December 22, 2023 • 30 mins
How does a company harness job market trends and use data to boost their recruitment process? Simply ask Cory Stahle, an expert from Indeed. In his discussion with William Tincup, Cory shares some fascinating trends and impactful findings from Indeed's 2024 U.S. Jobs and Hiring Trends Report.
As a well-versed economist, Cory starts off by shedding light on how Indeed handles data on an extensive scale. He shares the fact that their data collection process is high-frequency, comprehensive, and grounded in real-time, offering a more precise perspective on job postings than traditional government surveys. He also underscores the fact that Indeed provides its datasets to the public, enabling users to leverage this pool of information and spot patterns in recruitment trends.
Cory's analysis reflects a job market that is undergoing significant shifts as technology and economic trends unfold. His insights underscore the potential for data and AI to revolutionize recruitment - a testament to the power of digital advancement in the hiring space. So, if you're looking to gain a competitive edge in recruitment, perhaps it's time to start exploring the data at your fingertips.
Listen & Subscribe on your favorite platform
Apple | Spotify | Google | Amazon
Visit us at RecruitingDaily for all of your recruiting, sourcing, and HR content.
Follow on Twitter @RecruitingDaily
Attend one of our #HRTX Events
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
William Tincup (00:32):
This is William Tincup, and you're listening to
the Recruiting Daily podcast.
Today, we have Corey on fromIndeed, and we'll be talking
about Indeed's 2024 U.
S.
Jobs and Hiring Trends Report.
So, uh, this is going to bereally, really fun.
Can't wait.
And so let's just jump into someintroductions.
Uh, Corey, would you do us afavor and introduce yourself and
(00:55):
what
Cory Stahle (00:55):
you do at Indeed?
Yeah.
Thanks, William.
It's great to be here.
So, just by way of introduction,I'm sure a lot of listeners have
maybe heard of Indeed and kindof the job searching side of
things.
If they haven't,
William Tincup (01:08):
Corey, at this point, I really feel sorry.
Cory Stahle (01:11):
You know what I'm saying?
I know.
I know.
It's, uh, it's pretty rough outthere.
It's like that progressive.
William Tincup (01:15):
What is that?
You've been living under a rockfor, you know?
Yeah.
But anyway,
Cory Stahle (01:19):
I'm sorry to interrupt.
Uh, yeah, no, no, uh, you know,so yeah, I mean, you talk about
Indeed, you know, a lot ofpeople think, you know, jobs,
job postings, and certainlythat's a big part of what we do,
you know, and kind of coming upwith the, you know, those kind
of connecting kind of thishiring between employers and job
seekers, uh, but we also have alot of other functions.
And I get to fortunately sit inone of those kind of little bit
(01:42):
of a different kind of funthings that we do at Indeed,
which is I work with what'scalled the hiring lab.
Which is a group and team ofeconomists that uses the data
that we have on Indeed to reallykind of track what's going on in
the labor market and in theeconomy kind of broadly.
And so that's kind of as we talkabout, um, stuff today, you
know, that's kind of wherethat's coming from is we're
(02:03):
looking at kind of what thetrends have been on Indeed and
in hiring in recent months, inrecent years, and then looking
at kind of where those trendsmight lead us going forward.
So
William Tincup (02:13):
ADP has, um, and they've done this for years is
because they have payroll.
So they can kind of tell people,uh, with the data that they're
sitting on kind of what's goingon with payroll, what's going on
with, you know, those types ofthings.
And y'all are sitting on amassive amount of doc, of data
as well in terms of who'shiring, uh, who's posting job
(02:34):
ads, et cetera.
Could you, could you tell us alittle bit more about the data
that y'all sit on?
Cory Stahle (02:40):
Yeah, so I mean, yeah, great, that's, uh, so we
got a lot of different data.
I would say that with theeconomic data that we're looking
at, a lot of what we're lookingat is job postings, you know, so
this isn't necessarilyproprietary data in terms of,
you know, these are job postingsthat job seekers can go on to
look at, Bye bye.
Bye bye.
But if you want to kind of get agood feel for what's going on,
(03:03):
on the larger picture, it wouldbe pretty difficult to comb
through millions and millions ofthese different job postings and
to break them out, you know,into, you know, this sector is
hiring more, this sector ishiring less.
That's something that would bevery, very hard to do as a job
seeker or an employer orsomebody working in recruiting
or HR.
So that's what we really do iswe take.
That data we have, we take thedata, you know, the job postings
(03:25):
that employers are posting, andfrom that we can tell a lot
about, you know, our sector'sposting more or less than they
used to.
You know, we can also look at,within a job posting, we can
see, you know, how much areemployers asking for.
Generative AI type of tools, youknow, how much are employers
advertising pay in a job postingand is that pay going up or
(03:45):
down?
So there are a lot of economicindicators that we can actually
draw out of job postings, youknow, as we start to kind of
roll up and look at those at ahigher
William Tincup (03:54):
level.
I love that because you can givepeople, you could paint a
picture about skills and comp ina way that.
Other folks that don't have thatdata set.
Um, can you, uh, can you, canyou kind of talk to about
traffic?
Like again, there's the postingitself and then there's
obviously people that then gointo the posting and do
something.
(04:14):
Is there, is there data therethat's meaningful or even
candidate flow that kind of gofurther into people that have
applied, etc.
Is that useful to the, uh, tothe audience?
Cory Stahle (04:26):
Yeah, so that is something that we've done a
little bit of work and, uh, justclarify, I think if I understand
correctly, you're asking aboutkind of people who click on job
postings, you know, kind of theflow of job seekers.
So yeah, so that is somethingthat we've looked at
historically.
