within the organization, and has no discussion about what should change.

Now ask: What if you are a member of a minority group, and the Xerox representatives ask you, and others around you, how well connected you are to determine whether you should be promoted or retained.

34 dearieme July 11, 2014 at 6:08 am

As soon as it’s widely used, it’ll be “gamed”. Such is naughty old homo sapiens.

35 prior_approval July 11, 2014 at 7:13 am

It’s already been gamed – or you think that example of LV= is not exactly an example of keeping things in the family? Though the gaming at that level is not exactly at the applicant level.

Much like the call I just got today from a friend, offering summer jobs at Siemens for students – it is pretty much the same thing, as the number of Siemens (and Mercedes) employees is noticeable around here.

36 Just Another MR Commentor July 11, 2014 at 7:21 am

I was in your neck of the woods the other week in fact.

37 Axa July 11, 2014 at 8:43 am

Once I met a call center middle manager. They hired anyone who asked a job but in the first week the management tested the new employees. Tests were about the dress code, bathroom breaks and such. The objective was to enforce stupid rules so the moderately intelligent employee left the job in the first week. In this way, the call center keep obedient employees.

If you’re smart enough to think about why you need a tie for your call center job, you’re too smart for the job.

38 Rahul July 11, 2014 at 10:21 am

How much does a typical call center job pay? Just curious.

39 Boonton July 11, 2014 at 12:55 pm

Back when Bush wanted to turn Social Security into a 401K type system, I think it would have been very important to know that nearly 10% of people would loose money even if they did exactly what they were supposed too (namely invest in a low cost index fund and leave the money there for as long as possible).

But this simulation is premised on a very unrealistic way of investing…namely taking a big lump of money when one is 30-something, putting in the market at once and forgetting about it until one becomes 60-something. Most people contribute yearly and their contribution goes up over time as they get raises.

What if you reran this simulation using yearly contributions starting at $1000 and going up 3% per year? Would you still get nearly 10% losing money? What would the 10% most unlucky people look like relative to the 10% most lucky?

If you made it a national policy, would the stock market’s volatility remain the same with the influx of more constant buyers? If its volatility decreases would it’s average return also decrease thereby undercutting the argument for privitizing social security?

40 Shane M July 11, 2014 at 6:56 pm

Somehow this got posted on the wrong thread?

41 Dana Harrison July 11, 2014 at 1:44 pm

As someone who just spent 5 years working for an assessment provider and a lot of time helping clients do better hiring, I’ll say that this absolutely makes sense to me.

My hunch is that the individuals who were with fewer social networks were more introverted in the sense of needing fewer (and perhaps overall less) personal connection with others. Those individuals with 4+ networks are more social. One would assume that for call center work, you’d want highly extroverted, social people, but this often is not the case.

Although call center people are communicating a lot, their work often does not involve forming relationships with customers. The call center work is often highly transactional (not relational) and many calls centers to a greater or lesser extent scripts the reps. In these environments, the highly extroverted person who needs a lot of personal connections, personal feedback, and relationships is stymied. You often find them unhappy with the work itself and bopping around the room or otherwise engaging with people on a more personal level than what they get from their calls with customers.

Common characteristics for strong call center employees include – but are not limited to – being comfortable with repetitive and stationary work, and caring about the customer but doesn’t take interactions personally. You often do not find these traits with social butterflies, but instead with the less extroverted individual. Therefore, correlation, if not causation, between less social people who will be more successful in this work and people who have fewer networks makes sense.

42 MikeP July 11, 2014 at 2:44 pm

Some managers initially questioned why new recruits were appearing without any prior relevant experience. As time went on, attrition rates in some call centres fell by 20 per cent and managers no longer quibbled. “I don’t know why this works,” admits Ms Morse, “I just know it works.”

Uh… Someone who has prior relevant experience has a history of leaving such a job. Someone who doesn’t does not.

Can Bayesian reasoning be a high school requirement? That would actually be useful.

Comments on this entry are closed.

Previous post:

Next post: