Hiring Metrics: Totally Useful or Totally Useless?

This is a real dilemma for me. On one hand, corporate recruiters clamor for metrics, but when I tell them what is necessary, their eyes glaze over. My advice to them is to be careful: you just might get what you wish for. You see, metrics attract attention. When competent people have competent metrics at their fingertips, they know what to do with them. Incompetent people just get confused. Depending on your competency, hiring metrics will attract attention that can either work for … or against … you.

About Metrics

The only reason to collect metrics is to make informed decisions: am I doing well? What do I need to improve? Are my activities helping achieve organizational goals? And, so forth. Metrics are not the “cool” thing to do, nor are they an interesting thing to play with. Metrics should provide specific feedback in specific areas that are actionable.

Consider quality initiatives. Many manufacturers measure product quality at the end of the production line. This is problematic because there are many small steps along the production line that could affect the final product. Measuring quality at the end of the line tells you whether the product is good or bad, but almost nothing about how to fix it. About 50 years ago, W.E. Deming and J.J Juran became leaders in quality improvement by breaking manufacturing into small process steps; collecting data at each step; working to control the process variability; and repeating the process until it was under control. In my experience, HR needs to incorporate some of their ideas.

To incorporate quality control into hiring, we must identify specific job competencies (i.e., the “how,” not the “what”), evaluate the accuracy of each testing-tool, evaluate demographic impact, gather performance evidence, work to control the variability, and so forth. Evaluating hiring quality takes professional practice. So why is it not done more often? I can only think of four reasons:

  1. People don’t have a clue how to do it
  2. They think it’s too much work
  3. They only care about surviving the guarantee period (this is what I hear most from external recruiters)
  4. They did not think of it themselves (see #1, #2 and #3)

The Status Quo

It does not help that people think recruiting is a field one can “break into” more than one can break into law, medicine, science, or engineering. So old-school recruiting practices remain strongly rooted in reviewing a job description, interviewing a hiring manager, and using HiHowAreYa interview questions. The usual result is unclear job skills that must be inferred from resume data or taken at face value. In the final run, candidates are often compared with one another instead of to the job, and usually the ones who walk and talk most like current employees get hired. As a result, only about half of new hires fail to achieve high performance. A few months or weeks later, enterprising internal recruiters send out one-size fits-all smile sheets to hiring managers expecting to get honest feedback from people who are highly motivated not to admit any kind of bad hiring decision. Is it any wonder why status quo metrics seldom tell you very much?

Useful Metrics

Let’s start with identifying specific skills that lead to results. This is done when a job analyst interviews job holders and their managers about what it takes to do the job; what leads to successes; and, what leads to failures. This is a learned art because the analyst must know how to listen for clues associated with competencies that either precede or are associated with performance. For example, a long story about a successful project may involve learning, decision making, interpersonal skills, planning, and so forth. It’s up to the analyst to listen to each story to isolate the key competencies against which new candidates should be measured.

Specific competencies serve as the least common denominator against which all metrics are collected. For example, metrics will tell us if our job analysis missed anything; if the job changed; if our tests and measures (i.e., that includes interviews) were accurate and trustworthy; what needs to be modified; if pass/fail scores were set too-high or too-low; or, if any of the tests and measures adversely impacted protected groups. And, yes, this includes manager-employee fit.

Article Continues Below

Foggy Bottom

Military experts often refer to the fog of war. That is when everything is happening at once and nothing is clear. We can use the same term for the fog of performance: foggy because managers’ ratings are often part opinion and part fact; hard performance numbers often conflict (e.g., customer satisfaction and time per call are often negatively related); and, results-data often occur long after competencies are applied. Nevertheless, a skilled analyst knows how to gather highly useful and actionable data that can be used to improve and modify quality of hire.

If you don’t know specifically what to measure, how to measure it, and what to do with the results, then metrics can become a muddy mess. It takes clear job analysis, validated tests, structured interviews, and effective manager feedback to make hiring metrics work.

  • http://www.HireBetter.com Gina Kleinworth

    Metrics are worth the work. If you do it right- you can avoid the cost of the mis-hire. That money can then go into those A-Players those metrics have helped you find.

  • Keith Halperin

    IMHO, metrics are useful if they:
    1) Are structured and presented to enhance the interests of the presenter(s)
    2) Aren’t taking up more than 5% os someone’s time to compile and create or are outsourced to someone whose time is less expensive/valuable.