I know one example of thatrecently is we looked at The
share of clicks on job postings,you know, so we're looking at
(04:47):
people who are interested in jobpostings in the United States.
In this case, we were lookingand we wanted to see how many
people who are clicking on aspecific job posting were living
overseas, you know, so we cankind of track foreign interest
and kind of get an idea.
You know, we're not necessarilygetting into the U.
S.
You know, really individualized,personalized data, like we take
(05:07):
protecting personal informationreally important, you know,
really seriously, um, but again,we kind of roll up, we look at
the aggregate level and, youknow, we can then have trends
like we see that, you know,during the pandemic, foreign
interest, you know, in U.
S.
jobs, you know, was pretty flat,but we've seen that actually
grow pretty significantly in thelast couple years, and I think
that's been a A prettysignificant part of the story
(05:30):
behind the resilience in thelabor market, you know, so we
can really take these things andwe kind of tie them together and
get a picture, you know, and onthe job posting side, we have
public data sets that wepublish, one's called the Indeed
Job Postings Index, so that'swhat allows us to kind of track
how job postings are moving,anybody can access that.
Um, on our website, which ishiringlab.
(05:51):
org, um, and then we also havewhat's called the Indeed Wage
Tracker, which I alluded toearlier, which gives a
measurement of kind of how wagesand comp and job postings are
changing over time.
Um, so again, just helps us tocreate more of a full picture
than I think governmentstatistics, uh, Right.
You know, where, where they kindof maybe, maybe can lack
(06:11):
sometimes, especially in termsof frequency.
So.
William Tincup (06:14):
I love it.
I love it.
First of all, I love that you'redoing something with the data to
help both the larger audience,but also the folks that, that
are indeed customers as well,like giving them some insight
into all of those things, skillsand wages and everything.
It's like.
Because left to their own, they,they might not know that these
macro issues are at play.
(06:35):
So I, first of all, I just lovethat.
Um, let's talk a little bitabout the report.
Uh, the, the 24 US jobs andhiring, uh, trends report.
What, um, if we were to rollthis up for folks with short
attention spans, which would bepretty much everybody, um, what
are some of the things thatyou'd like to highlight?
Cory Stahle (06:56):
Yeah, so I think, you know, kind of the, uh, quick
rundown of it, I think, settingthe stage for this report, I
think is important to understandas we were kind of preparing
this report, we were looking atwhat was going on, kind of in
this macro economy, this currenteconomic environment, and kind
of what's happened in 2023, andwhat's been really interesting
is, you know, we've seen a lotof different discussions in
(07:18):
2023, you know, we've heard, youknow, a lot of discussions
around return for rural.
return to office, generative AI,you know, how that's going to
maybe change the labor market.
You know, so there's been a lotof things to talk about that
have happened in 2023, but Ithink it's important to note
what did not happen in 2023, orat least, you know, this far.
We've still got, you know, amonth left here, but, uh, but
(07:39):
what we haven't seen based onthe data we're looking at is a
recession.
You know, we saw so many people,you know, making recession calls
at the end of last year.
And as we, you know, have gonethrough the year, the labor
market and by and large thewhole economy has been pretty
resilient.
And so really that's kind of thesetting the stage for this
report.
So the report basically issaying, okay, what are the
(08:01):
trends that we saw in 2023 thatneed to carry forward into 2024?
You know, what is it that, youknow, employers need to do in
terms of their demand?
You know, what do we need to doin terms of attracting people,
you know, into the labor market?
You know, what does.
Quitch rates and like thedynamics of the labor market
look like in a healthy labormarket.
(08:22):
You know, and then also wetalked a little bit about wage
growth.
And I think also generative AI,you know, is another trend that
we'll really be watching.
So that's kind of the stage.
So I would say in short, uh, youknow, here's kind of the, the
wrap up really, really shortversion.
You know, what is it that needsto go right in 2024?
We need to continue to see, youknow, strong demand from
employers, more people enteringinto the labor market.
(08:45):
And we need to, you know, keepour eye on some of these
emerging things like generativeAI.
William Tincup (08:50):
So, um, and I'm just probably done in the, in
the report, but did you, are webifurcating between hourly and
corporate or salaried?
Uh, in any way, because I, youknow, what I hear usually from
practitioners is more, uh, onthe hourly side, it's, it is,
it's, it's very difficult andseems to be getting more
(09:11):
difficult, but on the salariedside, depending on the job or
wherever, uh, not as hard.
There's like a, and again,that's just, you're sitting on
data.
I'm sitting on anecdotes.
So.
I'm going to trust you more thanthey trust me, but basically,
basically it's kind of like thestory of like, okay, if you're
hiring for salary, you're goingto have some options.
(09:31):
Got it.
If you're hiring for hourly,you've got to re your, your
options are limited and you'vegot to raise the pay.
Cory Stahle (09:39):
Yeah.
Yeah.
And I think ultimately, youknow, as we look across indeed,
about 50%, a little over 50percent of jobs now include a
salary, you know, partially as aresult of different kind of pay
transparency regulations in somestates.
Um, but that means that thereare almost 50 percent that still
do not contain a salary.
So we don't necessarily havelike the breakout, um, in those
(10:00):
terms, but what we do look at iswe kind of look at, and I think
this is another interestingproxy.
We look at.
The breakout between jobs thatare more likely to be done
remote versus the ones that areless likely to be be done remote
and I would say there'sprobably, you know, some
comparison there to be madelower remote jobs tend to be a
little more skilled labor,probably a little more, you
(10:22):
know, in person hourly typejobs, whereas some of those
highly remote jobs tend to bemore kind of knowledge worker
type positions, um, andpotentially more salary jobs.
And I think as we kind of Talkabout that.
