A new study from MIT says “Fifty -eight percent of organizations now apply analytics to create a competitive advantage. . . . Significantly, these same organizations are more than twice as likely to outperform their peers.”

Called “Analytics: the Widening Divide,” the study makes a strong point that companies that don’t adopt analytics are in deep trouble. Not only are they failing now, but they’re also failing to develop the capability that will allow them to compete in the future.

With analytics becoming more complex and more valuable all the time, the widening gap may become insurmountable.

The study also looks at where analytics are being applied. Top uses are in Finance and Operations. HR is lagging.

As our economy shifts more and more to a knowledge and service economy, excellence at human capital management is more and more critical for success. There’s an opportunity for HR leaders to help their companies win in the next wave of competition by adopting the analytic tools that will help them optimize the quality of their workforce and their spend on that workforce.

In years of working with recruiters, one of the most common filters I’ve found them using is to discard the resumes of job hoppers.  The conventional wisdom is that people who’ve had a long string of short term jobs aren’t worth hiring. Either they’ll quit quickly, or they’re no good and they’ll be let go.  In either case, they aren’t worth pursuing.

It’s a strong and consistent bias.

When I google the phrase “job hoppers,” the top result link is”Never Hire Job Hoppers. Never. They Make Terrible Employees.”

I used to work for an ATS company, and a popular feature was a filter that automatically removed job hoppers from the search results.

But what if the bias is wrong?  What if there’s nothing wrong with people who, in their past, have changed jobs frequently?

The Evolv analytics team recently asked this question. They looked at more than 20,000 people hired through Evolv, with an eye towards how many jobs they had held in the previous 5 years, and how long they stayed on the job they were hired into.

The results, available here, were really surprising.  It turns out that there is no correlation between how long someone lasts on the job and how long they’ll last on their next job. A candidate who’s had five jobs in five years is no more likely to quit that someone who’s had one job for five years.

People who are unemployed when they apply for a job also have the same expected tenure as any other candidate.  (This one surprised me, as I would have expected unemployed people to be a little more likely to want to keep their job.)

In other words, a very common screening tool turns out to be totally non-predictive.

This is one more reason why employers need to use data, not hunches, when making hiring decisions.

AllAnalytics.com did a very nice write up on Evolv, and how we’re helping our clients hire better. You can read the whole article here.

The article covers what we do: data driven workforce selection.  We use data from the hiring process to determine which candidates are most likely to be star performers.

Then it covers the results our process delivers:

• 15% greater productivity
• 10% higher scores on quality of service
• 12% increase in employee retention

These are numbers that employers need to sit up and notice.  The economic impact of this improvement is huge. The Results Companies have a measurably better product than their competition, and at a lower cost.  We’re seeing the same thing across our installed base.

It’s not surprising that using a science and data driven process results in better hires, which is why employers are turning to Evolv to help them get an edge on their competition, or simply to lower their costs of recruiting and turnover.

According to The Atlantic and ERE.net games are beginning to enter the workplace as a means to forecast candidates’ on-the-job performance.

I love the idea of a game-like pre-employment assessment.  Traditional assessments are typically made up of bland, transparent questions that sap energy and create a negative hiring experience.  It’s no wonder that in Evolv’s most recent Retail Trade Business Survey we found that the number one recruiting objective for retailers in 2012 is “Better branding and candidate experience.”

But all assessments – even engaging, branded, gamified “candidate experiences” – are just predictions; forecasts of future performance.  So, how good can a forecast be?

Traditional assessments built for industries and broad categories of behavior.  They also tend to be unchanging for years at a time. Because of this, they tend to be weak at predicting performance for a hires into a specific job at a specific company. They’re fine for identifying the worst of the pack, but do little to raise the bar on the average hire.

Evolv “closes the loop” on our assessments to make them unique.   We correlate individuals’ on-the-job performance with their assessment responses, and then optimize the assessment to more accurately identify top performers.  The result is an assessment that is optimized to the unique characteristics of your business: identifying those who will significantly outperform the pack

Gamification is great. But the ultimate objective of candidate screening must be to build a better workforce.  So bring on the games – let’s have some fun!  But be sure that whatever you do with your candidate experience, you don’t lose sight of the key objective: select the workforce that will make your company a winner.

For those who wonder just how powerful analytics and big data can be . . .

Target has gotten so good at data mining that they can spot when women are pregnant and when they’re going to give birth. (Read the NY Times article here.)

They look for things like purchases of unscented lotion, because pregnant women get a more sensitive sense of smell.

The Target line on this:  “We are very conservative about compliance with all privacy laws. But even if you’re following the law, you can do things where people get queasy.”

To mask that queasy feeling, Target disguises how much they know, placing the diaper coupons with coupons for other things, so you don’t know they’re on to you.

Never the less, there’s a story of a father who’s high school daughter started getting coupons for cribs and diapers.  Turns out Target knew something that Dad didn’t know – yet.

Evolv customers don’t need to worry.  We don’t do anything creepy with your data. Unless, of course, higher productivity and lower attrition freaks you out.

Check out this Wired Magazine article.

Superstar Jeremy Lin was overlooked for most of his career, and now he’s apparently the hottest thing in the NBA.

Meanwhile, the NFL’s runs a thing called the scouting combine, where prospects are measured on everything from doing pushups to psychometric responses.  It turns out that all that testing isn’t one bit predictive of performance.

It gets worse: top draft picks only outperform lower draft picks 52% of the time.  Yikes.  These are multi-million dollar decisions, and they might as well toss coins as use all their fancy tests.

Best line in the article: “There is talent everywhere. We just don’t know how to find it.”

That’s just not true.

