#125 - Dr. John Boudreau & Pete Ramstad - Why Beyond HR is More Relevant Than Ever

WEBVTT
00:00:00.000 --> 00:00:17.360

00:00:17.278 --> 00:00:20.318
Welcome to Directionly Correct, a PNL podcast with Colin Scott.
00:00:20.618 --> 00:00:25.078
Today's guest, John Boudreau, Senior Research Scientist at the Center for Effective
00:00:25.078 --> 00:00:30.638
Organizations at USC, and Pete Ramstad, former CHRO and Strategy and Organizational Consultant.
00:00:30.838 --> 00:00:34.638
Hey, Directionly Correct fans. If you're looking to stay connected to the podcast,
00:00:34.978 --> 00:00:39.078
make sure to subscribe to our Substack newsletter, watch our videos on YouTube,
00:00:39.078 --> 00:00:41.838
or listen to the podcast on your app of choice.
00:00:42.418 --> 00:00:46.958
This podcast is dedicated to you to help educate and entertain you on how to
00:00:46.958 --> 00:00:48.538
effectively do people analytics.
00:00:48.838 --> 00:00:52.738
We are a listener-supported podcast, so please consider becoming a patron.
00:00:52.938 --> 00:00:56.898
If you'd like to support the podcast by becoming a patron, you can find a link
00:00:56.898 --> 00:01:02.098
to sign up in the show notes or at patron.podbean.com slash directionallycorrect.
00:01:02.218 --> 00:01:03.418
Thanks for being a listener.
00:01:04.178 --> 00:01:08.418
All opinions are our own and do not reflect those of any other organization.
00:01:08.418 --> 00:01:13.518
John, I'm so tickled with what you're wearing today in terms of the callback.
00:01:14.138 --> 00:01:14.498
Right.
00:01:14.818 --> 00:01:17.258
Yeah, I think how you found out about the podcast.
00:01:19.038 --> 00:01:21.678
How did you find out about the podcast? What's that?
00:01:22.358 --> 00:01:25.518
Well, maybe I'll tee up the story and then John can kind of take it from there.
00:01:25.518 --> 00:01:27.778
But I can't remember whose episode it was.
00:01:27.898 --> 00:01:31.258
But at one point I was talking about, because in my mind, I see a person like
00:01:31.258 --> 00:01:33.458
John Boudreaux is like a famous person.
00:01:33.638 --> 00:01:38.118
Like in my world, John, John is famous. And so I'd seen him speak at a conference
00:01:38.118 --> 00:01:41.338
and he was wearing like a cowboy jacket is what I felt like.
00:01:41.438 --> 00:01:45.678
And we talked about on the podcast, come to find out, like I got an email thread
00:01:45.678 --> 00:01:49.858
sent to me from a few folks that had gotten around to John and we had talked
00:01:49.858 --> 00:01:51.718
about it and found out about it.
00:01:51.818 --> 00:01:56.538
So I appreciate you being a good sport, John, showing up in the cowboy vest for the podcast.
00:01:57.238 --> 00:02:01.658
Yeah. Well, it appears to be my defining characteristic, like I said, these days.
00:02:01.838 --> 00:02:07.498
So I was delighted. I wrote people back and said, yeah, this is the arc of my career.
00:02:08.018 --> 00:02:12.238
I used to be famous for academic articles, and then I was famous for leadership
00:02:12.238 --> 00:02:16.678
books and articles, and now I'm famous for what I wear at the conference.
00:02:17.358 --> 00:02:21.258
There's no such thing as bad publicity, right? I guess so. It's all the same.
00:02:21.518 --> 00:02:22.198
We'll see after this.
00:02:22.458 --> 00:02:28.658
John was present at a workday conference, Connected Commons,
00:02:28.898 --> 00:02:31.738
that someone spilled coffee on me.
00:02:31.738 --> 00:02:35.878
That's the famous one where like someone like he was not in the room at the
00:02:35.878 --> 00:02:40.798
time but he was there I definitely remember this in Pleasanton some time ago it's been a long time
00:02:41.899 --> 00:02:46.979
Well, John, you kind of queued it up in terms of like the progression of your fame and your career.
00:02:47.319 --> 00:02:51.959
I think one of the reasons why both of you being guests on the podcast today
00:02:51.959 --> 00:02:56.599
is such an honor for us is I saw the book that you all wrote,
00:02:56.779 --> 00:03:01.119
Beyond HR, as a very formative book in my career.
00:03:01.119 --> 00:03:07.339
But I imagine in many people's careers in terms of how they think about talent
00:03:07.339 --> 00:03:14.319
and kind of the equation of how we think about quantifying the value of talent
00:03:14.319 --> 00:03:17.439
to organizations and what role variation plays in that.
00:03:17.439 --> 00:03:22.959
And I know, Pete, you have a background in kind of economic analysis and how
00:03:22.959 --> 00:03:25.099
that frames your lens on the world.
00:03:25.239 --> 00:03:30.899
But can you guys talk to us a little bit about that book? Because I feel like
00:03:30.899 --> 00:03:32.879
it is so relevant at this moment.
00:03:33.699 --> 00:03:38.719
And I'm not sure anybody's paying attention to this type of research at this moment either.
00:03:38.959 --> 00:03:42.299
And so maybe that can tee you guys up to talk about it.
00:03:42.699 --> 00:03:46.819
Well, maybe I'm just going to start the story of a little bit of how the book began to come out.
00:03:46.819 --> 00:03:50.199
Pete and I, we had a whole bunch of common connections,
00:03:50.499 --> 00:03:55.479
many of which went through the board of a company called Personnel Decisions International,
00:03:55.499 --> 00:04:02.599
which was kind of an executive assessment board, very much anchored in the research
00:04:02.599 --> 00:04:07.379
at the University of Minnesota and people who were publishing there at the time.
00:04:07.379 --> 00:04:12.179
And back in the late 90s, several different contacts, Wayne Casio,
00:04:12.319 --> 00:04:16.039
George Milkovich, and others were saying to me, John, you've got to meet this
00:04:16.039 --> 00:04:19.919
guy, Pete Ramstad, who's working for Personnel Decisions International.
00:04:19.919 --> 00:04:21.779
And I said, okay, what's he like?
00:04:21.799 --> 00:04:26.379
And they said, oh, he's a CFO and accounting guy and all that.
00:04:26.459 --> 00:04:28.779
And I said, well, okay, I guess that's all right.
00:04:28.899 --> 00:04:32.879
I can certainly chat with a finance and accounting person.
00:04:33.459 --> 00:04:37.239
And Pete and I got to talking, and one of the first things he said was,
00:04:37.379 --> 00:04:43.239
you know, John, we've got, at that case, it would be file cabinets full of assessments of executives,
00:04:43.519 --> 00:04:47.379
thousands and thousands of these assessments that we do based on rigorously
00:04:47.379 --> 00:04:51.499
designed psychometric tests and simulations and other things like that.
00:04:51.599 --> 00:04:55.199
And he said, you know, I come from finance, and I'm familiar with bond ratings,
00:04:55.259 --> 00:05:00.039
and the rating of a financial bond is the prediction of the future performance
00:05:00.039 --> 00:05:02.059
of an asset that's uncertain.
00:05:02.419 --> 00:05:05.579
You can't really say exactly what a bond will be worth in the future.
00:05:05.759 --> 00:05:10.619
But we have these things called bond ratings, triple A, double A, B, right?
00:05:10.979 --> 00:05:17.439
And those ratings are a kind of rigorous way to predict the future value of
00:05:17.439 --> 00:05:19.319
this unpredictable asset.
00:05:19.619 --> 00:05:23.719
And he said, it seems to me executive assessments are like that.
00:05:24.345 --> 00:05:28.385
And this information we have in our file cabinets could be seen as bond ratings
00:05:28.385 --> 00:05:33.505
on executives, an uncertain asset with future value.
00:05:33.745 --> 00:05:38.265
And I was kind of hooked on P at that point.
00:05:38.285 --> 00:05:43.525
And as we've talked about, Cole in particular, there's a great deal of that
00:05:43.525 --> 00:05:48.745
kind of finding a metaphor in a discipline that leaders already understand.
00:05:48.745 --> 00:05:52.265
I call it a place where leaders are already using their smart brain.
00:05:52.945 --> 00:05:58.485
So if you get them started thinking about financial assets, they lock into a
00:05:58.485 --> 00:06:02.345
financial set of decision models, a financial set of mental models.
00:06:02.465 --> 00:06:03.545
Those are pretty rigorous.
00:06:03.805 --> 00:06:06.325
They've learned them. They're held accountable for using them well.
00:06:06.585 --> 00:06:11.445
And so when they think about money, they've kind of been expected and trained
00:06:11.445 --> 00:06:16.305
to think about money in a certain rigorous way based on principles that come
00:06:16.305 --> 00:06:17.985
to them from this finance discipline.
00:06:18.865 --> 00:06:23.365
And so one of the tricks I've often tried to use is to find a metaphor like
00:06:23.365 --> 00:06:28.905
that, get executives into their smart brain, and then talk to them about something
00:06:28.905 --> 00:06:30.185
like executive assessment.
00:06:30.425 --> 00:06:34.485
And you can do it for all kinds of things in HR, performance management, rewards.
00:06:34.845 --> 00:06:38.665
Right now, there's a great deal of discussion about what do we do under uncertainty?
00:06:38.665 --> 00:06:43.505
And there's plenty of tools from the finance world about how you hedge and how
00:06:43.505 --> 00:06:46.845
you create a portfolio that has a certain payoff under many,
00:06:46.965 --> 00:06:50.525
many different future scenarios, even though you can't predict any one of them.
00:06:50.745 --> 00:06:54.285
So that's my introduction to PEED. And that was the beginning of a lot of work
00:06:54.285 --> 00:06:58.585
we did together that ultimately produced Beyond HR in 2007.
00:06:58.865 --> 00:07:03.465
And to show you how naive I was, I started building some of these models on
00:07:03.465 --> 00:07:07.845
trying to work on PDI showing value, just from my economics perspective.
00:07:08.585 --> 00:07:13.185
And one of the IL psychologists, one of the dozens and dozens of well-explained
00:07:13.185 --> 00:07:14.645
IL psychologists who had a pediatrician
00:07:15.550 --> 00:07:21.030
Oh, you're going down into the utility trap. You're going into the utility rabbit hole.
00:07:21.850 --> 00:07:25.330
And he said, you ought to read these papers by Wayne Cascio.
00:07:25.950 --> 00:07:29.190
I go, oh, I know Wayne. He's on the board. What does he do? I had no idea.
00:07:30.570 --> 00:07:35.310
I knew Wayne personally. I'd interacted with him several times as a board member
00:07:35.310 --> 00:07:40.750
at that time of the subsidiary PDRI, but I had no idea. And so the whole introduction
00:07:40.750 --> 00:07:43.930
to utility analysis and the integration of the I.O.
00:07:44.090 --> 00:07:51.150
World at that time with the economic world was kind of my, through the lens of being at PDI.
00:07:51.430 --> 00:07:54.210
I happened to be a graduate student in accounting theory at that time.
00:07:54.630 --> 00:07:59.910
And so I just happened to be in the theory of economics and accounting and by
00:07:59.910 --> 00:08:01.470
happenstance, immersed in an I.O.
00:08:01.530 --> 00:08:05.110
Psychology firm. I think it was going to be a temporary job for two years while
00:08:05.110 --> 00:08:06.050
I went to graduate school.
