Latest / The Science of Leading / Models Suggest, Humans Decide

Transcript

Models Suggest, Humans Decide · The Science of Leading

• 0:00
  Welcome back to Science of Leading, I'm Claire Monroe with Edwin Carrington.
• 0:03
  Edwin, we've spent two episodes making the case for data in people decisions.
• 0:08
  Today I want to push on the worry that comes up every time this topic gets raised,
• 0:12
  that all of this quietly dehumanizes the process.
• 0:15
  It's a legitimate worry, Claire, and the answer isn't to retreat from data.
• 0:19
  It's to be precise about what data is actually for.
• 0:23
  A model suggests.
• 0:24
  A human decides.
• 0:26
  The moment that order flips, you've stopped being data driven and started being data
• 0:31
  blind, just with better graphics.
• 0:33
  Can you make that distinction concrete?
• 0:35
  Because "models suggest, humans decide" sounds good as a phrase but I want to know
• 0:40
  what it looks like in an actual room.
• 0:42
  Take a predictive flight-risk model flagging an employee.
• 0:45
  The model's doing exactly what it should, surfacing a pattern a manager might miss.
• 0:50
  But if that manager happens to know this employee just received a life-changing
• 0:54
  personal opportunity, something no dataset captured, that context has to be allowed
• 1:00
  to override the flag.
• 1:01
  The model did its job.
• 1:03
  The human still has to finish it.
• 1:05
  There's a promotion example in the piece that I think captures this even more
• 1:08
  sharply, two people, one clear data-favored choice, and it doesn't go the simple
• 1:13
  way.
• 1:14
  An assessment strongly favored promoting one person, Person A,
• 1:17
  based on the behavioral and performance data.
• 1:20
  But Person B held critical domain expertise the data couldn't fully weigh,
• 1:25
  and was navigating a family health situation at the time.
• 1:28
  The resolution wasn't picking a side and ignoring the other.
• 1:32
  It was a phased promotion plan that honored both realities,
• 1:36
  informed by the data, but not enslaved to it.
• 1:39
  That's such a specific example, I want to flag for anyone listening that it's
• 1:42
  illustrative, the piece doesn't name a real company here.
• 1:44
  But the underlying lesson still holds regardless.
• 1:47
  It does, because the lesson isn't about that specific pair of people,
• 1:51
  it's about the shape of the decision.
• 1:54
  Data should analyze.
• 1:56
  It shouldn't dictate.
• 1:58
  The moment leadership treats an assessment score as a verdict instead of an input,
• 2:02
  you've built a system that's technically sophisticated and practically worse than
• 2:07
  the gut-feel process it replaced.
• 2:09
  So what actually keeps that line intact, practically, inside a real HR process?
• 2:14
  Because "use judgment" isn't a policy, it's a hope.
• 2:17
  Real safeguards, not just good intentions.
• 2:21
  Diverse review panels, so no single perspective quietly becomes the default filter.
• 2:28
  Fairness checks on the models themselves, specifically to catch algorithmic bias
• 2:33
  before it compounds at scale.
• 2:35
  And clear documentation of the actual decision criteria, so a choice can be
• 2:41
  explained and defended after the fact, not just felt to be reasonable in the moment.
• 2:45
  Documentation feels like the piece people skip most, because it's paperwork,
• 2:50
  right up until someone needs to explain a decision six months later and can't
• 2:53
  reconstruct why it was made.
• 2:55
  Exactly, and that's precisely when the absence of documentation turns into a real
• 3:00
  liability, legally and organizationally.
• 3:03
  The habit of writing down why has to exist before you need it,
• 3:07
  not after.
• 3:08
  So if someone's listening and worried that getting more data driven means getting
• 3:11
  less human, what's the actual reassurance?
• 3:13
  The reassurance is that the goal was never to remove the manager from the room.
• 3:18
  It was to give that manager better evidence to bring into it,
• 3:21
  and the discipline to notice when the evidence and their own judgment are pointing
• 3:25
  in different directions.
• 3:27
  That tension is the system working, not the system failing.
• 3:31
  If you want a real behavioral starting point to bring into your own next people
• 3:35
  decision, OAD's free Top Performer Profile is about seven minutes,
• 3:39
  results in twenty-four hours, no credit card.
• 3:41
  That's O-A-D dot A-I.
• 3:43
  Let the data speak first.
• 3:45
  Just don't let it have the only vote.
• 3:47
  Perfect place to close out this series.
• 3:49
  Thanks, Edwin.
• 3:49
  Always, Claire.