Talent is everywhere.  And we know how to find it.  That’s why Evolv is in business.

The problem is the resistance to using data to make decisions that have traditionally been made by people.  There’s a fear of upsetting a status quo that you have been supporting. If you’ve spent 20 years as an MLB talent scout, it would be pretty embarrassing to say that you actually aren’t sure you know what you’re doing.

But the times are changing. I don’t think the “I’ve got a hunch” approach will work out for folks much longer.

The New York Times just had an interesting article on if Amazon is going to offer analytics in the Cloud.

I think the question is not if, but when.  The value created by analytics is far greater than that of putting computing power in the cloud.  Analytics can be used to find powerful correlations and give a much deeper level of insight to business managers as they try to make better decisions. That sounds great, right?

But it’s so vague, so squishy, so cloud-like.  Perhaps what I’m concerned about is best evinced in the quote by Kyle McNabb, a vice president at research firm Forrester. “They could . . . possibly offer to match your data to other large data sets and find something useful.”

Possibly find something useful?  That’s not the kind of thing you take to your boss with a purchase order.

I’ve seen this before. I went to a presentation by a leading analytics firm a few months ago, and the presenter/salesman said they often found “amazing things” when they looked at data.  A question from the audience was if he could give an example of a business outcome tied back to these amazing things.  He couldn’t. He said they were working at the “strategic” level, and removed from operations.

At Evolv, our focus is on finding things that are both amazing and useful.  We support data driven workforce selection: using performance data to determine which candidates will make the best hires. We’re delivering 10% to 15% improvements to productivity combined with reductions in attrition of 30% or more.

All employers know that the top performers are radically more productive than the bottom performers. Evolv employers also know who is more likely to fall into that top rank before they hire them.

A challenge I see with analytics and big data is that it’s very expensive.  It can be hard to build the business case for amazing things.

At Evolv, we know recruiting analytics and hourly hiring intimately.  We have deep expertise on the business of our clients, and know what drives their profitability.  We use that expertise to drive better outcomes.  Our focus allows us to do so in a cost effective manner.  The amazement we deliver is when we quantify the savings we deliver and show the quality improvements our process allows.

I’m very optimistic about the future of Analytics in the Cloud.  I think the driving force for it will be focused, vertical solutions by providers that can optimize specific business functions where the impact is very valuable.  Evolv is one of those.

There’s been a lot of snark about the recent court decision that says being too smart is a reason to NOT hire someone.  Apparently, Robert Jordan is too smart to be a policeman.  (Read about it here.)

On the  one hand, there’s nothing surprising here.  The idea of being overqualified for a job is old hat.  The logic is that the person will get bored, start bucking for a promotion, or quit as soon as something better comes along.  When a hiring manager extends an offer, they’ll want to know that the person is going to stick around and do the job happily.  Turnover is expensive and disruptive. Our analysis is that for call center workers, the cost of someone leaving is easily thousands of dollars.

Additionally, using assessments to predict performance has been shown again and again, by Evolv and others, to be far more accurate than interviews. You want the best workforce, don’t you?

So, as someone who has seen how effective these tests can be at building a better workforce, I’d normally be all in favor of this.

But I’m not.

The problem is that for most jobs, there are different paths for management and the rank and file.  It’s highly unlikely that someone working the phones in the call center or doing the night shift at a gas station is actually on a career path to become the CEO.

In police work, it’s the opposite.  Even the Police Commissioner of New York started as a beat cop.  Same with the Chiefs of Police in San Francisco, Los Angeles, Chicago, Dallas . . . . You get the picture.  By capping the IQ of the people at the bottom, “because they’ll get bored,” they’re deciding to cap the IQ of the people at the top, because you have to have experience to get there.   I don’t know much about police work and managing police forces.  But I’d guess that being smarter would be better for managing a department.

The solution might be to do what a lot of large employers do to attract the best and the brightest: have management rotation programs.  Tell would be cops that if they want to be on a management track, they need to do rotations around the department for several years. This way, you’ll get smart, motivated people looking at all different aspects of police work.  And you’ll be able to hire people who score high on IQ tests.

In the meantime, hopefully those managing police screening will think about the implications of screening out smart people.   That is, if they’re smart enough.

With all the coverage of SuccessFactors being bought by SAP, some people may have missed the news that SuccessFactors bought Jobs2Web.  Spending $110 million for a company that did about $10 million in revenue in 2010, SuccessFactors was making a pretty big bet on analytics.

I think this is good news for HR.

Described as the Marketo of recruiting by Brian Sommers, Jobs2Web is a jobs distribution platform that pushes jobs to dozens or hundreds of job boards and other sites, and helps recruiters build “talent communities.”  But a lot of companies do that.  When I saw Jobs2Web at HRTech last year, the thing that impressed me most was their analytics: the ability to show exactly where candidates are coming from, how much they cost and how many get hired.

Recruiting analytics is a critical function that many companies skip altogether, or do so poorly that they might as well skip it. The challenge is getting good data into the platform, and then knowing how to turn that data into real insight.  The fact that Lars saw recruiting analytics as an investment worth $110 million shows something we at Evolv see as well: there is an increasing awareness from HR leaders that data driven HR is the goal they need to be moving towards.  I believe data driven workforce selection is the most important aspect of that because quality of hire drives pretty much everything else a company does over the long term.

Evolv’s customers are already seeing the benefits of deep recruiting analytics. By tying analytics not just to the advertising spend, but the assessment, hiring process, and post hire performance, we’re giving them outsized returns on their entire recruiting process.  We’re also showing them, in real time, how the quality of their workforce selectivity ties directly to the productivity of their workforce.