00:08:06.650 --> 00:08:09.310
They were a client of mine for my CPA firm days. That's how I knew them.
00:08:09.390 --> 00:08:13.050
That's how they knew me. But anyway, I ended up spending 16 years at PDI.
00:08:13.490 --> 00:08:17.210
I really had a really tremendous experience with a lot of really,
00:08:17.370 --> 00:08:20.010
really smart, smart people inside and outside of PDI.
00:08:20.230 --> 00:08:24.030
It feels like the world's changing so quickly that these sort of things are
00:08:24.030 --> 00:08:28.010
even more important, like trying to derive the value of a specific employee,
00:08:28.090 --> 00:08:29.850
the specific job functions, this sort of stuff.
00:08:29.850 --> 00:08:33.930
So, like, how has, like, the recent advent of, like, say, like,
00:08:34.030 --> 00:08:39.170
Gen.AI or, like, the oncoming of Gen.AI going to impact these sort of assessments
00:08:39.170 --> 00:08:42.030
on, like, how we select people and what we value?
00:08:42.610 --> 00:08:46.710
Yeah, let me just hit my crackhead because here's what I'm excited about and
00:08:46.710 --> 00:08:47.950
here's what I'm worried about, okay? Okay.
00:08:48.652 --> 00:08:53.152
And the other thing is I like to look at historical analogies, okay?
00:08:53.712 --> 00:08:57.532
And I'm not sure we're well enough organized to take advantage of AI well.
00:08:57.972 --> 00:09:01.552
And I would say that in the 70s, we weren't well enough organized to take advantage
00:09:01.552 --> 00:09:04.272
of the new enterprise software systems that were coming through.
00:09:04.452 --> 00:09:11.772
So, you know, all of a sudden, you know, HR and finance got great software tools in the 70s and 80s.
00:09:11.892 --> 00:09:15.672
You know, the difference was finance had been doing consolidations and financial
00:09:15.672 --> 00:09:19.992
reporting and stuff for 50 years. so they had well-horned practices they could automate.
00:09:20.232 --> 00:09:23.792
And the joke I say is that all of a sudden, HR starts brainstorming turnover
00:09:23.792 --> 00:09:28.852
codes because they got to code turnover. They didn't ever code turnover before.
00:09:29.092 --> 00:09:34.612
And one of the problems we have is that we try to do, this is something I feel
00:09:34.612 --> 00:09:37.452
really strongly about. I want to differentiate some of the work that John and I do.
00:09:37.632 --> 00:09:39.292
We kind of link too much to the
00:09:39.292 --> 00:09:43.812
accounting rather than the essence of the economics behind the accounting.
00:09:44.352 --> 00:09:48.272
Trying to apply accounting measures to HR will not work well at all.
00:09:48.472 --> 00:09:50.612
I can just tell you a thousand ways that it goes wrong.
00:09:50.932 --> 00:09:55.212
But if you distill the theory behind why the accounting is that way,
00:09:55.812 --> 00:10:01.712
then you get down to the raw elements that give you a lot of insight into how HR can do a better job.
00:10:02.052 --> 00:10:03.572
What are we talking about there?
00:10:03.972 --> 00:10:07.872
I'll give you an example. I hear people say, oh, people are a most important
00:10:07.872 --> 00:10:10.012
asset if only people were on the balance sheet.
00:10:10.572 --> 00:10:14.732
Well, the essence of strategy is to be as balance sheet light as possible.
00:10:15.132 --> 00:10:18.412
So people are on the balance sheet. Nobody wants plants on their balance sheet.
00:10:18.512 --> 00:10:19.332
Nobody wants inventory.
00:10:20.292 --> 00:10:25.152
Economic value is driven by driving down the value of your assets on the balance sheet.
00:10:25.892 --> 00:10:29.812
And most economic and most balance sheet restructurings are reducing assets.
00:10:30.494 --> 00:10:34.694
People want to be asset-like companies. So it sounds great.
00:10:34.814 --> 00:10:37.354
Is that any better than being a liability, though, which is,
00:10:37.414 --> 00:10:38.574
I think, what they are right now?
00:10:39.034 --> 00:10:40.254
No, they're holding costs.
00:10:40.554 --> 00:10:41.754
They're holding costs? Okay.
00:10:41.874 --> 00:10:46.874
Yeah, they're not a liability. There's no future cost to people that's associated with that.
00:10:47.014 --> 00:10:49.974
I mean, you might have some pensions in some certain kind of situations where
00:10:49.974 --> 00:10:52.914
you've made certain contractual obligations financially.
00:10:53.394 --> 00:10:58.494
But the idea is we don't have brands on the balance sheet, and nobody has any
00:10:58.494 --> 00:10:59.714
trouble investing in brands.
00:11:00.494 --> 00:11:02.194
Yeah. What's the value of Coca-Cola?
00:11:02.194 --> 00:11:04.714
They show up in the stock price, though, right? What's that?
00:11:05.294 --> 00:11:08.594
Those show up in the stock price, though, right, with the differentials between
00:11:08.594 --> 00:11:12.254
how much a firm is evaluated versus what they're trading at.
00:11:12.514 --> 00:11:14.894
Is, like, attributed to the brand, is that correct?
00:11:15.494 --> 00:11:18.274
It's attributed to intangible assets. We don't know which ones.
00:11:18.734 --> 00:11:24.134
And there are people, you know, but you just have to get into the guts of it,
00:11:24.194 --> 00:11:25.474
you know what I mean, to really know what you're doing.
00:11:25.534 --> 00:11:29.274
Yeah, but Coca-Cola, everybody understands there's tremendous value in that brand.
00:11:29.854 --> 00:11:35.674
No question about it. A huge part of the Tesla loss evaluation in the last four
00:11:35.674 --> 00:11:39.074
months has been a perceived erosion of the value of that brand.
00:11:40.174 --> 00:11:46.134
But my point is, if we just try to do simplistic things, like if we only had
00:11:46.134 --> 00:11:49.714
people on the balance sheet or try to apply accounting per se as it stands or
00:11:49.714 --> 00:11:54.894
try to apply gap to people, we're never going to get any place any better in analytics at all.
00:11:54.894 --> 00:11:58.454
So what are some like simple things that organizations can do to,
00:11:58.454 --> 00:12:03.294
you know, better judge the quality of their talent and the future value of their talent?
00:12:03.514 --> 00:12:05.194
Yeah, understand their bond ratings.
00:12:05.574 --> 00:12:10.194
Yeah. So my first suggestion is, I'll go a lot of different ways,
00:12:10.394 --> 00:12:14.754
but one of the, just to give you a real tangible example, I would really focus
00:12:14.754 --> 00:12:18.074
on analytics on holding line leaders accountable for talent decisions.
00:12:18.474 --> 00:12:23.114
Okay. And I wouldn't try to measure much else around that. So when I talk to
00:12:23.114 --> 00:12:27.654
CHROs, I say, would you like to have measures as powerful as they have in finance?
00:12:28.354 --> 00:12:32.394
And yes. Well, how many of the measures in finance tell you what the finance
00:12:32.394 --> 00:12:34.834
department is doing? And the answer is almost none.
00:12:35.134 --> 00:12:38.614
What they do is they hold the line leaders accountable for the asset management.
00:12:39.713 --> 00:12:47.573
Finance cares about. And, you know, even things as simple as time to hire, days to fill, right?
00:12:48.633 --> 00:12:53.893
When I changed that from days to fill by job type to days to fill by line manager,
00:12:54.473 --> 00:12:57.673
it changed the whole dynamic of the discussion with my executive team.
00:12:57.793 --> 00:13:01.033
Well, kind of along these lines, I know I'm prepping for this,
00:13:01.173 --> 00:13:03.333
Pete, and I'd love to get your perspective too, John.
00:13:03.573 --> 00:13:09.073
We talked about like, what are the human capital topics that boards of directors care about?
00:13:09.293 --> 00:13:13.813
So, you know, pushing the analytics to the line leader level is kind of the,
00:13:14.333 --> 00:13:17.173
I guess, the nirvana that maybe we should be pursuing.
00:13:17.453 --> 00:13:22.053
But if you're aggregating that up at the company level, what are the types of
00:13:22.053 --> 00:13:24.933
things that boards care about or should be caring about?
00:13:25.133 --> 00:13:29.293
And what do you guys, from your experience, think about that?
00:13:29.493 --> 00:13:29.933
Go ahead, John.
00:13:30.353 --> 00:13:30.953
You want to be just on?
00:13:32.133 --> 00:13:32.713
That's it?
00:13:32.713 --> 00:13:37.633
Well, I think I'll just give a quick, I guess my quick example is that I think
00:13:37.633 --> 00:13:43.113
boards, particularly since COVID, were directed a much greater attention to
00:13:43.113 --> 00:13:46.573
people and work and we might even say an NHR.
00:13:46.793 --> 00:13:51.573
And I think you saw an elevation of attention to the HR leader during COVID
00:13:51.573 --> 00:13:56.853
that continued with the George Floyd, kind of the George Floyd incident and,
00:13:56.873 --> 00:14:01.313
and, and, and all the kind of flurry of attention that came beyond that.
00:14:01.313 --> 00:14:05.993
And so you really saw with boards an elevation of attention to this thing called HR.
00:14:06.133 --> 00:14:09.833
Not that they weren't there before, but I think it really brought HR leaders,
00:14:09.993 --> 00:14:13.793
CHROs, et cetera, into the attention of boards.
00:14:13.913 --> 00:14:20.693
And that's kind of produced these days various calls for scorecards or dashboards
00:14:20.693 --> 00:14:25.753
that boards would find useful and a sense that those may be different than the
00:14:25.753 --> 00:14:30.153
scorecards or dashboards that HR might be using or even that we might use with line leaders.
00:14:30.153 --> 00:14:33.573
I think for me, I don't know that we have a very well-developed answer.
00:14:34.258 --> 00:14:38.978
Honestly, I think that the bottom line would be that for me,
00:14:39.098 --> 00:14:42.418
it would be much more about what boards understand and how they think about
00:14:42.418 --> 00:14:45.218
this rather than about any certain measures.
00:14:45.458 --> 00:14:49.718
So a quick example, I've seen a number of lists in which boards say we'd like
00:14:49.718 --> 00:14:54.858
to know more about the internal promotion rate and how much are we moving people
00:14:54.858 --> 00:14:56.418
between jobs in the organization.
00:14:56.698 --> 00:15:01.538
And again, one of my favorite thought experiments would be to say to boards,
00:15:01.798 --> 00:15:03.778
assume we had a perfect measure of that.
00:15:04.258 --> 00:15:08.378
Okay, you've asked for it and assume we had a very, very good measure of the
00:15:08.378 --> 00:15:10.518
proportion of positions filled internally.
00:15:11.338 --> 00:15:16.078
What would be the right number? So as a board member, what kind of, why did you ask for that?
00:15:16.618 --> 00:15:20.998
What's your logic about that number and how it fits with your concept of value?
00:15:20.998 --> 00:15:24.318
And I think what I would get is answers all over the map.
00:15:24.618 --> 00:15:27.478
You know, some would say, oh, well, I guess I thought it should be close to
00:15:27.478 --> 00:15:32.398
100%. And that seems good to me because we're, you know, we're developing our people internally.
00:15:32.618 --> 00:15:35.038
And somebody else will say, well, no, no, I don't think 100%.
00:15:35.038 --> 00:15:38.358
That's awfully expensive when there's really good talent outside.