We have seen, you know, reallythe bifurcation this year in the
labor market, you know, wherecoming out of the pandemic, you
(10:42):
know, all sectors rose together.
They all sunk together duringthe pandemic and they all rose
together.
But what we've seen in 2023 iskind of this, this kind of fork
in the road to some degree.
You know, we've seen softwaredevelopment jobs drop below
their pre pandemic levels,whereas we've seen, you know,
jobs in areas like nursing andconstruction and manufacturing,
(11:04):
you know, keep, you know, prettyhigh levels of employer demand,
you know, which ultimately Ithink relates back to what
you're talking about in terms ofthe difficulty to find workers,
um, as, you know, there's just,you know, stronger demand that
we're seeing in some of thosetype of low remote in person
type of opportunities.
So
William Tincup (11:21):
did y'all, uh, just as a matter of a point of
honor, is did y'all?
Do this report last year at
Cory Stahle (11:27):
the year?
We did actually.
Yeah.
So we did produce kind of fivetrends to watch last year.
Um, so that is something thatpeople can kind of go back and
look.
I would say last year we kind offocused on a long term outlook.
Um, kind of the takeaway was.
You know, demographics arereally weighing on the United
States labor market andpopulation in general.
(11:51):
You know, we have this agingpopulation, which over the
course of decades, unlesssomething changes, you know,
with the birth rate, we're goingto see labor markets continue to
be pretty tight over the comingdecades.
You know, business cycle mightgo up and down, but if you have
fewer workers to take jobs,that's going to create different
challenges in the future.
So we took that long termperspective, but for this year,
(12:12):
we kind of wanted to focus.
Uh, mainly on 2024, so, so yeah,so we have kind of this kind of
long look last year versus theshort look this year, which I
think in some ways they kind ofmake a good pairing of reading
together.
Oh, a
William Tincup (12:24):
hundred percent.
I always love it when people goback and look at like their
predictions.
And, and analyze theirprediction, they're very
critical, analyze theirpredictions like, yeah, wildly
missed.
Uh, yep.
Got that one.
Right.
Got that one.
Right.
This one's kind of, we're still,you know, we're still almost
there.
Um, so, so first of all, where'sthe, uh, where's the report
(12:45):
located?
It's obviously on the
Cory Stahle (12:47):
website.
Yeah, yeah, so people can go tohiringlab.
org.
Um, and so we've got not onlythis report, um, but also like
the data sets I've mentioned.
So like hiringlab.
org slash data, um, is a, a dataportal where people can kind of
do their own analysis, so tospeak, and you can kind of sort.
And I love getting in there andjust playing with it because you
(13:08):
can look how job postings aredifferent between different
states.
You can look how they'redifferent between different like
metropolitan areas.
You can look how job postingsvary by different sectors, you
know, we've talked a lot aboutsectors, you know, and so it can
be really helpful to go in thereand to, you know, kind of slice
up the data that you want.
Um, but yeah, as far as thesereports, they're all just up on
hiringlab.
org along with a lot of otherresearch that we publish using
(13:31):
this data pretty regularly.
Oh, I
William Tincup (13:33):
love that.
And I love, I especially lovethe fact that they can, um, use
your, use your aggregate data tothen model and think about
themselves.
Cause that's, you know, that's,that's for them, especially, uh,
CH, uh, CHROs and VPs of talent.
They're trying to figure outforecasts.
They're trying to figure outlike what is going on.
(13:55):
We know what's going on withinthe four walls of our company,
but we have no idea what's goingon the four walls of the United
States.
And so we might learn somethings at an industry event or
something like that, but toactually play, I say play, to
actually use aggregate data inthis way, it's just, first of
all, Thanks, because I don'tknow of anybody else that does
(14:15):
that, that allows people to, tolook and kind of do their own
research, uh, based on such alarge data set.
Cory Stahle (14:24):
Yeah, no, I love it.
I think it's absolutely amazing,and I would say from a past
life, I used to work as aneconomist in state government,
um, and a big part of my jobWith that was to go around to
different businesses and HR, youknow, and heads of talent and
to, you know, kind of talk aboutthe tools that the government
had available, you know, andthings that were out there, um,
(14:45):
for businesses to use and thegovernment has some really,
really good, you know, reallygreat tools as you look at it,
but then Now coming into myposition within Indeed, I'm
saying, you know, there's a lotof value in the government
tools, but it's really hard tobeat the scale and the size of
what we're doing at Indeedbecause, you know, you talk
about, you know, the governmentdata, you know, we, there's been
(15:07):
a lot of discussion lately aboutgovernment data surveys, having
problems with response rates,and the result has been a little
bit of noisiness in the series,and then you compare that to the
Indeed data where we're talkingabout basically, you know, Any
job posting that's on theinternet, it's probably being
included in this analysis.
You know, we're really pullingtogether a near universe of
(15:28):
jobs, so you can get like areally good picture.
You don't have to worry aboutsurveys.
And not only that, but I mean, Ican look at what the data was
doing, you know, last Friday,rather than waiting, you know,
three or four months.
Because I know one of thebiggest employment data sets,
you know, in from the governmentis the QCEW data.
And that's kind of the bestcensus overall universe of
(15:51):
employment.
But that tends to have like asix month delay, you know, in
six months when you're trying toforecast and kind of think into
the next quarter, you know, canbe so yeah.
So I think there's just so muchvalue.
In these data sets, in thesehigh frequency, you know, based
off of, you know, large amountsof data, you know, and rolling
that up, and I know for kind ofthe even more technical users,
(16:12):
you know, we also have all thisdata shared on GitHub, you know,
so we've got the job postingsindex, the Indeed wage tracker,
and all of that, so, you know,if there's anybody who is doing
that type of modeling youmentioned and wants to pull it
into whatever coding language orenvironment they're working
with, you know, we make thateasy to do as well with our
GitHub.
