Tell Me the Talent-acquisition Metrics You Invented (Even the Bad Ones)

As Jessie Garcia and I were prepping for our leadership workshop at the Spring ERE Conference in San Diego on April 18, we had a spirited discussion around the “best metrics for influencing your C-suite.” To say we are both metrics-driven is a bit of an understatement (you might say we are metrics freaks).

Jesse wrote a thought-provoking piece using a framework of 1) labor correlations, 2) service excellence & quality, 3) cycle time, 4) conversions (“batting average”), and 5) brand & retention.  Read Jessie’s article here. Readers added great comments making the case for 6) revenue (or other financial) correlations.  These six themes provide a model for you to evaluate how well you are expressing your recruiting results by speaking the language of business rather than “recruiter speak.”

We were also reminiscing about the importance of not just recycling the same ‘ol metrics but looking for opportunities to innovate and bring new thinking to our field and express our value in new ways.  One example of this was the creation of “Time to Find.” And here is its story …

Definition

Time to find is the time from when a requisition is opened (or an “order” is placed) to the time the winning (hired) candidate is presented to the hiring manager. If you have a strong talent pool and/or are a great sourcer, Time to find could be hours or a couple days. If you are a post-and-pray recruiter, then you’re probably looking at a time to find measured in weeks or months.  Recruiters always complain that they don’t control time to fill. However, they totally control time to find.

What Spurred Its Creation

I was facing an interesting quandary. As the recruiting leader in a large healthcare organization, I was hearing from operations that “it takes your recruiters too long to get us nursing candidates.” I was hearing from recruiters that “hiring managers sit on my candidates and aren’t responsive.” It seemed unlikely to me that both accusations were true, so I started thinking … how can we get to the factual reality? How can I get to the truth?

The answer was possible with our ATS. Most ATSs time-stamp recruiters’ actions, including when they send each candidate to a hiring manager. So, we started analyzing data. Unfortunately for the recruiters, at that point in time, it was taking us 3-4 weeks on average to send any candidates to hiring managers, and we discovered that on average we weren’t sending the “winning” candidate until day 65. Our overall time to fill was about 73 days, so it didn’t seem like hiring managers sitting on candidates was the real problem — our less-than-stellar sourcing was the culprit.

The Impact

As I mentioned above, when we started measuring Time to Find for nursing hires, we were at about 65 days. Over a two-year period, we cut this by half, and were striving toward getting under 20 days — a very aggressive goal in a tight RN labor market. (Side note: On our internal recruiter scorecard, the amazing Jessie Garcia was always at or near the top of scorecard for shortest time to find).

Article Continues Below

These sourcing improvements brought our overall cycle time down to under 40 days, and greatly increased our hiring manager and candidate satisfaction survey scores. We took pride in our sourcing competency. And we certainly were not a corporate recruiting “black hole.” There was a huge element of fulfillment, too, by creating something that no one else seemed to be using. So what started as a mission to discover factual reality and truth ultimately lead to creating a metric that drove sourcing excellence.

Your Inventions?

So, let’s hear from you. Please comment below and tell us about a metric you “invented” and why. Or even something you tried to measure that really never took off — we can still learn from “failed” experiments. You never know when your ideas and experiences may be applicable to another’s situation, or you may spur new thinking and ultimately new inventions. I’m 99 percent sure that Jessie will be happy to buy a coffee or an adult beverage for those ideas that are the most creative! I’ll chip in, too. See you in San Diego

 

image from bigstock

Tony D. Blake, M.S., SPHR, is a dynamic senior leader who is known for aligning “people strategy” with business strategy to drive outstanding organizational outcomes. One CEO described him as an “Intrapreneur.” He has over 30 years of professional experience, including 20 years in human resources leadership.  His industry experience includes healthcare, financial services, telecommunications, and aerospace. In March 2016, he founded Strategic Impact Group.
 

Topics

6 Comments on “Tell Me the Talent-acquisition Metrics You Invented (Even the Bad Ones)

  1. At a previous job, I partnered with a member of our HR Analytics team to come up with a better measure of Recruiter workload, something better than the typical “Requisition Load.” After creating a regression model (in full disclosure, he did the heavy statistical lifting since he’d just finished his Industrial/Org psych degree and mine is almost 2 decades into gathering dust). We took into account job, job family, location, pay and a variety of other factors. But when all was said and done, the biggest predictor of recruiter workload was the grade of the job. In other words, filling a sr. director level req was the equivalent of filling 5 regular individual contributor reqs. That’s pretty commonsensical, but it was nice to see the data bear it out.

    1. Thanks very much, Ben, for your comment. Using less sophisticated methods, we tried achieving something similar at DaVita, but never came up with anything that seemed very credible (and thus never got any internal traction). Based on your experience, for organizations that don’t have a solid salary grade structure it seems one could also create a meaningful “recruiter load factor” by looking at the base compensation of the position (since one would assume that pay is highly proportional to grade). Oh by the way, I’ve used “total compensation recruited” as a recruiter PRODUCTIVITY FACTOR on the back end. Great thought and comment…thanks again, Ben!

  2. I have measured something like your “time to find” and found it quite helpful in motivating recruiters to work quickly and also Managers to take action on good resumes.

    Another we have measured is the % of requisitions that our Managers insist go straight to Agency (Contingency Search) rather than allow our TA model – which is RPO supported – to have a chance to fill. We found that we were still filling that requisition with our own candidate (versus an Agency candidate) more than 50% of the time… and used that information as leverage to reduce the number of requisitions that were going straight to Agency. Eventually it was (mostly) the RPO Recruiter – not the Manager – making the recommendation for going straight to Agency on the jobs that we truly could not fill on our own.

    1. Thanks, Allen! Great observation and comment for orgs that use RPOs. AND, this metric could also be very applicable to in-house TA functions that send a decent volume of reqs out for search (perhaps unnecessarily). Thanks for the “time to find” validation, too. All the best…Tony

  3. 2 things from my days at Cummins:

    1) We partnered with Workforce planning to combine attrition analysis with projected revenue growth and revenue per employee to predict where and what buckets of heads we would hire – because they business never got this to us until we were already behind. It helped tremendously in dense job areas, like supply chain, manufacturing engineers, and even HR + Finance.

    2) We created (through excel manually) what we called a “weekly change report” for recruiting operations. You can cut and slice data all you want in Taleo, but you can’t show data over a time range and how it changes. We were then able to tell how our Reqs changed week over week, and ultimately how productive each resource was. The static reports simply can’t show this.

Leave a Comment

Your email address will not be published. Required fields are marked *