00:15:38.878 --> 00:15:43.398
And so, gosh, I think maybe closer to 50%. And I don't think either of those
00:15:43.398 --> 00:15:46.758
is really going to be a right answer. but it would tee up the idea that,
00:15:46.858 --> 00:15:50.658
well, to really answer that question, we're going to have to dig in to how internal
00:15:50.658 --> 00:15:53.158
movement actually affects the value proposition.
00:15:53.598 --> 00:15:57.198
And we're going to have to now agree that there's going to be some places where
00:15:57.198 --> 00:16:01.458
it ought to be maybe close to zero because the best talent in the world is actually
00:16:01.458 --> 00:16:04.418
outside and some places where it should be really high.
00:16:04.738 --> 00:16:09.478
But just asking for that number, and I think HR is often quick to provide the number
00:16:10.141 --> 00:16:14.501
oh goodness, the board asked for it. This is terrific. They finally recognize
00:16:14.501 --> 00:16:21.001
HR measurement as key and to just rush to providing that almost as a service
00:16:21.001 --> 00:16:25.881
provider when the opportunity exists to step back and do something like I described,
00:16:26.061 --> 00:16:28.841
to step back and say, why is it you asked for that?
00:16:29.021 --> 00:16:33.301
What is your mental model about how value is created in a workforce if that's
00:16:33.301 --> 00:16:34.601
one of the things you asked for?
00:16:34.741 --> 00:16:37.141
Same kind of thing with things like culture instead of going.
00:16:37.141 --> 00:16:43.001
So I want to build on that. So I think resisting the reaction to the board is really important.
00:16:43.421 --> 00:16:45.921
So John and I would say you have to teach the board in advance.
00:16:46.101 --> 00:16:50.961
The board already knows discounted cash flow. The board already knows the principles of capital allocation.
00:16:51.301 --> 00:16:55.821
So they come in already knowing the mental models so they know how to process the data.
00:16:56.221 --> 00:17:00.841
And one of the challenges that CHROs have to do is they have to teach the mental models.
00:17:01.621 --> 00:17:05.221
Otherwise, the board's just going to ask for things and apply data in random ways.
00:17:05.461 --> 00:17:10.441
I'll give you an example. And it's one that I know you're kind of kidding.
00:17:10.541 --> 00:17:13.801
You don't want to talk about it much anymore. But the whole work from home discussion a little bit.
00:17:15.481 --> 00:17:17.221
Hey, we'll talk about it. I mean, come on.
00:17:17.681 --> 00:17:24.161
Yeah. If you take it down to its economics, changing your policy on work from
00:17:24.161 --> 00:17:27.521
home changes your supply of talent. Oh, yeah.
00:17:27.981 --> 00:17:29.121
Now you're talking my language.
00:17:29.581 --> 00:17:32.401
Okay. It's supply and demand curve.
00:17:32.801 --> 00:17:38.321
We also know if you don't have any selection validity, it doesn't matter if
00:17:38.321 --> 00:17:40.721
you're choosing from a pool of 20 or a pool of five.
00:17:41.641 --> 00:17:46.461
Because zero selection validity means that the size of the pool doesn't matter.
00:17:46.661 --> 00:17:48.021
Just random guessing there. Right.
00:17:48.261 --> 00:17:51.141
You know, I kind of, the example I use when I teach it to line managers is,
00:17:51.181 --> 00:17:53.561
you know, you can give me a hundred cantaloupes to choose from.
00:17:53.701 --> 00:17:56.041
I don't know what I'm doing. So you might as well give me five.
00:17:57.001 --> 00:18:00.361
Give Martha Stewart a hundred cantaloupes, you're going to end up with a dandy cantaloupe, you know?
00:18:00.581 --> 00:18:05.681
And so all of a sudden, you know, you're, the, the impact of demand only matters
00:18:05.681 --> 00:18:07.401
if you have selection validity. Yep.
00:18:07.881 --> 00:18:10.701
And can I nerd out on this for just a second?
00:18:10.961 --> 00:18:17.061
Because what you're really describing is the concept from utility analysis of SDY.
00:18:17.681 --> 00:18:24.901
So being able to quantify the measure of the value of a standard deviation in
00:18:24.901 --> 00:18:28.541
performance once you know the validity coefficients of what you're doing.
00:18:28.541 --> 00:18:34.101
And a great callback to Wayne Cascio and to John Boudreaux's work and Alexis
00:18:34.101 --> 00:18:37.541
Fink, prior guest on the podcast as well, in investing with people.
00:18:38.121 --> 00:18:42.821
Love this concept. And so maybe you could build on that for us and why that's so important, Pete.
00:18:43.221 --> 00:18:48.721
Yeah. Well, then the next step is you don't go through the process of trying to estimate SDY.
00:18:48.921 --> 00:18:49.481
Mm-hmm.
00:18:49.981 --> 00:18:54.761
Okay. But you say, what's the break-even SDY? Okay. So this is what they learned
00:18:54.761 --> 00:18:58.161
from the whole options analysis with the
00:18:59.104 --> 00:19:01.504
Oh, I can't think of the name of the options evaluation biologist.
00:19:01.664 --> 00:19:05.084
What they do is every day, every option, you can see an implied...
00:19:05.084 --> 00:19:07.184
The Black Shoal. Yeah, Black Shoal.
00:19:07.324 --> 00:19:10.444
Thank you. Yeah, Black Shoal. It's actually, to date myself,
00:19:10.604 --> 00:19:14.044
we had to study the derivative of that model as part of our...
00:19:14.044 --> 00:19:15.664
When I was in graduate school.
00:19:16.184 --> 00:19:19.644
But it's... What they do is they say, all right, if options are trading here,
00:19:20.064 --> 00:19:22.764
this implies a volatility here. Okay?
00:19:23.064 --> 00:19:24.304
So you could take...
00:19:24.304 --> 00:19:27.564
Let's assume we have this much demand reduction, which you can measure.
00:19:28.104 --> 00:19:30.924
So we have this much validity, which you can estimate.
00:19:31.484 --> 00:19:36.504
Therefore, your break-even SDY on work from home is X.
00:19:37.204 --> 00:19:43.184
And if it's 200,000 or if it's 5,000, well, those result in different kinds of decisions.
00:19:43.384 --> 00:19:46.824
And John and I further recommended to refine those kind of models.
00:19:46.944 --> 00:19:49.644
Why not run experiments in your company? You talked on some of your podcasts
00:19:49.644 --> 00:19:52.964
about A, B and stuff like that. Maybe different divisions have different kinds of dynamics.
00:19:53.224 --> 00:19:55.764
But the idea is to put a decision model in front of it.
00:19:56.164 --> 00:19:58.504
And by the way, that's where bond ratings really became valuable.
00:19:58.664 --> 00:20:01.764
They were interesting. All of a sudden, then we had capital asset pricing models
00:20:01.764 --> 00:20:03.104
that made them really powerful.
00:20:03.424 --> 00:20:06.584
You know, it's the decision models that make the information powerful.
00:20:07.304 --> 00:20:10.904
And we try to get so much data out there, we don't spend enough time working
00:20:10.904 --> 00:20:14.744
the decision models to a level where a board or a line manager,
00:20:15.044 --> 00:20:16.044
et cetera, can understand.
00:20:17.004 --> 00:20:19.564
Surely you can shift those sort of metrics, too, by just, like,
00:20:19.644 --> 00:20:22.924
throwing a bunch of money at people as well. Like you may not want to come in
00:20:22.924 --> 00:20:26.064
the office five days a week, but if you throw enough money at people,
00:20:27.244 --> 00:20:30.504
you know, you change your selection pool, but that's going to cost you. Of course.
00:20:30.684 --> 00:20:32.564
That's another way to affect supply and demand, right?
00:20:32.884 --> 00:20:33.124
Yeah.
00:20:33.244 --> 00:20:36.204
That's, that's, that's one of the ways to affect supply and demand.
00:20:36.364 --> 00:20:39.304
And you may find you have a very, very nonlinear curve. I think we're seeing
00:20:39.304 --> 00:20:43.044
lots of signs of nonlinear curve that there's a lot of benefit giving people
00:20:43.044 --> 00:20:45.764
two days work from home. You know what I mean? Yeah.
00:20:46.704 --> 00:20:49.784
Those kinds of things. They do that in marketing all the time.
00:20:49.864 --> 00:20:53.304
They estimate utility curves. They estimate, and they run experiments.
00:20:53.484 --> 00:20:57.464
So they build the decision models and then populate them with data based on
00:20:57.464 --> 00:20:58.544
their own internal companies.
00:20:58.844 --> 00:21:02.684
Why isn't this used more often? Why don't companies use this?
00:21:02.764 --> 00:21:07.484
If all the leaders are well-adapted with MBAs and all this sort of stuff,
00:21:07.644 --> 00:21:10.304
they ostensibly think in this sort of manner.
00:21:10.644 --> 00:21:12.904
It's not regularly applied in HR.
00:21:13.164 --> 00:21:13.824
Why is that?
00:21:14.850 --> 00:21:15.950
Not taught for one.
00:21:16.630 --> 00:21:20.570
Yeah. Let me talk about that. Because when I originally reached out to John
00:21:20.570 --> 00:21:24.290
about this and we were talking about some of these topics and in his infinite
00:21:24.290 --> 00:21:27.890
humility, he was like, Cole, nobody cares about this stuff anymore.
00:21:28.090 --> 00:21:31.770
And we figured this all out, you know, in the 90s or the early 2000s.
00:21:31.890 --> 00:21:33.430
Why do you think this is even relevant?
00:21:33.590 --> 00:21:37.070
And the thing I said is because I feel like there's just been this great forgetting
00:21:37.070 --> 00:21:41.810
that if something hasn't happened in the last three, five, 10 years in HR,
00:21:41.810 --> 00:21:45.730
it's like we never knew it at all. And so there was all this mass amount of
00:21:45.730 --> 00:21:46.750
knowledge that was gained.
00:21:46.930 --> 00:21:52.310
I don't know anybody who's doing utility analysis anymore or that would even know to do it.
00:21:52.430 --> 00:21:55.410
It's not just that they don't do it. It's like that they would even know to do it.
00:21:55.510 --> 00:22:00.050
So I would love you guys to talk about, you know, just the power of variation
00:22:00.050 --> 00:22:01.710
and being able to quantify that,
00:22:02.070 --> 00:22:06.750
like how some of these ideas are very applicable, especially at this moment
00:22:06.750 --> 00:22:11.250
in time and why it's so relevant to get over this great forgetting that we've had.
00:22:11.250 --> 00:22:16.630
So most people learn utility analysis as a way to cost justify the selection
00:22:16.630 --> 00:22:17.830
system they want to install.
00:22:18.390 --> 00:22:22.190
So if you're doing your benefit analysis to drive your agenda,
00:22:22.670 --> 00:22:26.230
that you want to justify your training programs, you want to justify your selection
00:22:26.230 --> 00:22:29.550
process, you want to justify HR investments.
00:22:29.810 --> 00:22:32.810
That's 90% of the people that even know utility analysis.
00:22:33.270 --> 00:22:37.250
That's the default that they come in in terms of what they're trying to do with it.
00:22:37.890 --> 00:22:41.530
And even then, they don't do a good job of explaining it to people who don't understand it.
00:22:41.630 --> 00:22:45.710
So you're coming in already with a bias that the CFO doesn't understand the
00:22:45.710 --> 00:22:49.510
model, CFO doesn't want to spend on HR, and you're going to show them some formula
00:22:49.510 --> 00:22:52.570
they've never seen or understood before and expect them to make a decision with it.