William Tincup (16:30):
I love that.
So now or in the future, do yousee a position for generative
AI, uh, to help folks, uh, withprompts in terms of you've given
them keys to the kingdom, theaggregate data that, uh, that is
just wonderful, but they stillmight not know.
I mean, I'm assuming youactually, you know, have a
(16:51):
degree, maybe have studied thisstuff for a while, you know,
might be pretty good at it, youknow, this, that, and the other,
but also, The averagepractitioner might not be as
skilled at like, okay, they haveaccess to this data, but then
what questions should they beasking or what pro I'm thinking
of prompts in terms ofgenerative AI, it's like, what
(17:12):
prompts should I be asking thisto get, to tease out The things
that I should really know.
Cory Stahle (17:19):
Yeah, no, those are some great questions, I think.
Unfortunately, I do have thedegrees and I do have the
skills, you know, to do theeconomics piece.
I think, you know, we're stillin an era where I don't know
that I'd feel comfortableaccounting myself as an expert
in writing prompts, you know, ingenerative AI.
Um, but I think that is one ofthe nice things about what we do
with the hiringlab.
(17:40):
org slash data.
I mean, it's really as simple askind of just changing the drop
down boxes, you know, selecting,you know, the sectors you want.
And I think that that type ofkind of democratization of data,
making it easy so that, yeah,like, I mean, AI can do some
great things, you know, it canreally help with some of the
coding and all that.
But I think also just makingsure that, like, we're telling a
(18:01):
story.
With these data visualizationsand making it so easy that, you
know, when somebody clicks on tothe site, they say, oh, I see
clearly what this data isrepresenting.
You know, that's reallysomething that we try to aim for
anytime we publish a graph,anytime we publish any of that
stuff.
So, um, I think, I think, youknow, generative AI is going to
be a part of potentially helpingpeople to Right.
(18:22):
Pull out some different trends,um, but right now, you know,
we're doing our best to, youknow, make those trends as clear
as possible, um, in a way thatyou can kind of compare and
really say, you know, if you'reworking in healthcare and you
want to see how healthcare isstacking up against retail in
terms of hiring, you know, youjust drop down that box, uh, you
know, in that, uh, Kind of inthat dashboard and you can see,
(18:44):
hey, you know, healthcare ornursing, whatever part of
healthcare you're looking at ismaybe a little stronger than
retail right now and you canmake those comparisons kind of
in the blink of an eye withouthaving to kind of go through
that iterative process with agenerative AI tool.
Right,
William Tincup (18:58):
right, right.
When people start a kind of aresearch product, uh, project,
I'm always fascinated with likethe thesis or, or kind of the,
like what we're, what's, whatare we starting with?
What do we think will happen?
And then at the end of thereporting process and the data's
back there going through it, uh,things that surprised them or
things that they kind ofinvalidated, like, eh, is that
(19:20):
Trex?
Um, was there anything like thatthere for you?
Like when you started this, I,when you started with this idea,
did you already have kind of anidea of how it would play out?
And if so, or if not, is thereanything that kind of, I
wouldn't say shocked, but wasthere anything that just kind of
came out of left field for you?
Cory Stahle (19:40):
Yeah, I think, you know, we've talked a little bit
briefly about kind of the fivetrends to watch, you know, we've
talked a lot about kind of themacroeconomic trends, but kind
of carrying forward off yourlast question with generative
AI, I think it's been a reallyinteresting thing.
So in addition to this trendsreport, we've also done some
recent work on generative AI andthe Potential exposure and
(20:02):
impact we see that having in thelabor market, um, you know, and
so what we kind of did is Ithink we approached it very,
very differently.
You know, if you want to getinto some, uh, some theory and
methodology, you know, I'm, I'mwilling to talk all day about
that stuff because, uh, graphsmethodology, you know, those are
all my, you know, happy nerdwords.
Uh, but, you know, as we look atgenerative AI and as we started
(20:23):
kind of look, um, Understandingand saying, okay, how can we
measure the impact potentiallyof some of these tools?
What we realized was that a lotof the existing methodologies
out there were really focused onsaying you have a job that's
called a janitor.
You know, janitors use brooms,janitor of AI can't You know,
use a broom.
(20:44):
So, you know, therefore,generative AI, you know, is not
going to take that job or pieceof that job.
What we did, though, that wasdifferent in our research, is we
looked at the skills in everysingle job, you know, everything
from the janitor to the computerprogrammer.
And what we found was really,really interesting.
So this gets to your point.
I know this is kind of a longbuild up to your, to your
(21:05):
question here as far assomething that was surprising,
but what I thought was reallysurprising was kind of two
things.
First, we saw that as weanalyzed every skill across 55
million job postings, what wesaw was that every job has at
least some skills that chatGPTor generative AI is potentially
(21:25):
able to do, you know, to somedegree or another.
Um, so I thought that wasreally, really interesting.
You know, you think about how,you know, pretty much every job
now we use email, you know,email is an area where
generative AI, you know, isreally poised to potentially,
you know, revolutionize andtransform the way we're doing
that.
And so it's kind of interestingto think that, you know, it's
not necessarily a question ofcan an entire job be.
(21:48):
You know, change or transform,but really like, what are the
skills within an individual job,you know, and what we found, you
know, from this research againwas that, you know, while every
single job was exposed to somedegree or another, some were
obviously exposed more thanothers.
ChatGPT, you know, thesegenerative AI tools are better
at writing code than they areat, you know, doing, you know,
(22:09):
care, you know, personal caretype of home health tasks or
something.