    They should NOT be used in an attempt to get senior executives to take HR/Staffing seriously- if you have to convince someone they should let you play, you’ve already lost.
    The language of power is “power,” not metrics. A lot more people got ahead studying office politics than through studying how to do a good spreadsheet. That being said, if a good spreadsheet makes you look good in the eye’s of the boss, metric away……

    Cheers,

    Keith

  • Dave Pollock

    Hmmm. Not a word about goals… What about outcomes? I really don’t care what the competencies are or what the test-du-jour (sorry – ‘evaluative indices’) tell me. If, at the end of the day, the departmental goals aren’t met – the metrics are not only meaningless, they’ve wasted time and resources.

    We seem to conveniently forget we’re talking about hiring people, not manufacturing widgets. There’s a reason for getting an appropriate sample size in any study. And when we’re talking about studying the outcomes of human behavior (result of the hire)… in an uncontrolled environment (your workplace) I have 2 words for that one: Good Luck. How many hires will it take – IN EACH POSITION -to support (or not) your evaluative/hiring methodologies? Too few? A waste of time/resources. Too many? An even bigger waste.

    Define your butt off when it comes to competencies and validity on behavioral testing but when it comes to the outcome of the hire… (departmental goals) what did they accomplish? A perfect match on competencies and great scores on behavioral testing! Congrats! You’re fired.

  • http://www.paulmnorman.com Paul Norman

    Without good metrics in place, performance becomes measured by emotion (and that’s a dangerous place to be because emotions can change quickly). It’s also key to understand the goals of the organization. Some companies are more focused on speed or volume than quality; some companies cherish the candidate experience more than speed. Good metrics on speed, quality, and cost can help a company assess if their focus is the right strategy.

  • http://www.ScientificSelection.com Dr. Wendell Williams, MBA, Ph.D

    To Keith and Dave.

    Metrics is just another word for trustworthy feedback and information. Wouldn’t it be nice to know, for example, if XYZ was job-critical, that Billy Bob scored high on XYZ but failed to apply XYZ on the job? Isn’t that the kind of information someone might want to investigate?

    A lack of large numbers? This is not a statistical validation study, it is structured actionable feedback.

    Finally, anyone who believes establishing goals without also clarifying how to achieve them is destined to re-live MBO. Yes, goals are important, but so is the process of getting there.

    Opinions are free, but it often helps to think them through first.

  • Dave Pollock

    Thanks for the insight about opinions and thinking. Without a superior intellect, I never could have thought of that.

    Speaking of goals, where do you refer to them in the article? Oh… you don’t.

    XYZ may indeed be job-critical. So contact Billy Bob. Just don’t imply that you mention common sense recruiter follow up in your article. You don’t.

    Establishing goals means establishing objectives to get there. Or, if you prefer, “clarifying how to achieve them [goals]…”. Doing this where clearly defined, measurable results can be quantified (in the workplace) is fine for the floor manager. Your article is titled HIRING metrics… and proceeds to subtly bash the recruiter for trying to get feedback from the floor manager… who should be able to use his/her own “clearly defined, measurable results” to determine the worth of the hire!

    Unless the recruiter is also responsible for employee evaluation and retention we are talking about the relationship between two processes: Hiring and whatever goes into productivity – as measured by goal achievement. And the only thing said about it is, “Measuring quality at the end of the line tells you whether the product is good or bad, but almost nothing about how to fix it.”

    I maintain that neither does your article. Unless you establish a valid (<- pesky word, isn't it?) relationship between HIRING metrics and WORKPLACE goals, I don't see the value in "hiring metrics" alone. In fact, your suggestion that the recruiter "…evaluate demographic impact, gather performance evidence, work to control the variability, and so forth" looks suspiciously like an attempt at validation… in the workplace, not in the HIRING process.

    Advice and How To's are free, but it often helps to think them through first.

  • http://www.ScientificSelection.com Dr. Wendell Williams, MBA, Ph.D

    Ouch!

  • Keith Halperin

    @Dr. Williams:
    I wish to disagree with your assertion:”corporate recruiters clamor for metrics”. Perhaps their supervisors and superiors do (and Dr. Sullivan certainly does), but in my experience there has been little clamor for such among those of us who actually do recruiting.

    @Paul:
    While many pieces of data are interesting and or useful, that doesn’t mean the cost of collecting, processing, and organizing it is worth the expenditure of time and resources to do so. Thus, the 5% Rule:
    It shouldn’t take more than 5% of your time to gather and process metrics. If it does, the process should be “down-sourced”(streamlined), through-sourced (automated) or out-sourced (sent away).

    @Dave:
    I agree with you- evaluating hiring quality and and retention is typically beyond the scope of a typical recruiter’s duties, i.e., “We put butts in chairs, we don’t measure how comfy they are or how long they’re sitting there.”

    Cheers,

    Keith