00:22:53.190 --> 00:22:57.390
You take that same exact model and apply it to a really important decision around
00:22:57.390 --> 00:23:01.370
how it can affect supply and demand to talent and how that impacts us in one, three, and five years.
00:23:01.910 --> 00:23:05.210
And rather than say, here's the answer, you say, here's the break-even points.
00:23:05.710 --> 00:23:10.630
And then the CFO's job and the line manager say, do we think we're above or below that number?
00:23:11.190 --> 00:23:16.210
You show them where the decision changes, not what you expect to have it happen.
00:23:18.230 --> 00:23:21.910
So that's the kind of stuff. And we just haven't had a lot of people analytics
00:23:21.910 --> 00:23:26.350
and CHROs that are prepared to do that and know how to apply it and say,
00:23:26.430 --> 00:23:29.250
oh, this is really – nobody says work from home.
00:23:29.370 --> 00:23:33.230
There's a supply-demand problem that utility analysis really helps us answer well.
00:23:34.010 --> 00:23:35.010
Well, now they do.
00:23:35.530 --> 00:23:36.390
Yeah. Well, I just,
00:23:36.530 --> 00:23:39.430
Would you see Cole say how that changes the process?
00:23:39.710 --> 00:23:42.990
I totally understand it. John, I want to kick it to you here for a second.
00:23:43.150 --> 00:23:45.210
First of all, I think you've done really great work in this space,
00:23:45.210 --> 00:23:50.670
but also you have published some things more recently thinking of like work without jobs, right?
00:23:50.870 --> 00:23:53.830
And kind of what's going on there and the decomposition of work.
00:23:53.910 --> 00:23:56.690
And then I believe, and you can correct me if I'm wrong on this,
00:23:56.810 --> 00:24:02.110
I believe that came out slightly before Genre of AI kind of entered the picture.
00:24:02.110 --> 00:24:04.250
Has that accelerated the trend?
00:24:04.490 --> 00:24:08.530
And can you speak to that at all in its relationship to how we even think about
00:24:08.530 --> 00:24:12.230
what constitutes work and who does work and all of that?
00:24:13.788 --> 00:24:17.168
Yeah, maybe we're having fun with analogies today,
00:24:17.228 --> 00:24:22.168
So I'll kind of put an analogy out there first, and then we can talk about the implications.
00:24:22.588 --> 00:24:30.308
So I think, again, if I were talking to non-HR executives about work without jobs and about this,
00:24:30.488 --> 00:24:35.868
the basic idea is to consider whether looking at work with a unit of analysis
00:24:35.868 --> 00:24:41.728
that would be what's inside jobs, so it would be tasks or projects or something like that.
00:24:41.728 --> 00:24:45.208
I always like to think about a job description. Let's say
00:24:45.208 --> 00:24:45.948
We print it out
00:24:45.948 --> 00:24:49.908
And then we slice it up. And all the things that are inside it,
00:24:49.968 --> 00:24:54.108
all the tasks that are inside it become now like strips of paper.
00:24:54.248 --> 00:24:57.548
And let's say we did that for a team in a distribution center.
00:24:57.908 --> 00:25:01.248
So we take all these things we call jobs, we'd lay them all out,
00:25:01.348 --> 00:25:06.268
we'd snip them up, and then we'd say, you get to arrange these tasks in the
00:25:06.268 --> 00:25:07.608
most logical way possible.
00:25:08.088 --> 00:25:11.528
I've done this. I've talked to a lot of school teachers about this who came
00:25:11.528 --> 00:25:15.708
through COVID and said it really shined a light on how much of our time is spent
00:25:15.708 --> 00:25:22.288
on administration and meetings and stuff like that and how all of the data says
00:25:22.288 --> 00:25:27.348
that you need a couple of hours of prep for hour of being in front of a classroom, at least.
00:25:27.908 --> 00:25:32.088
And so if you really were to arrange this, you'd make prep much bigger.
00:25:32.388 --> 00:25:36.088
You'd make in front of the classroom much bigger. You'd drop the meetings down.
00:25:36.088 --> 00:25:39.828
You know, and so, and, but, and that, and you get to that a bit more easily
00:25:39.828 --> 00:25:44.028
if you're allowed to say, well, you don't have to say how many teachers do we
00:25:44.028 --> 00:25:45.408
need with this job description?
00:25:45.788 --> 00:25:49.028
You can say how much more time in prep do we need?
00:25:49.148 --> 00:25:53.148
How much more time in the classroom do we need? And then go reconstruct the
00:25:53.148 --> 00:25:56.828
work to reflect what, what you've discovered that.
00:25:57.048 --> 00:26:01.828
And it really does have, so, so that's kind of, and the analogy would be,
00:26:01.988 --> 00:26:06.828
I, you know, right when we were writing this book, I was working with a pet food company.
00:26:07.948 --> 00:26:09.548
And we were trying to figure out how to
00:26:10.160 --> 00:26:16.000
express this analogy to their leaders. And as Pete said, what I find is we need
00:26:16.000 --> 00:26:17.880
metaphors to engage their smart mind.
00:26:18.220 --> 00:26:21.220
So I would say, well, here's how I would present it to them.
00:26:21.460 --> 00:26:25.780
Remember the days when you created pet food and the choices you had was how
00:26:25.780 --> 00:26:29.860
much bone to put in, how much meat to put in, and how much brain to put in.
00:26:30.380 --> 00:26:34.180
And with those units of analysis, you could mix up pet food.
00:26:34.500 --> 00:26:38.680
It was pretty blunt edged. And then someone looked inside each of those components
00:26:38.680 --> 00:26:40.440
and found the nutritional elements,
00:26:41.040 --> 00:26:42.020
You know, the vitamins,
00:26:42.320 --> 00:26:47.400
Et cetera, the amount of fiber. Well, once you understand that across those
00:26:47.400 --> 00:26:51.720
ingredients and you start to realize you could create food based on nutritional
00:26:51.720 --> 00:26:54.900
level outcomes instead of based on,
00:26:55.180 --> 00:27:00.520
well, how many ounces of meat, how many ounces of grain, how many ounces of bone, whatever.
00:27:01.000 --> 00:27:05.960
And then you have an entirely different way, an entirely more effective way
00:27:05.960 --> 00:27:10.020
of understanding what you're creating and the value you're creating in the pet food.
00:27:10.140 --> 00:27:15.860
So the fundamental thing, Cole and Scott, I guess, in the deconstruction of
00:27:15.860 --> 00:27:18.800
work is this idea that, again, it goes to mental models.
00:27:18.960 --> 00:27:26.000
If we allowed leaders, employees, and others to think about work in a way that
00:27:26.000 --> 00:27:29.960
what I like to think of is if the job description is an ice cube, we allow it to melt.
00:27:30.766 --> 00:27:34.326
And we allow those drops of water to refreeze in different ways.
00:27:34.506 --> 00:27:39.986
And even if you did the hypothetical idea of snipping up those job descriptions,
00:27:40.226 --> 00:27:44.586
there's often situations where that's going to offer leaders and employees freedom
00:27:44.586 --> 00:27:48.706
to talk about and to consider things that they wouldn't otherwise.
00:27:48.986 --> 00:27:53.146
Just like you see it in nursing. If you take the job description of nurses,
00:27:53.366 --> 00:27:58.166
it included looking in on well patients and including writing schedules for
00:27:58.166 --> 00:28:00.766
the nursing shifts, that sort of thing.
00:28:01.046 --> 00:28:04.246
All of those things, you know, automate the schedule writing,
00:28:04.546 --> 00:28:08.926
give the looking in on well patients to candy stripers or somebody like that.
00:28:09.086 --> 00:28:12.926
And now what you're left with is a whole bunch of stuff that really needs nurses
00:28:12.926 --> 00:28:15.906
to do and that we could emphasize with them.
00:28:16.026 --> 00:28:19.206
And in fact, once you see what nurses do, if you still have a shortage,
00:28:19.446 --> 00:28:23.846
you realize that your hospital executives were often physicians before.
00:28:24.206 --> 00:28:29.226
And so you get them recertified as physicians. And now that work,
00:28:29.466 --> 00:28:33.046
that not so much the job, you know, there's no job description of a hospital
00:28:33.046 --> 00:28:38.946
administrator who spends a day on the floor as well, right? That wasn't in the job description.
00:28:39.106 --> 00:28:43.266
But once they're recertified, if you deconstruct the administrator,
00:28:43.506 --> 00:28:50.166
you realize that the person inside that job has these capabilities that go with
00:28:50.166 --> 00:28:53.366
being a physician, even though they're not reflected in the job description.
00:28:53.706 --> 00:28:58.646
So again, the idea is that if we're looking at the people and the jobs in some
00:28:58.646 --> 00:29:02.866
more molecular level, I like the concept of capabilities better than skills.
00:29:03.206 --> 00:29:05.966
I mean, it's pretty simple questions for me.
00:29:06.446 --> 00:29:12.446
The capability side on the part of the person would be, what can you do or what
00:29:12.446 --> 00:29:13.966
could we prepare you to do?
00:29:14.206 --> 00:29:17.006
So very specific, here are things that you can do.
00:29:17.186 --> 00:29:21.986
Not so much do you know the Spanish language, but do you know how to simultaneously
00:29:21.986 --> 00:29:24.886
translate? Do you know how to translate a document, right?
00:29:25.066 --> 00:29:30.406
And the same thing on the work side is what do we need done is a fairly straightforward
00:29:30.406 --> 00:29:33.786
question, and that would get us inside those job descriptions.
00:29:34.086 --> 00:29:38.126
And then I'll hearken back and say, I think a lot of the psychological tools
00:29:38.126 --> 00:29:42.226
that have been applied to jobs, and I think, frankly, a lot of the confusion
00:29:42.226 --> 00:29:46.786
is that we weren't working at the molecular level of things that need to get
00:29:46.786 --> 00:29:48.026
done and things that get done.
00:29:48.206 --> 00:29:51.746
So if you think about the example Pete and I used of sweepers at Disney,
00:29:52.006 --> 00:29:57.266
right? The sweeper job description has a lot of stuff in it about cleaning and
00:29:57.266 --> 00:29:59.186
some stuff in it about entertaining guests.
00:29:59.626 --> 00:30:05.786
You can't really see the strategic value until you break those apart and draw the payoff curve.
00:30:06.632 --> 00:30:11.492
How much better off would Disney be if we made cleaning the park 30% better?
00:30:11.932 --> 00:30:15.252
Well, not that much, probably. We're already good enough, probably,
00:30:15.412 --> 00:30:18.352
on keeping the park clean. It's not unimportant, but we're pretty good.
00:30:18.592 --> 00:30:23.692
How much better would Disney be if we enhanced the ability of sweepers to answer
00:30:23.692 --> 00:30:28.572
questions, to entertain guests in a surprising way? And you say,
00:30:28.652 --> 00:30:30.032
oh, yeah, that part of the job.
00:30:30.532 --> 00:30:36.432
Well, that's really valuable. And so I don't think there's enough to get on a soapbox.
00:30:36.572 --> 00:30:43.392
I just don't think this discussion is undertaken often enough by leaders and
00:30:43.392 --> 00:30:46.452
workers about what's really adding value here.
00:30:46.692 --> 00:30:51.912
And where is what you're doing? Where is good enough good enough?