Um, but I think the other thingthen, and this is kind of the
second thing I think that'sreally surprising, especially
given everything we've heardabout generative AI this year,
is as we looked across jobpostings, what we found was
that, It was about 0.
05 ish percent of job postingswere actually asking for
(22:34):
generative AI type oftechnologies just even a few
weeks ago.
You know, so I think it'sinteresting that, you know, a
lot of the coverage has been,hey, you know, generative AI is
coming, you know, it's this hugething, it's transforming
everything already.
And, and certainly like it'sgrown pretty fast this year, but
it is still like a super, supersmall fraction of the number of
jobs that we're seeing.
(22:54):
Like employers are still not, byand large, asking for people.
With these types of skills, andwe also haven't necessarily seen
a massive drop off postings of,you know, ChatGPT taking jobs or
anything away that oftentimes,you know, the media fears.
So that to me, you know, is kindof surprising, but also maybe
I'm just a little bit of anoptimist because I think that
(23:16):
these technologies are so early.
on in the game that we canreally shape them into what we
want them to be and help them tobe things that, you know,
enhance our productivity, youknow, and aren't things that,
you know, really hurt us, but weneed to be aware of, you know,
what their capabilities are tobe able to do that.
It's,
William Tincup (23:32):
it's really, it's really interesting because
it reminds me of the beginningof the internet.
So there was this whole pre, uh,I'd say pre monetization of the
internet, uh, that was going onwith message boards and whatnot.
And then there was a, there wasa time in which the internet
then basically, uh, kind of tookover.
(23:52):
And it was the buildup to thatthat's fascinating because it
feels a lot like generative AIin that people just had no idea.
They knew that it was coming.
They knew to impact, in thiscase, their life or their
business or whatever, but theydidn't know how, and I feel a
lot of the times when I talk topeople about generative AI,
they're like, I know I shouldpay attention to it, I'm
(24:17):
tinkering with it, Um, it'sinteresting, but I still don't
know if it gets to your point.
It's like, I still don't know.
Like, is that a job or is that askill within another job?
And, uh, and where do thosethings lie?
And does it, does that just kindof play out over time?
Um, I did want to ask you aquestion, uh, that's a little
bit similar, but.
(24:39):
But, but, but, uh, butdifferent.
Have, have, have y'all seen anuptick in generative eye either
on the candidates side of thingsin terms of creating
personalized resumes to fitjobs?
Or is that something you couldsee?
Or generative eye in terms ofjob postings?
(25:00):
Is that, well, you know, for me,on the outside looking at,
again, anecdotes, your data.
Okay, so we're different.
Um, it seems like if I were acandidate.
I would, I could apply to morejobs and do it more personally
using the keywords that are inthe job description, job
posting, and then mirror themwith the, my resume.
(25:24):
Um, it seems like I could dothat as a candidate.
I don't know if that's true.
The other part of that, if Iwere on the corporate side or
the hiring side, it seems like Icould create a much better job
posting.
Now I don't, I don't know ifeither of those are true or if
you have any insight into eitherof those.
Cory Stahle (25:45):
Yeah, no, I think those are both great questions.
I know on the, the research sideof things and what we've done
within my team of economists, Iknow we've focused a lot to this
point, you know, really on thejob posting side of things, you
know, really looking to see kindof what employers are, what
their reaction is, you know, ouremployers are adopting and
that's where that, you know, 0.
05 ish percent number comesfrom.
(26:06):
Um, you know, so we look atthat, but I, but I do think that
you also, you know, you bring upa good point that there are a
lot of other things beyond just.
Having it mentioned as a skilland other areas, you know, where
it could potentially have animpact.
And I know that, you know,indeed, um, you know, is doing a
lot, you know, we've usedartificial intelligence tools
now for years as part of kind ofour matching and trying to, you
(26:28):
know, align people together.
I can't necessarily, I'm not theperson to speak to the specifics
exactly of what we're doing withall that.
But I know that, um, I've seen alot of things that we're doing
to you.
Make the job posting better tomake it easier for people who
are posting jobs, becausecertainly, you know, generative
AI has a lot of promise for, youknow, as you talked about, you
know, creating better alignedresumes, creating job postings
(26:51):
easier, and maybe job postingsthat, you know, better attract
the type of worker that, uh, theemployer actually is trying to
attract, you know, so I thinkthat those are going to be
really interesting things Toreally watch over the coming
years, I mean, as you kind ofsaid, you know, nobody really
knows like we're so early inthis, but I think you've pointed
out two things that are reallygoing to be interesting, um, as
(27:12):
we kind of move forward withthis is to see how people adopt
this and people, how people usethese types of technologies.
So
William Tincup (27:19):
last question is what's the, what's the next
thing you're researching?
What do you get a littledramatic foreshadowing?
What are you, what are youlooking
Cory Stahle (27:26):
into next?
Yeah.
So, I mean, so again, we have ateam of economists kind of all
over the world.
Um, you know, so we haveeconomists and, you know,
Germany and England and, um, youknow, just kind of everywhere,
even in Japan.
And so we are constantlyresearching a bunch of different
topics.
Uh, for me in the U S though, Iwould say I can speak probably
(27:46):
personally to what I'm workingon right now.
I think really looking at someof the kind of skills.
pieces of job postings.
You know what we've done withthe generative AI piece and
really unpacking the skills issomething that we had kind of
done a little bit and we've kindof looked at in the past.
Um, but it's something that Ithink we still feel like there's
(28:07):
more to look at and to reallyhelp people to understand, you
know, what skills look like overtime, you know, and especially,
um, how.
The strength of skills haschanged over time, you know,
just because an employer puts askill in a job posting doesn't
necessarily mean that it's, youknow, a must have skill.
