Dubious Data™️: Recruiting Metrics Edition

Probably the only edition I’ll do but it sounds more official that way.

As much as I love numbers, I hate numbers. When you trick yourself into paying attention to the wrong stuff – because some metrics sound plausible – you end up making dumb decisions.

Example: I mentioned this in a blog post (or LinkedIn post, depending on where you read my nonsense) a week back: Time-To-Fill.

Sounds important. And it is! As a hiring metric, not a recruiting metric. Time-To-Fill measures the speed and effectiveness of your hiring process. Two things that drag this down are:

1. A slow hiring process. Hurry up, kids.

2. Declined offers. They just weren’t that into you. Start over.

But it usually gets lumped in with top-of-the-funnel recruiting metrics, so my go to answer is “5-10 days plus the length of your interview process.”

As Kelli Hrivnak pointed out: the metric you’re looking for is Time-To-Source. How long it takes to find that first slate (short list) of candidates.

Both are important, but not the same. If you’re trying to reduce Time-To-Fill but keep harping on finding candidates faster, you’re going nowhere.

What else falls into that bucket?

👉Number of Submissions

What’s the right number here? 1 sounds amazing. Can’t beat that efficiency. Unless you want a broad perspective of the candidate market, maybe go with 10. Or 20. But that’s a loooot of time. So 3? 5?

No matter what, it’s arbitrary. And it comes down to how well you know that skill you’re hiring for.

👉Size of the “Rolodex” of candidates.

Shout to anyone who knows what a Rolodex is. Bonus points if you ever owned one.

I blame crappy staffing sales people for this one. It goes like this:

Client: “I need to hire XYZ skill set.”

Sales hack: “Yeah I’ve got 10 of those guys, no problem.”

They had the search results before even doing a search! Uncanny.

In reality: everyone has a massive database nowadays. We’re overloaded with public data sources, damn near everyone in the world can be found. Having someone’s name and number doesn’t mean they’re a valid candidate; they have to be available and want the job. They have to be…(wait for it)…recruited.

What you really need the gauge is someone’s ability to *build* a Rolodex quickly. 

👉Source of Candidates

Who cares?

👉Percentage of candidates currently employed.

This is problematic for 2 reasons:

1. Bias against people who aren’t working. (Another rant for another day.)

2. People who *are* working always want more money. No one takes lateral moves unless it’s a once in a lifetime opportunity. (Unless you’re curing cancer, it’s probably not.)

👉Interview-to-Hire Ratio

The idea of evaluating targeting based on interview-to-hire metrics seems appealing, but it gets skewed to hell based on:

1. Searches where the requirements are refined/changed as the process goes on.

2. The “need” for some orgs to always talk to 4+ candidates before making a hiring decision. (Which is sometimes justified / sometimes ridiculous. Another rant for another day).

Perhaps looking at the percentage of candidates who make it through multiple interviews instead.