A Geek’s Guide To Staffing

They say if the only tool you have is a hammer, everything looks like a nail.  Since I started my professional career as a systems reliability engineer, everything looks like a system, including staffing models. Guilty.

Here’s my geeks-eye view on why hiring and retaining top talent is so important, and why “settling” is so detrimental.

The Proof

In the case of system reliability, we think in terms of Probability of Success, with perfect being 1.0000.

If you have a system built on two elements, who’s Probability of Success are both .9999, you have a system Probability of Success of .9999 x .9999 = .9998

But what happens when one of those elements is not very reliable, and has a Probability of Success of .7000 (70%)

Then the system Probability of Success drops to .9999 x .7000 = .6999

Even a system with 4 near perfect elements and one running at 70% you end up with .9999 x .9999 x .9999 x .9999 x .7000 = .6997

Note that system reliability or Probability of Success is always less than that of the lowest critical component.  The chain is only as strong as the weakest link.

The Corollary 

So if you think about the different skills required of any team member, when they are weak in one area then their overall performance is pulled down.  Further, if you have one person underperforming, the whole enterprise is pulled down. It doesn’t matter how many really good other skills a team member has or how may other really good team members you have.  If something important isn’t working, it pulls the whole system/organization down to that level.

In the systems world there are three approaches to reducing the impact of a weak component: redesign, replacement, redundancy.  You either redesign the component for better performance, find a higher performing version of the same part, or run the low performer in parallel mode (at twice the parts count and cost).

(the parallel mode goes something like this .9999 x < .7000 || .7000 >  = .9099  not quite .9998, but a lot better than .6999)

In the case of organizational effectiveness the corollary is training, replacement, staff increase with staff increase being least efficient yet most prominent method.

Why is it most prominent? It’s easiest.

Training takes time and requires someone with both the functional skills and training skills not to mention a receptive student.

Replacement is always difficult and awkward, especially when the incumbent is performing exceptionally well in other areas of responsibility.

So someone else is hired to fill the gap.  This is hugely inefficient because the economic order quantity is 1 person, who inevitably has overlapping skills with the incumbent and/or other team members.  And since Probability of Success never exceeds 1.000, this capacity is wasted.  Further, in times of economic stress, the head-count becomes unsustainable/non-competitive, forcing reductions.  Reductions lead to a return to underperforming at a time where maximum performance is a matter of survival. So underperforming yields more economic stress, yields more staff reduction…. the traditional death spiral.

The Pudding

So, if you’re in a hiring situation and thinking of “settling”, do the math.

If you’ve got someone who’s not firing on all cylinders, do the math.

Advertisements

One thought on “A Geek’s Guide To Staffing

  1. This math explains why you hire people who have critical thinking, reasoning and learning capabilities, not to mention passion and vision, rather than (just) functional capabilities. Get capable, agile people to fill spots and don’t settle for someone who may be great functionally but that’s all they know.

Share your thoughts

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s