And, you know, this is like, Ineed to have this, right?
(28:28):
And so like, trying to unpack,you know, it's like, what are
kind of the mandatory skills,you know, versus kind of the
nice to have skills, you know,so we're kind of looking into
that, also looking into likeeducational requirements, you
know, trying to understand howthose are changing over time.
We've, we've seen a lot Uh,discussion around states and
other companies, you know,dropping the requirements for
(28:48):
education, um, in their jobpostings and for their jobs.
And that's something that we'relooking to see, um, if that
bears out in the Indeed data aswell, so that we can kind of
track how those are changing.
Because I think, ultimately, allof these are good.
Measures for, you know, kind ofthe recruiting intensity in the
labor market and where thingsare and where, you know, HR and
(29:08):
talent attraction is right now.
And so I think those are just afew of the things that are kind
of at the top of my mind.
Yeah, it's,
William Tincup (29:14):
it's interesting because it's opening the funnel.
You know, if you ban the box,uh, with, with felons, you're
going to have 70 million peoplethat, that, okay, so you've,
you've, you've opened the funnelthere.
If you've, again, take degreesaway, you're opening the funnel
there.
Uh, and if you add compensation,To those things, you're going to
(29:36):
get a higher uptick of peoplebecause it's not a, you know,
guessing game in terms of whatthe, what the job pays.
So it's like opening up thosefunnels.
What are all the differentlevers we can pull to open the
funnel so that you get morecandid flow and then you can
qualify them as you go throughthe process.
So it indeed does a wonderfuljob of being able and letting
(29:57):
their customers have the abilityto have some qualifying
questions.
Being able to push peoplethrough, et cetera, but, uh,
this has been fantastic, Corey.
Thank you so much for your time.
Cory Stahle (30:09):
Yeah.
Thanks for having me.
Anytime I can get on and talkabout methodology and graphs and
stuff, you know, I'm going to behappy.
So.
William Tincup (30:16):
And for the audience, it's a hiringlab.
org.
Please go there because there'sjust some great information
there.
And uh, thanks again forlistening until next time.
This is William Tincup, and you're listening to
the Recruiting Daily podcast.
Today, we have Corey on fromIndeed, and we'll be talking
about Indeed's 2024 U.
S.
Jobs and Hiring Trends Report.
So, uh, this is going to bereally, really fun.
Can't wait.
And so let's just jump into someintroductions.
Uh, Corey, would you do us afavor and introduce yourself and
(00:55):
what
Cory Stahle (00:55):
you do at Indeed?
Yeah.
Thanks, William.
It's great to be here.
So, just by way of introduction,I'm sure a lot of listeners have
maybe heard of Indeed and kindof the job searching side of
things.
If they haven't,
William Tincup (01:08):
Corey, at this point, I really feel sorry.
Cory Stahle (01:11):
You know what I'm saying?
I know.
I know.
It's, uh, it's pretty rough outthere.
It's like that progressive.
William Tincup (01:15):
What is that?
You've been living under a rockfor, you know?
Yeah.
But anyway,
Cory Stahle (01:19):
I'm sorry to interrupt.
Uh, yeah, no, no, uh, you know,so yeah, I mean, you talk about
Indeed, you know, a lot ofpeople think, you know, jobs,
job postings, and certainlythat's a big part of what we do,
you know, and kind of coming upwith the, you know, those kind
of connecting kind of thishiring between employers and job
seekers, uh, but we also have alot of other functions.
And I get to fortunately sit inone of those kind of little bit
(01:42):
of a different kind of funthings that we do at Indeed,
which is I work with what'scalled the hiring lab.
Which is a group and team ofeconomists that uses the data
that we have on Indeed to reallykind of track what's going on in
the labor market and in theeconomy kind of broadly.
And so that's kind of as we talkabout, um, stuff today, you
know, that's kind of wherethat's coming from is we're
(02:03):
looking at kind of what thetrends have been on Indeed and
in hiring in recent months, inrecent years, and then looking
at kind of where those trendsmight lead us going forward.
So
William Tincup (02:13):
ADP has, um, and they've done this for years is
because they have payroll.
So they can kind of tell people,uh, with the data that they're
sitting on kind of what's goingon with payroll, what's going on
with, you know, those types ofthings.
And y'all are sitting on amassive amount of doc, of data
as well in terms of who'shiring, uh, who's posting job
(02:34):
ads, et cetera.
Could you, could you tell us alittle bit more about the data
that y'all sit on?
Cory Stahle (02:40):
Yeah, so I mean, yeah, great, that's, uh, so we
got a lot of different data.
I would say that with theeconomic data that we're looking
at, a lot of what we're lookingat is job postings, you know, so
this isn't necessarilyproprietary data in terms of,
you know, these are job postingsthat job seekers can go on to
look at, Bye bye.
Bye bye.
But if you want to kind of get agood feel for what's going on,
(03:03):
on the larger picture, it wouldbe pretty difficult to comb
through millions and millions ofthese different job postings and
to break them out, you know,into, you know, this sector is
hiring more, this sector ishiring less.
That's something that would bevery, very hard to do as a job
seeker or an employer orsomebody working in recruiting
or HR.
So that's what we really do iswe take.
That data we have, we take thedata, you know, the job postings
(03:25):
that employers are posting, andfrom that we can tell a lot
about, you know, our sector'sposting more or less than they
used to.
You know, we can also look at,within a job posting, we can
see, you know, how much areemployers asking for.
Generative AI type of tools, youknow, how much are employers
advertising pay in a job postingand is that pay going up or
(03:45):
down?