00:30:51.912 --> 00:30:54.872
You're not going to enhance value by getting better at that.
00:30:54.872 --> 00:30:59.112
Pete's favorite example is getting tax documents in.
00:30:59.612 --> 00:31:03.292
It doesn't help to get them in three months early. You know,
00:31:03.392 --> 00:31:06.372
that is not adding value, you know, even if you could do it,
00:31:06.472 --> 00:31:07.632
Right? You could pay your penalty early there.
00:31:08.732 --> 00:31:09.852
Yeah, exactly.
00:31:10.572 --> 00:31:15.152
It's amazing, like, how much mileage you can get out of just talking about the
00:31:15.152 --> 00:31:19.132
mission, vision, strategy of an organization. Like that will funnel down to
00:31:19.132 --> 00:31:22.612
everything else that you want to happen, whether it's someone sweeping the park
00:31:22.612 --> 00:31:25.472
or actually, you know, entertaining the guests.
00:31:25.632 --> 00:31:28.992
In that case, like probably Disney wants you to provide a great experience for
00:31:28.992 --> 00:31:30.832
their customers in an entertaining way.
00:31:31.052 --> 00:31:33.512
Oh, Pete, you faded out there. That's incredible.
00:31:33.752 --> 00:31:35.412
Yeah, I adjusted my camera. Sorry about that.
00:31:37.032 --> 00:31:40.372
There's always been a fair amount of like job crafting that goes around that
00:31:40.372 --> 00:31:41.952
people do on their own. Right.
00:31:42.092 --> 00:31:46.772
But what you're talking about is a much more systematic way to get at what the
00:31:46.772 --> 00:31:50.572
organization actually needs out of employees and redistribute the actual talent.
00:31:51.052 --> 00:31:56.292
Yeah. And I think get a good understanding of it. I think that's the key, Scott, is I'm not.
00:31:56.882 --> 00:32:00.302
I'm much more interested in the way people think about this stuff than I am
00:32:00.302 --> 00:32:04.682
in any particular measurement or outcome of it, you know, and I've always seen
00:32:04.682 --> 00:32:08.362
measurement as a, measurement is a vehicle to getting people to think better,
00:32:08.642 --> 00:32:12.082
to, to, to make decisions better, to understand things better.
00:32:12.262 --> 00:32:13.842
That's the ultimate goal.
00:32:14.062 --> 00:32:19.262
And the, in my, my sense, the ultimate evaluation is, are people making better
00:32:19.262 --> 00:32:22.222
decisions than they would if you hadn't measured this thing?
00:32:22.322 --> 00:32:26.982
And in this case, I'd say, would people make better decisions if there was kind
00:32:26.982 --> 00:32:33.162
of an institutionalized capability and requirement that we look at your work
00:32:33.162 --> 00:32:36.122
and we look at the work and ask these kinds of questions?
00:32:36.122 --> 00:32:40.682
In addition to whether it could be automated, we actually get a handle on what
00:32:40.682 --> 00:32:44.582
is the real payoff and what do people think?
00:32:44.582 --> 00:32:48.982
I think there's probably a lot of the equivalent of Disney sweepers running
00:32:48.982 --> 00:32:52.962
around, maybe not at Disney, but you can imagine the sweepers think that where
00:32:52.962 --> 00:32:54.642
they add value is sweeping better.
00:32:54.962 --> 00:32:57.862
It's in the job description. It might even be in the title. Well,
00:32:57.962 --> 00:33:01.202
that to me is the key is whether we measure stuff or not,
00:33:01.442 --> 00:33:06.222
would it be helpful to have a process by which the sweepers and their leaders
00:33:06.222 --> 00:33:10.402
could talk about where they add value at that deconstructed level?
00:33:10.562 --> 00:33:13.902
Is that a cultural thing? I think you have to be careful about mission,
00:33:14.042 --> 00:33:16.602
vision, and values just a little bit, because what happens is,
00:33:16.782 --> 00:33:19.862
as companies have gone to scale and with the technology,
00:33:20.142 --> 00:33:24.702
we're seeing a lot of people try to build enterprise-wide performance management
00:33:24.702 --> 00:33:27.442
around things at the enterprise scale level.
00:33:27.482 --> 00:33:31.722
When you think about Disney, they've got cruise lines. They've got movie studios.
00:33:31.982 --> 00:33:35.482
They've got television studios. They've got theme parks, obviously.
00:33:36.542 --> 00:33:42.442
And one of the things that John and I are seeing is that we are eliminating
00:33:42.442 --> 00:33:47.602
the strategic context by aligning it to the very highest level of the enterprise.
00:33:48.622 --> 00:33:52.802
And it's very, very common. I think about my experience at Toro,
00:33:52.842 --> 00:33:56.462
where he had, at that time, nine significant business units.
00:33:56.642 --> 00:33:59.442
Some were industry leaders. Some were in growing industries.
00:33:59.662 --> 00:34:03.902
Some were third or fourth place in their industries. Some were in margin challenge
00:34:03.902 --> 00:34:08.742
industries. And it was, you know, the idea of having just one common set of
00:34:08.742 --> 00:34:10.502
mission, vision, and values and trying to align it.
00:34:10.722 --> 00:34:14.122
You can't be strategic if you're trying to do a one-size-fits-all when you've
00:34:14.122 --> 00:34:16.742
got eight different business units in eight different contexts.
00:34:17.082 --> 00:34:22.082
I see a lot of COEs trying to get to scale, in effect, robbing the opportunity
00:34:22.082 --> 00:34:25.842
to be strategic at the level that really matters to line and business leaders.
00:34:26.522 --> 00:34:29.522
This is also where, like, Japanese companies will have, like,
00:34:29.522 --> 00:34:33.602
a one-year plan, a three-year plan, five-year plan, a 500-year plan for how
00:34:33.602 --> 00:34:34.262
they're going to operate.
00:34:34.422 --> 00:34:39.262
Obviously, trying to project 500 years out isn't probably that accurate.
00:34:39.662 --> 00:34:43.402
But it gives you some sort of time horizons to think about. What sort of skills
00:34:43.402 --> 00:34:46.742
you may need in the future, what you want to do in the future, and...
00:34:47.358 --> 00:34:50.218
Be able to adjust as different technologies come down the pipeline
00:34:50.218 --> 00:34:55.318
Yeah what i hear you guys saying is that airline pilots are the most critical
00:34:55.318 --> 00:34:59.518
role in any company that has airline pilots that's that's my takeaway.
00:34:59.518 --> 00:35:05.678
I found in working as a chro that the concept of pivotal pools is hard to communicate
00:35:05.678 --> 00:35:09.798
what works though it's the same concept and it's defined the strategy pivot
00:35:09.798 --> 00:35:14.938
points and then talk about how close a job is to it and so the idea there's
00:35:14.938 --> 00:35:16.938
such a fairness issue around
00:35:17.358 --> 00:35:19.338
I'm pivotal and I'm not, that gets in the way.
00:35:19.598 --> 00:35:23.898
But if you can identify the pivot points and then identify job proximity to
00:35:23.898 --> 00:35:26.038
it, you have a much better chance of being successful.
00:35:26.198 --> 00:35:28.838
That's the one thing that we've learned in the 20-plus years since we wrote
00:35:28.838 --> 00:35:33.138
the book is the concepts are right, but there's just some really slight nuances
00:35:33.138 --> 00:35:34.778
in how it gets implemented to make it effective.
00:35:35.378 --> 00:35:37.338
Yeah, cool. You got served. What's up?
00:35:37.798 --> 00:35:43.678
Taking my joke and educating me with it. I appreciate it. Let's do some confusion.
00:35:45.418 --> 00:35:46.578
The Confusion Matrix.
00:35:47.698 --> 00:35:50.998
All right, we got five questions for everybody here.
00:35:51.098 --> 00:35:54.978
And if Cole likes your answer, he will provide you a social media tile that
00:35:54.978 --> 00:35:56.538
you can share with your friends and family.
00:35:56.978 --> 00:36:00.238
All right, anyone jump in here. This is kind of wild here.
00:36:00.618 --> 00:36:05.558
So what's a belief that you held early in your career that you're completely
00:36:05.558 --> 00:36:07.618
changed your mind about now?
00:36:07.618 --> 00:36:10.578
Oh, that's easy. And related to what we did,
00:36:10.778 --> 00:36:16.998
I think when all of us started doing utility analysis, what we believed was
00:36:16.998 --> 00:36:22.698
that the only reason executives and others weren't making better decisions is
00:36:22.698 --> 00:36:25.638
that we had not translated psychology into dollars.
00:36:25.978 --> 00:36:29.758
That if we just translated this stuff into dollars in a credible way,
00:36:29.938 --> 00:36:33.718
that that would change the way leaders made decisions about people.
00:36:33.718 --> 00:36:36.878
I think that was, you know, kind of at the heart of all of that.
00:36:37.038 --> 00:36:42.158
And I think one of the reasons it's fallen out of favor is that, in fact,
00:36:42.358 --> 00:36:47.998
it's much, much more about what Pete is talking about, about priming leaders
00:36:47.998 --> 00:36:53.418
on a way to think before we come in with a fully blown analysis.
00:36:53.418 --> 00:36:55.838
And I think that's still true today.
00:36:56.098 --> 00:37:01.758
So for me, I think the belief that changed is today what I believe is that the
00:37:01.758 --> 00:37:08.798
real bottleneck is that leaders come into discussions about people using simplistic myopic models.
00:37:09.418 --> 00:37:13.858
It's not so much their fault, but they've never really been held accountable the way they have.
00:37:13.998 --> 00:37:18.098
And that's what needs to be changed, whether we think about measurement,
00:37:18.258 --> 00:37:20.298
whether we think about how we define work, et cetera.
00:37:20.578 --> 00:37:26.538
It isn't just that they're desperately waiting to make better decisions about
00:37:26.538 --> 00:37:31.478
people, and all they need is the dollar values or a good scorecard or something like that.
00:37:31.638 --> 00:37:33.838
I'd say just that I knew anything.
00:37:36.318 --> 00:37:40.258
Actually new stuff and apparently sure yeah.
00:37:41.678 --> 00:37:42.698
I think the other one is,
00:37:42.818 --> 00:37:48.718
For me, I had a hypothesis that the HR profession was going to change much faster.
00:37:49.398 --> 00:37:52.718
If I go back and look at 40 years and look at the research that we did,
00:37:52.818 --> 00:37:56.218
a lot of the research I did with Ed Lawler, where we did a survey of HR leaders
00:37:56.218 --> 00:37:59.978
every three years or so, really hammered home the point to me that,
00:38:00.398 --> 00:38:03.898
in fact, a profession is often defined very traditionally.
00:38:03.898 --> 00:38:10.818
The bulk of HR work in organizations is very, very traditional and very hard
00:38:10.818 --> 00:38:12.178
to move out of that tradition.
00:38:12.738 --> 00:38:16.118
And that's, I think of it as an iceberg of HR, right?
00:38:16.138 --> 00:38:19.938
If you think about all the HR people from the Tobys of the world in the office
00:38:19.938 --> 00:38:24.458
or, you know, whatever is your metaphor for that, and then up to the headline
00:38:24.458 --> 00:38:27.458
leaders that you feature on your program, right?
00:38:27.778 --> 00:38:30.558
The headline leaders are kind of the iceberg above the water.
00:38:30.898 --> 00:38:34.978
You know, futuristic stuff, influence, all that kind of thing.
00:38:35.138 --> 00:38:39.678
And the iceberg below the water is much, much bigger and tends to be pretty
00:38:39.678 --> 00:38:43.618
traditional and it tends to be a difficult thing to move.
00:38:44.158 --> 00:38:47.198
That's the way people think about HR. They think about Toby from the office
00:38:47.198 --> 00:38:48.698
or someone out in the kitchen, right?
00:38:49.318 --> 00:38:54.378
Yeah. And that really affects how people are prepared to hear ideas. I've been shocked.
00:38:54.558 --> 00:38:57.998
So I'm the same I'm the CFO or I'm the CHRO.
00:38:58.318 --> 00:39:02.118
And people react to me differently with a different title next to my name.
00:39:02.558 --> 00:39:09.058
And it isn't because I can't speak. I'm just as financially competent as a CHRO as I was as a CFO.
00:39:09.438 --> 00:39:14.618
But the idea, there's some real barriers to that reputation and that history.
00:39:16.278 --> 00:39:19.398
So one of the things John and I have talked a lot about is just some of the
00:39:19.398 --> 00:39:22.078
experiences I've had being on both sides,
00:39:22.078 --> 00:39:26.978
doing it with exactly the same situation and the halo of the department that
00:39:26.978 --> 00:39:31.418
I come from just changing entirely the situation that I find myself in.
00:39:31.878 --> 00:39:35.618
It's the guy with the money bags versus the guy that can punish you,
00:39:35.778 --> 00:39:36.018
Right?
00:39:38.038 --> 00:39:39.338
It's really interesting.
00:39:40.018 --> 00:39:43.038
Okay, let's say you have one month, no email, no responsibilities,
00:39:43.538 --> 00:39:46.838
no professional obligations. What are you doing? What are you doing?
00:39:47.558 --> 00:39:51.438
It's easy. I'm going to Lake O'Woods, northern Minnesota. I'm going up to Angle Inlet.
00:39:52.078 --> 00:39:57.298
I'm going up to the Voyager National Park. I'm getting as far away from northern
00:39:57.298 --> 00:40:00.718
Minnesota and away from everything as I possibly can.
00:40:02.366 --> 00:40:03.546
Doing some ice fishing?
00:40:04.486 --> 00:40:09.486
Yeah, I'm not a big ice fisher. I mean, there's too many. We lose at least a
00:40:09.486 --> 00:40:11.746
dozen or two people every year in Minnesota here.
00:40:12.646 --> 00:40:15.306
Funny thing, I have a friend of mine who lives on a big ice fishing lake,
00:40:15.346 --> 00:40:19.726
and she's a big extrovert. And she used to cross-country ski and just go talk to the ice fishermen.
00:40:19.886 --> 00:40:23.026
And I said, Paula, do you realize how much these guys don't want to talk to
00:40:23.026 --> 00:40:25.506
anybody if they're sitting for an hour on a lake?
00:40:26.386 --> 00:40:28.826
You know, just in a dark house.
00:40:28.826 --> 00:40:32.146
But she said they worked exactly.
00:40:32.606 --> 00:40:34.966
She said possible. So, no, that's what I would do.
00:40:35.166 --> 00:40:35.946
What are you doing, John?
00:40:36.586 --> 00:40:40.306
Yeah, I suppose, you know, I've kind of moved to my lake house in Santa,
00:40:40.406 --> 00:40:42.666
the equivalent of my lake house in Santa Fe, New Mexico.
00:40:43.126 --> 00:40:47.646
So, yeah, I'm probably out on one of the many, many trails out here trying to
00:40:47.646 --> 00:40:48.706
get a little closer to nature.
00:40:49.286 --> 00:40:55.286
Or I may be in a museum somewhere marveling at the nuances of an indigenous
00:40:55.286 --> 00:40:58.386
culture, you know, while we get baskets or pots, et cetera.
00:40:58.866 --> 00:41:01.746
How did you two ever get together? Both of y'all just seemed to want to isolate,
00:41:02.006 --> 00:41:06.046
self-isolate out in the woods. I mean, it was a historical trauma or something.
00:41:06.906 --> 00:41:12.506
We self-isolated a lot when we were together. I mean, if you tried to look for
00:41:12.506 --> 00:41:15.186
Pete and me of an evening after we'd been running a workshop,
00:41:15.366 --> 00:41:19.466
what you would find is the pair of us walking sometimes for hours,
00:41:19.646 --> 00:41:23.246
just talking together in whatever city it was.
00:41:23.346 --> 00:41:26.846
We'd just head off, and before we knew it, a couple of hours would have been
00:41:26.846 --> 00:41:30.146
passed, and we'd have been self-isolated. you know, in that way together.
00:41:30.993 --> 00:41:34.433
I'm going to guess Cole is going to go golfing. Is that right, Cole?
00:41:35.073 --> 00:41:38.913
I was just thinking, I would get so anxious, I would end the month with like
00:41:38.913 --> 00:41:41.333
seven more projects that I was working on. Yeah.
00:41:41.893 --> 00:41:46.433
That sounds right, honestly. Yeah. Okay, well, we'll wrap it up with this one.
00:41:46.533 --> 00:41:50.873
Like, what is one everyday skill or habit that you rely on that has nothing
00:41:50.873 --> 00:41:54.213
to do with academia or business, but makes you better at your work?
00:41:54.633 --> 00:41:58.593
As silly as it sounds, one of the best things I learned in high school is how to type.
00:41:59.153 --> 00:42:00.653
How to type. Oh, yeah.
00:42:00.653 --> 00:42:04.793
You know, back in those days, it was taught with manual typewriters.
00:42:05.033 --> 00:42:09.053
And early on in my career, I was in software development.
00:42:09.233 --> 00:42:12.373
And for that ability, I was trying to move and become more like,
00:42:12.533 --> 00:42:16.873
silly as it sounds, it's probably one of the most important things I learned early in my career.
00:42:17.033 --> 00:42:19.333
And it's served to me well ever since.
00:42:19.713 --> 00:42:26.913
Were those QWERTY typewriters like the same format that you have on a laptop now?
00:42:28.373 --> 00:42:35.193
I love that. Pete, yeah, I love the generational difference that Cole is asking
00:42:35.193 --> 00:42:36.713
that actual question, right?
00:42:36.893 --> 00:42:40.273
Like for you and me, it's like, well, of course, you know, QWERTY typewriters,
00:42:40.393 --> 00:42:42.493
that was the 1960s, 1970s.
00:42:42.553 --> 00:42:47.653
You know, we were using those to do punch carts, to feed into massive mainframe
00:42:47.653 --> 00:42:51.973
computers and get a little COBOL computer program to run it.
00:42:52.033 --> 00:42:56.413
Yeah, so it's been QWERTY. What was amazing to me is that QWERTY did hold up,
00:42:56.613 --> 00:43:01.393
that that weird keyboard design that was designed.
00:43:01.393 --> 00:43:06.493
QWERTY does hold up. You can modify it to the Dvorak system as well.
00:43:06.653 --> 00:43:08.053
I got a friend that uses that.
00:43:08.613 --> 00:43:11.413
And boy, you're talking about trying to do things left-handed.
00:43:13.033 --> 00:43:14.493
I mean, oh my God.
00:43:15.813 --> 00:43:20.013
I've always been grateful for my math background. So I was a math major undergrad,
00:43:20.333 --> 00:43:24.613
and it gave me comfort. was just, you know, everything else I did built on that afterwards.
00:43:24.853 --> 00:43:28.193
So that's what made the economics accessible. That's what made IO psychology accessible.
00:43:28.815 --> 00:43:31.315
It was my foundation of math.
00:43:31.775 --> 00:43:36.935
So one of the things as this podcast has gotten more popular when we first started,
00:43:36.935 --> 00:43:39.895
it was basically just our friend group who we had on as guests.
00:43:40.095 --> 00:43:43.695
And we have people from different generations coming on the podcast.
00:43:43.695 --> 00:43:46.135
And so I have to explain it to people.
00:43:46.235 --> 00:43:48.755
It has a very millennial feel to it.
00:43:51.155 --> 00:43:56.755
Because people aren't expecting this type of podcast relationship when doing
00:43:56.755 --> 00:43:58.815
this. So I appreciate the call out, John.
00:43:59.035 --> 00:44:02.035
Yeah, it was all meant positively. Right.
00:44:02.555 --> 00:44:02.995
Absolutely.
00:44:03.755 --> 00:44:08.695
You said punch cards. I've had to admit if I didn't relate this one sweet story
00:44:08.695 --> 00:44:10.395
about my grandparents and how they met.
00:44:10.595 --> 00:44:15.715
My grandparents, my grandfather worked for IBM. This is like in the 1920s or
00:44:15.715 --> 00:44:17.395
something like that in Beaumont, Texas.
00:44:17.695 --> 00:44:22.755
And my grandmother would come and purposely mess up the punch cards on the IBM
00:44:22.755 --> 00:44:25.875
machine just so he would have to come around the office and she could talk to
00:44:25.875 --> 00:44:27.735
us. That's cute. and they wound up getting married.
00:44:29.355 --> 00:44:30.395
The rest is history.
00:44:31.015 --> 00:44:35.155
So I've learned a new phrase from Cole. I would call that a meet cute.
00:44:35.875 --> 00:44:38.735
Oh, gosh. Wait, wait, wait, wait, wait, what?
00:44:39.555 --> 00:44:42.875
He didn't learn from me. He's clearly been listening to the podcast.
00:44:43.075 --> 00:44:46.895
That was Holly Lamb's episode. She explained what a meet cute was.
00:44:47.015 --> 00:44:49.015
I did not know what it was. I know what it is now.
00:44:49.695 --> 00:44:53.295
So I think Scott had just described a meet cute if I understand it correctly.
00:44:53.355 --> 00:44:57.055
Oh, man. I got to listen to this Holly Lamb episode.
00:44:58.255 --> 00:45:01.795
Let's do some nerdery. The nerdery.
00:45:03.115 --> 00:45:06.535
I will tee us up with the first one.
00:45:07.015 --> 00:45:11.075
Yeah, I think I'm going to start us with the how many people does your company
00:45:11.075 --> 00:45:12.575
have or how many people does it need?
00:45:12.675 --> 00:45:17.055
I think the reason why I chose this one is sort of because the conversation
00:45:17.055 --> 00:45:21.875
we were having around work without jobs and what's going on here.
00:45:21.875 --> 00:45:28.615
So a lot of these AI companies are making commitments to say how few people
00:45:28.615 --> 00:45:34.195
they actually need per the revenue per employee that they're going to dictate.
00:45:34.355 --> 00:45:38.415
And there's this company called Agency AI that has made a formal commitment
00:45:38.415 --> 00:45:41.955
to cap their full-time headcount at 100 people.
00:45:42.015 --> 00:45:45.855
And it's the company's guiding principles, basically saying no matter how much
00:45:45.855 --> 00:45:49.275
the company grows, they're never going to have more than 100 people.
00:45:50.095 --> 00:45:53.395
And so there's a lot of different things going on here. They give a few examples,
00:45:53.735 --> 00:45:59.875
MidJourney, Cursor, CalAI, 11 Labs of how much revenue that they have and how
00:45:59.875 --> 00:46:02.335
few employees the organization actually has.
00:46:02.455 --> 00:46:08.155
But I think about this in terms of what we show here isn't really about HR per se.