So there are a lot of economicindicators that we can actually
draw out of job postings, youknow, as we start to kind of
roll up and look at those at ahigher
William Tincup (03:54):
level.
I love that because you can givepeople, you could paint a
picture about skills and comp ina way that.
Other folks that don't have thatdata set.
Um, can you, uh, can you, canyou kind of talk to about
traffic?
Like again, there's the postingitself and then there's
obviously people that then gointo the posting and do
something.
(04:14):
Is there, is there data therethat's meaningful or even
candidate flow that kind of gofurther into people that have
applied, etc.
Is that useful to the, uh, tothe audience?
Cory Stahle (04:26):
Yeah, so that is something that we've done a
little bit of work and, uh, justclarify, I think if I understand
correctly, you're asking aboutkind of people who click on job
postings, you know, kind of theflow of job seekers.
So yeah, so that is somethingthat we've looked at
historically.
I know one example of thatrecently is we looked at The
share of clicks on job postings,you know, so we're looking at
(04:47):
people who are interested in jobpostings in the United States.
In this case, we were lookingand we wanted to see how many
people who are clicking on aspecific job posting were living
overseas, you know, so we cankind of track foreign interest
and kind of get an idea.
You know, we're not necessarilygetting into the U.
S.
You know, really individualized,personalized data, like we take
(05:07):
protecting personal informationreally important, you know,
really seriously, um, but again,we kind of roll up, we look at
the aggregate level and, youknow, we can then have trends
like we see that, you know,during the pandemic, foreign
interest, you know, in U.
S.
jobs, you know, was pretty flat,but we've seen that actually
grow pretty significantly in thelast couple years, and I think
that's been a A prettysignificant part of the story
(05:30):
behind the resilience in thelabor market, you know, so we
can really take these things andwe kind of tie them together and
get a picture, you know, and onthe job posting side, we have
public data sets that wepublish, one's called the Indeed
Job Postings Index, so that'swhat allows us to kind of track
how job postings are moving,anybody can access that.
Um, on our website, which ishiringlab.
(05:51):
org, um, and then we also havewhat's called the Indeed Wage
Tracker, which I alluded toearlier, which gives a
measurement of kind of how wagesand comp and job postings are
changing over time.
Um, so again, just helps us tocreate more of a full picture
than I think governmentstatistics, uh, Right.
You know, where, where they kindof maybe, maybe can lack
(06:11):
sometimes, especially in termsof frequency.
So.
William Tincup (06:14):
I love it.
I love it.
First of all, I love that you'redoing something with the data to
help both the larger audience,but also the folks that, that
are indeed customers as well,like giving them some insight
into all of those things, skillsand wages and everything.
It's like.
Because left to their own, they,they might not know that these
macro issues are at play.
(06:35):
So I, first of all, I just lovethat.
Um, let's talk a little bitabout the report.
Uh, the, the 24 US jobs andhiring, uh, trends report.
What, um, if we were to rollthis up for folks with short
attention spans, which would bepretty much everybody, um, what
are some of the things thatyou'd like to highlight?
Cory Stahle (06:56):
Yeah, so I think, you know, kind of the, uh, quick
rundown of it, I think, settingthe stage for this report, I
think is important to understandas we were kind of preparing
this report, we were looking atwhat was going on, kind of in
this macro economy, this currenteconomic environment, and kind
of what's happened in 2023, andwhat's been really interesting
is, you know, we've seen a lotof different discussions in
(07:18):
2023, you know, we've heard, youknow, a lot of discussions
around return for rural.
return to office, generative AI,you know, how that's going to
maybe change the labor market.
You know, so there's been a lotof things to talk about that
have happened in 2023, but Ithink it's important to note
what did not happen in 2023, orat least, you know, this far.
We've still got, you know, amonth left here, but, uh, but
(07:39):
what we haven't seen based onthe data we're looking at is a
recession.
You know, we saw so many people,you know, making recession calls
at the end of last year.
And as we, you know, have gonethrough the year, the labor
market and by and large thewhole economy has been pretty
resilient.
And so really that's kind of thesetting the stage for this
report.
So the report basically issaying, okay, what are the
(08:01):
trends that we saw in 2023 thatneed to carry forward into 2024?
You know, what is it that, youknow, employers need to do in
terms of their demand?
You know, what do we need to doin terms of attracting people,
you know, into the labor market?
You know, what does.
Quitch rates and like thedynamics of the labor market
look like in a healthy labormarket.
(08:22):
You know, and then also wetalked a little bit about wage
growth.
And I think also generative AI,you know, is another trend that
we'll really be watching.
So that's kind of the stage.
So I would say in short, uh, youknow, here's kind of the, the
wrap up really, really shortversion.
You know, what is it that needsto go right in 2024?
We need to continue to see, youknow, strong demand from
employers, more people enteringinto the labor market.
(08:45):
And we need to, you know, keepour eye on some of these
emerging things like generativeAI.
William Tincup (08:50):
So, um, and I'm just probably done in the, in
the report, but did you, are webifurcating between hourly and
corporate or salaried?
Uh, in any way, because I, youknow, what I hear usually from
practitioners is more, uh, onthe hourly side, it's, it is,
it's, it's very difficult andseems to be getting more
(09:11):
difficult, but on the salariedside, depending on the job or
wherever, uh, not as hard.
There's like a, and again,that's just, you're sitting on
data.
I'm sitting on anecdotes.
So.
I'm going to trust you more thanthey trust me, but basically,
basically it's kind of like thestory of like, okay, if you're
hiring for salary, you're goingto have some options.
(09:31):
Got it.
If you're hiring for hourly,you've got to re your, your
options are limited and you'vegot to raise the pay.
Cory Stahle (09:39):
Yeah.