00:46:09.217 --> 00:46:13.577
With AI coming and potentially deconstructing a lot of the work that's done,
00:46:13.817 --> 00:46:20.317
is there a potential in the future that the whole concept of people analytics
00:46:20.317 --> 00:46:25.097
or of even like human variability that we've talked about in this podcast isn't
00:46:25.097 --> 00:46:28.037
really necessary because the sample sizes are too small?
00:46:28.237 --> 00:46:34.377
Because organizations themselves are such small organizations that the traditional,
00:46:34.957 --> 00:46:39.457
analyses that we've done that required large sample sizes aren't really relevant anymore.
00:46:39.697 --> 00:46:43.357
And so I wanted, I thought the two of you are some of the most brilliant people
00:46:43.357 --> 00:46:45.077
in talking about these things.
00:46:45.237 --> 00:46:47.777
What do you see that's happening? What do you see that's coming?
00:46:47.977 --> 00:46:52.277
And how does this overlay with, you know, some of the work that you've done in the past?
00:46:52.637 --> 00:46:58.517
Well, I think it is. I mean, my first reaction to that is it's certainly attention-getting,
00:46:58.517 --> 00:47:01.177
And that may be the, that may be the point.
00:47:01.797 --> 00:47:07.117
Again, I think I always find it illuminating to take the analogy to something
00:47:07.117 --> 00:47:13.437
that isn't people and ask if the executive would feel the same way about being public about that.
00:47:13.577 --> 00:47:16.697
Right. So what if we, you know, pick anything, right? We're going to,
00:47:16.797 --> 00:47:21.897
we're going to never use more than 100 suppliers, no matter how big we get.
00:47:22.597 --> 00:47:28.097
Because, you know, we feel that efficiency in supplier management is very,
00:47:28.237 --> 00:47:30.077
very important to strategic success.
00:47:30.697 --> 00:47:35.597
And so that, you know, so, so, so, and I think if you said that you would be
00:47:35.597 --> 00:47:41.197
presented with a requirement that you walk that through a rigorous model, you know, well,
00:47:41.357 --> 00:47:44.417
what, you know, what, what is it, what is our relationship with suppliers?
00:47:44.577 --> 00:47:46.217
How, how do they create value?
00:47:46.437 --> 00:47:49.337
What is the cost of having more than a hundred suppliers? You know,
00:47:49.457 --> 00:47:51.497
what, what point does it tip over to be inefficient?
00:47:51.897 --> 00:47:55.237
To have 3,000 different suppliers. And of course, yeah, you know,
00:47:55.317 --> 00:48:00.217
there's a bureaucratic point at which the value of having more of those suppliers
00:48:00.217 --> 00:48:03.117
just doesn't justify the cost.
00:48:03.477 --> 00:48:08.077
But is it 100? Would you come out with that number in particular and say we'll
00:48:08.077 --> 00:48:09.177
never be bigger than that?
00:48:09.337 --> 00:48:14.197
And so my reaction is kind of, well, it's attention-getting,
00:48:14.237 --> 00:48:18.277
but I think I would wish for a much more rigorous approach to that.
00:48:19.577 --> 00:48:24.977
And it may be less. It might be more, but that we step back and we really ask,
00:48:25.097 --> 00:48:31.637
what is the framework we use to decide when adding people actually creates greater
00:48:31.637 --> 00:48:33.957
value, particularly if you're a growing organization?
00:48:34.277 --> 00:48:38.317
I also think what it really begins to tee up is executives are trying to get
00:48:38.317 --> 00:48:44.817
their head around having AI or automation be a part of the strategic planning
00:48:44.817 --> 00:48:49.237
process for, let's call it talent or capacity.
00:48:49.237 --> 00:48:56.657
And so what happens when you have AI as a part of your human capital portfolio
00:48:56.657 --> 00:48:59.057
or automation or something like that?
00:48:59.177 --> 00:49:03.297
You see it in the Shopify example saying you don't get to hire until you prove
00:49:03.297 --> 00:49:06.777
to me that AI couldn't do the thing you're hiring for, you know,
00:49:06.897 --> 00:49:08.897
sort of turning that completely on its head.
00:49:09.578 --> 00:49:17.058
I was just on an investor call with a company that has an AI agent,
00:49:17.258 --> 00:49:18.618
basically, right, that they built.
00:49:19.118 --> 00:49:22.418
And it's an agent that can do the kinds of things that first-year law students
00:49:22.418 --> 00:49:25.978
can do, first-year market researchers can do, that kind of thing, right?
00:49:26.118 --> 00:49:30.898
Pretty defined kind of work that you can tell an agent, go do this,
00:49:31.058 --> 00:49:34.938
right? like go look up all the cases that are relevant to this particular court
00:49:34.938 --> 00:49:40.718
case, or go do a market analysis of our competitors, both of which are really well-developed.
00:49:40.818 --> 00:49:42.658
Now, how do they go to a company,
00:49:42.898 --> 00:49:43.098
Right?
00:49:43.638 --> 00:49:48.518
They've chosen to go to companies by having the AI apply for the job.
00:49:48.918 --> 00:49:54.198
So if you're a hiring manager, you get an applicant from this organization,
00:49:54.198 --> 00:49:59.938
and you ask it questions, and the AI responds to your questions and you decide
00:49:59.938 --> 00:50:04.818
if you want to proceed with that person or that entity into the next stage of
00:50:04.818 --> 00:50:05.778
interviews or whatever.
00:50:06.258 --> 00:50:10.378
And at the end of the day, you know, the organization recognizes this is an
00:50:10.378 --> 00:50:12.938
automated agent that answered the questions perfectly.
00:50:13.598 --> 00:50:17.918
And that seems pretty well suited to doing that kind of first year work.
00:50:18.038 --> 00:50:22.638
You know, so I think part of what's going on here is how are we going to think
00:50:22.638 --> 00:50:26.698
about AI just to take an example of automation and what if we thought about
00:50:26.698 --> 00:50:31.438
it as actually a part of the what we might call the labor pool or the work yeah
00:50:31.438 --> 00:50:35.858
you know how much of your planning is yeah go ahead scott it's
00:50:35.858 --> 00:50:39.058
Like just another note of the network right at that point yeah like i wonder
00:50:39.058 --> 00:50:43.778
how much of it is like signaling as well like we are a forward-thinking company
00:50:43.778 --> 00:50:47.758
and the people we do have here you're highly valued because you're one of the
00:50:47.758 --> 00:50:49.338
hundred yeah or whatever
00:50:49.338 --> 00:50:55.498
Yeah yeah i'm not gonna trust the AI agent has passed the Turing test until
00:50:55.498 --> 00:51:00.078
it starts complaining about applicant tracking systems and how hard it is to
00:51:00.078 --> 00:51:03.158
apply and how they have to update their resume every time.
00:51:03.318 --> 00:51:05.358
That's when we'll really know that they're good.
00:51:05.918 --> 00:51:06.598
Oy vey.
00:51:08.338 --> 00:51:08.898
Yeah.
00:51:09.638 --> 00:51:12.358
Yep. Lobbying for return to punch cards.
00:51:13.518 --> 00:51:17.378
I would love it if we found out this company only has like 10 employees.
00:51:17.758 --> 00:51:21.718
They're just like out here, like shooting off the cuff here.
00:51:22.226 --> 00:51:27.286
My first reaction to that, there's really good economic theory about why certain
00:51:27.286 --> 00:51:30.326
industries have companies of certain size and other industries have companies
00:51:30.326 --> 00:51:31.426
that don't have different size.
00:51:31.646 --> 00:51:36.846
And it has to do, just a simple example, why do we only have five or six mega
00:51:36.846 --> 00:51:39.946
CPA firms and we have dozens and dozens of large law firms?
00:51:40.046 --> 00:51:42.586
And it has to do with the structure of the industry and things like that.
00:51:42.906 --> 00:51:46.326
And I think AI will change the structure of industries without any doubt.
00:51:46.526 --> 00:51:49.666
But it's pretty easy to, not easy, but it's pretty straightforward to apply
00:51:49.666 --> 00:51:54.086
that knowledge that we already have. and say, how will AI change the structure
00:51:54.086 --> 00:51:57.186
of companies within an industry on a number of factors?
00:51:57.306 --> 00:52:01.646
There's some really good theory around that that can be applied to what's the
00:52:01.646 --> 00:52:02.866
change now that we have AI.
00:52:03.506 --> 00:52:05.806
Well, Scott, what do you want to hit us with next?
00:52:06.186 --> 00:52:09.926
Let's go to surprising combinations of research.
00:52:10.226 --> 00:52:15.626
Contents and contacts are related to impact. I have very few articles these
00:52:15.626 --> 00:52:19.226
days that really get me going, but I read this and I thought it was absolutely
00:52:19.226 --> 00:52:24.506
amazing. So like think about those like really big ideas in science and technology.
00:52:24.926 --> 00:52:28.346
They really come to life when you have different worlds colliding.
00:52:28.546 --> 00:52:32.746
So this paper looks at essentially tens of millions of citations and patent
00:52:32.746 --> 00:52:35.366
data. And they use this hypergraph model.
00:52:35.586 --> 00:52:38.946
Scroll down a little bit, Cole. Like there's a really cool visual down here.
00:52:39.666 --> 00:52:42.986
Yeah. So this hypergraph model essentially takes the contents,
00:52:43.086 --> 00:52:46.426
like the keywords of different citations and patents.
00:52:46.466 --> 00:52:50.606
It matches it up to the context. So it's like where it was actually published
00:52:50.606 --> 00:52:56.506
and they were able to predict with 95% accuracy, like if something was going to be like a hit or not.
00:52:56.646 --> 00:53:01.166
So the research with the most outsized impact, so the most citations or,
00:53:01.226 --> 00:53:06.626
you know, the highest cited patents was strongly like, like linked to unexpected combinations.
00:53:06.986 --> 00:53:11.926
It's most often when researchers published in one field about problems in another field.
00:53:12.006 --> 00:53:16.126
So I think about Pete here, like he's a finance guy applying it to the world of IO.
00:53:16.126 --> 00:53:21.846
So a novelty is a strong indicator of future influence, but you can make an
00:53:21.846 --> 00:53:25.806
impact on either these areas, both in the like contents that you provide,
00:53:25.986 --> 00:53:28.806
the novel contents or novel context separately.
00:53:28.806 --> 00:53:31.826
So it feels like it's hitting on two different areas.
00:53:32.726 --> 00:53:37.526
Let's see. Scientists tend to be five times more likely to cite from their home
00:53:37.526 --> 00:53:41.446
journals. So they kind of signal that they like they like IOP or,
00:53:41.526 --> 00:53:46.706
you know, whatever sort of journal you have here. and surprising teams with diverse backgrounds.
00:53:46.906 --> 00:53:49.666
These like expeditions is what they call them. Published
00:53:49.666 --> 00:53:50.586
Work for another.
00:53:50.586 --> 00:53:53.766
Field tend to make the biggest waves as possible.
00:53:54.046 --> 00:53:58.326
But like a really, really cool sort of study going on here about how we can
00:53:58.326 --> 00:54:00.986
actually improve other fields and like kind of mix things up
00:54:01.665 --> 00:54:05.585
Yeah, I remember this made me think about this a little while ago.
00:54:05.705 --> 00:54:09.745
We had Ryan Hammond on as a guest and he had a background in org theory.
00:54:10.185 --> 00:54:13.625
And one of the things that he mentioned is he had a colleague of his,
00:54:13.785 --> 00:54:15.225
I believe, was getting a Ph.D.
00:54:15.305 --> 00:54:21.585
In economics. and their professor threw down on their desk a book from org theory
00:54:21.585 --> 00:54:25.685
and said, read through this somewhere in here is your dissertation.
00:54:26.445 --> 00:54:30.345
And it's basically like saying, if you were able to cut across disciplines,
00:54:30.345 --> 00:54:34.625
you're going to be able to find a really cool and unique dissertation.
00:54:34.945 --> 00:54:38.485
And so I thought that this is a fascinating line of research.
00:54:38.645 --> 00:54:43.365
And you guys, again, like what Scott said, really emulate this effectively,
00:54:43.365 --> 00:54:45.265
but I'd love to get your thoughts on it.
00:54:45.265 --> 00:54:49.345
Well, I think I would say that that's been my experience.
00:54:49.825 --> 00:54:55.745
That for me, I mean, it's not that there haven't been enormous insights that
00:54:55.745 --> 00:54:59.265
have come to me through what we might call the traditional HR area.
00:54:59.345 --> 00:55:03.705
And I've been very, you know, very fortunate to have worked in great institutions
00:55:03.705 --> 00:55:08.405
and with great mentors and others in that particular field. But I would agree
00:55:08.405 --> 00:55:12.925
that it's the cross-disciplinary stuff where I have found my eyes being opened.
00:55:13.285 --> 00:55:17.285
I had a period where I was walking across the street to the business school
00:55:17.285 --> 00:55:23.485
at Cornell, working with a bunch of engineers and others who were operations research engineers.
00:55:23.485 --> 00:55:28.325
And you'd go in and you'd say, well, have we thought about the reactions people have?
00:55:28.365 --> 00:55:32.125
I had one really talented PhD student, and he said, I don't think we've looked
00:55:32.125 --> 00:55:36.405
enough at how people react to what they see going on around them,
00:55:36.525 --> 00:55:37.845
like on an assembly line.
00:55:38.525 --> 00:55:42.125
And what he found was that the mathematics of the assembly line didn't really
00:55:42.125 --> 00:55:46.705
reflect that the person doing the work would be able to see stuff piling up
00:55:46.705 --> 00:55:50.905
to their left or to their right and slow down or speed up their work,
00:55:51.005 --> 00:55:52.365
which is a perfectly natural thing.
00:55:52.525 --> 00:55:56.365
Like, right, you know, if you see the supply next to you that you have to work
00:55:56.365 --> 00:55:57.965
on going up, you're going to work faster.
00:55:58.305 --> 00:56:02.645
If you see that your work is piling up after you and the person next after you
00:56:02.645 --> 00:56:05.925
isn't getting through the pile, you're going to slow up a little bit.
00:56:06.751 --> 00:56:10.471
And it turned out that when you put those things in, you had a much better predictor
00:56:10.471 --> 00:56:16.491
of how the supply chain worked and why it worked, even when the math might say,
00:56:16.611 --> 00:56:18.631
no, no, no, you're going to have big pileups here.
00:56:19.271 --> 00:56:23.751
And you'd say, well, no, we're not getting those. Well, what could possibly explain that, right?
00:56:24.091 --> 00:56:29.491
And it was the psychology of the person doing that work.
00:56:29.791 --> 00:56:34.991
And I think a lot of the metaphors we're seeing today have to do with that crossover,
00:56:34.991 --> 00:56:38.211
whether it's to math or economics or accounting.
00:56:38.491 --> 00:56:43.711
Again, for me, I talk about leaders' smart brains and how they come at disciplines
00:56:43.711 --> 00:56:49.831
like production or engineering or finance or accounting even with mental models
00:56:49.831 --> 00:56:53.211
that are far more sophisticated than they are when they think about people.
00:56:53.491 --> 00:56:57.791
And the power of looking across disciplines, and this has really been really
00:56:57.791 --> 00:57:01.371
impactful for me over the years, is you can look at different kinds of evolutions.
00:57:01.671 --> 00:57:07.651
So the evolution management system, I've talked a lot we've written about how
00:57:07.651 --> 00:57:10.811
finance systems you can learn a lot about how time zones evolve
00:57:11.931 --> 00:57:15.571
there's a lot of facts around time zones evolved
00:57:16.251 --> 00:57:20.931
to what's going on in people analytics 100 years from now a lot of information
00:57:20.931 --> 00:57:24.951
that came through what the land grant universities and how we manage crops to
00:57:24.951 --> 00:57:31.591
what not just refinance but the whole idea of deals for ratings and things like that
00:57:32.311 --> 00:57:35.271
there's just a lot of So not only can you cross disciplines,
00:57:35.451 --> 00:57:42.231
but you can cross decades or centuries and then learn from how those things
00:57:42.231 --> 00:57:46.851
evolved and make some pretty good predictions about what are some of the things
00:57:46.851 --> 00:57:50.031
that should or should not work and how people on analytics should evolve and
00:57:50.031 --> 00:57:51.331
could evolve along the way.
00:57:51.331 --> 00:57:54.731
I love Douglas Hubbard's book, how to measure anything.
00:57:54.911 --> 00:57:57.851
And like one of his tenets is, well, if you have a problem, like go look at
00:57:57.851 --> 00:58:00.351
another field, they've probably already dealt with it and they probably dealt
00:58:00.351 --> 00:58:01.451
with it better than you could.
00:58:02.011 --> 00:58:04.911
You know, so like, yeah, probably problems already been solved.
00:58:06.351 --> 00:58:08.991
I'm going to call this episode why time zones evolved.
00:58:14.111 --> 00:58:17.571
Well, when it took you, when it took you five months to get from New York to
00:58:17.571 --> 00:58:19.731
California, it didn't matter what time it was when you got there.
00:58:19.731 --> 00:58:23.731
Yeah, that makes sense So the railroads were the big drivers
00:58:24.918 --> 00:58:29.618
Well, it sounds like that's a very directionally correct response.
00:58:29.938 --> 00:58:32.858
Scott, you want to hit us with our last article today?
00:58:33.238 --> 00:58:37.978
Five reasons why leaders never say directionally correct. So this guy essentially
00:58:37.978 --> 00:58:41.678
argues that saying directionally correct is just kind of a reckless term.
00:58:41.858 --> 00:58:44.778
I'll quickly run through his rationale. Five different reasons.
00:58:45.218 --> 00:58:50.278
Directionally correct has no agreed upon definition, doesn't even mean what people think it means.
00:58:50.898 --> 00:58:54.758
Directionally correct decisions may or may not lead to good business outcomes.
00:58:54.978 --> 00:59:00.258
Using the phrase directionally correct degrades data culture by promoting incomplete data analysis.
00:59:00.718 --> 00:59:05.438
And directionally correct analysis never provides sufficient explanatory power
00:59:05.438 --> 00:59:07.578
that holds anyone accountable.
00:59:07.898 --> 00:59:09.658
Sorry, something just fell in the room here.
00:59:10.698 --> 00:59:12.558
Well, let me ask you guys.
00:59:13.518 --> 00:59:16.398
Did someone throw a rock through the window or something?
00:59:17.378 --> 00:59:18.618
We're right at time.
00:59:18.938 --> 00:59:21.518
With that article wrapped around it, right?
00:59:23.818 --> 00:59:28.538
Well, I see you guys both as leaders. Do you ever say the words directionally correct?
00:59:28.798 --> 00:59:34.478
My former boss, CEO, his favorite quote, he said it once a week,
00:59:34.498 --> 00:59:38.738
if not more often, change is good as long as it's in the right direction.
00:59:38.898 --> 00:59:40.498
I think he attributed it to Churchill.
00:59:41.978 --> 00:59:46.218
But whenever someone had a proposal, it was usually a way to be skeptical about
00:59:46.218 --> 00:59:47.258
the proposal. You know what I mean?
00:59:47.678 --> 00:59:51.998
Well, change is good, but is it in the right direction? So I lived with that
00:59:51.998 --> 00:59:59.018
for a number of years as the CEO's decision-makers go-to framework as a way
00:59:59.018 --> 01:00:00.938
to think about change for itself is not good.
01:00:01.038 --> 01:00:03.838
But anyway, that's my initial reaction when I see that.
01:00:04.618 --> 01:00:09.878
It's a spectrum, right? Like you can do the most deep analysis in the world,
01:00:10.038 --> 01:00:11.598
right, and come up with a perfect answer.
01:00:12.138 --> 01:00:15.978
But it'll be six months down the line. And if you need an answer today,
01:00:16.978 --> 01:00:18.618
like what are you going to do with that?
01:00:18.618 --> 01:00:24.418
The same CEO was very good about saying, I don't want any uncertainty that's
01:00:24.418 --> 01:00:26.758
coming from lack of deep thinking.
01:00:28.719 --> 01:00:29.839
Yeah we're gonna have uncertainty
01:00:29.839 --> 01:00:30.479
Yeah yeah yeah
01:00:30.479 --> 01:00:37.339
But but i i do not want to accept uncertainty that results from his team not
01:00:37.339 --> 01:00:39.699
doing their homework and i thought that was really insightful
01:00:39.699 --> 01:00:41.479
That's a good phrase i like that
01:00:41.479 --> 01:00:45.579
You know so it's a corollary of direct but that was his that was his counterpoint
01:00:45.579 --> 01:00:50.099
is yeah we're always gonna have risk but i don't want risks that absolutely
01:00:50.099 --> 01:00:53.019
are manageable with with good thinking and good analysis
01:00:53.019 --> 01:00:56.719
In advance well that's the whole point of analysis right like is uncertainty
01:00:56.719 --> 01:01:02.299
reduction right it's not uncertainty elimination right like we'd all be millionaires
01:01:02.299 --> 01:01:03.679
and billionaires if we could
01:01:03.679 --> 01:01:09.179
Well well i'll probably regret saying this but carlos kimony the guy who wrote
01:01:09.179 --> 01:01:13.299
the article in 2020 about directionally correct you have open invite to come
01:01:13.299 --> 01:01:17.219
on the podcast and we'll debate you that directionally correct is actually a
01:01:17.219 --> 01:01:22.719
great term so well done but well john and pete you guys are fantastic guests today.
01:01:23.019 --> 01:01:26.739
Thank you so much for joining us. Scott, any final words before I give John
01:01:26.739 --> 01:01:27.639
and Pete the final words?
01:01:27.999 --> 01:01:31.859
Oh, Pete, John, man, hang in there. Things will work out eventually.
01:01:32.119 --> 01:01:35.179
I believe in you guys. Just hang in there.
01:01:35.379 --> 01:01:39.819
I feel better already. Thank you, Scott.
01:01:41.319 --> 01:01:44.019
I just want to say thank you. And I appreciate what you guys are doing with
01:01:44.019 --> 01:01:47.339
these podcasts and the kind of guests you have, the kind of questions you ask and stuff.
01:01:47.419 --> 01:01:50.299
And I think it's really, really interesting stuff.
01:01:50.299 --> 01:01:52.679
A lot of fun today. A lot of fun. Thanks for coming.
01:01:52.819 --> 01:01:55.419
I really appreciate that. Thank you. Coming from you, that's high praise.
01:01:55.579 --> 01:01:57.479
That's Directionally Correct praise right there for you.
01:01:59.679 --> 01:02:03.019
Well, you've been listening to Directionally Correct, a People Analytics podcast
01:02:03.019 --> 01:02:05.239
with Colin Scott, and today's guests, Dr.
01:02:05.519 --> 01:02:08.139
John Boudreaux and Pete Ramstad. Thanks for joining us today, guys.