Yeah.
And I think ultimately, youknow, as we look across indeed,
about 50%, a little over 50percent of jobs now include a
salary, you know, partially as aresult of different kind of pay
transparency regulations in somestates.
Um, but that means that thereare almost 50 percent that still
do not contain a salary.
So we don't necessarily havelike the breakout, um, in those
(10:00):
terms, but what we do look at iswe kind of look at, and I think
this is another interestingproxy.
We look at.
The breakout between jobs thatare more likely to be done
remote versus the ones that areless likely to be be done remote
and I would say there'sprobably, you know, some
comparison there to be madelower remote jobs tend to be a
little more skilled labor,probably a little more, you
(10:22):
know, in person hourly typejobs, whereas some of those
highly remote jobs tend to bemore kind of knowledge worker
type positions, um, andpotentially more salary jobs.
And I think as we kind of Talkabout that.
We have seen, you know, reallythe bifurcation this year in the
labor market, you know, wherecoming out of the pandemic, you
(10:42):
know, all sectors rose together.
They all sunk together duringthe pandemic and they all rose
together.
But what we've seen in 2023 iskind of this, this kind of fork
in the road to some degree.
You know, we've seen softwaredevelopment jobs drop below
their pre pandemic levels,whereas we've seen, you know,
jobs in areas like nursing andconstruction and manufacturing,
(11:04):
you know, keep, you know, prettyhigh levels of employer demand,
you know, which ultimately Ithink relates back to what
you're talking about in terms ofthe difficulty to find workers,
um, as, you know, there's just,you know, stronger demand that
we're seeing in some of thosetype of low remote in person
type of opportunities.
So
William Tincup (11:21):
did y'all, uh, just as a matter of a point of
honor, is did y'all?
Do this report last year at
Cory Stahle (11:27):
the year?
We did actually.
Yeah.
So we did produce kind of fivetrends to watch last year.
Um, so that is something thatpeople can kind of go back and
look.
I would say last year we kind offocused on a long term outlook.
Um, kind of the takeaway was.
You know, demographics arereally weighing on the United
States labor market andpopulation in general.
(11:51):
You know, we have this agingpopulation, which over the
course of decades, unlesssomething changes, you know,
with the birth rate, we're goingto see labor markets continue to
be pretty tight over the comingdecades.
You know, business cycle mightgo up and down, but if you have
fewer workers to take jobs,that's going to create different
challenges in the future.
So we took that long termperspective, but for this year,
(12:12):
we kind of wanted to focus.
Uh, mainly on 2024, so, so yeah,so we have kind of this kind of
long look last year versus theshort look this year, which I
think in some ways they kind ofmake a good pairing of reading
together.
Oh, a
William Tincup (12:24):
hundred percent.
I always love it when people goback and look at like their
predictions.
And, and analyze theirprediction, they're very
critical, analyze theirpredictions like, yeah, wildly
missed.
Uh, yep.
Got that one.
Right.
Got that one.
Right.
This one's kind of, we're still,you know, we're still almost
there.
Um, so, so first of all, where'sthe, uh, where's the report
(12:45):
located?
It's obviously on the
Cory Stahle (12:47):
website.
Yeah, yeah, so people can go tohiringlab.
org.
Um, and so we've got not onlythis report, um, but also like
the data sets I've mentioned.
So like hiringlab.
org slash data, um, is a, a dataportal where people can kind of
do their own analysis, so tospeak, and you can kind of sort.
And I love getting in there andjust playing with it because you
(13:08):
can look how job postings aredifferent between different
states.
You can look how they'redifferent between different like
metropolitan areas.
You can look how job postingsvary by different sectors, you
know, we've talked a lot aboutsectors, you know, and so it can
be really helpful to go in thereand to, you know, kind of slice
up the data that you want.
Um, but yeah, as far as thesereports, they're all just up on
hiringlab.
org along with a lot of otherresearch that we publish using
(13:31):
this data pretty regularly.
Oh, I
William Tincup (13:33):
love that.
And I love, I especially lovethe fact that they can, um, use
your, use your aggregate data tothen model and think about
themselves.
Cause that's, you know, that's,that's for them, especially, uh,
CH, uh, CHROs and VPs of talent.
They're trying to figure outforecasts.
They're trying to figure outlike what is going on.
(13:55):
We know what's going on withinthe four walls of our company,
but we have no idea what's goingon the four walls of the United
States.
And so we might learn somethings at an industry event or
something like that, but toactually play, I say play, to
actually use aggregate data inthis way, it's just, first of
all, Thanks, because I don'tknow of anybody else that does
(14:15):
that, that allows people to, tolook and kind of do their own
research, uh, based on such alarge data set.
Cory Stahle (14:24):
Yeah, no, I love it.
I think it's absolutely amazing,and I would say from a past
life, I used to work as aneconomist in state government,
um, and a big part of my jobWith that was to go around to
different businesses and HR, youknow, and heads of talent and
to, you know, kind of talk aboutthe tools that the government
had available, you know, andthings that were out there, um,
(14:45):
for businesses to use and thegovernment has some really,
really good, you know, reallygreat tools as you look at it,
but then Now coming into myposition within Indeed, I'm
saying, you know, there's a lotof value in the government
tools, but it's really hard tobeat the scale and the size of
what we're doing at Indeedbecause, you know, you talk
about, you know, the governmentdata, you know, we, there's been
(15:07):
a lot of discussion lately aboutgovernment data surveys, having
problems with response rates,and the result has been a little
bit of noisiness in the series,and then you compare that to the
Indeed data where we're talkingabout basically, you know, Any
job posting that's on theinternet, it's probably being
included in this analysis.
You know, we're really pullingtogether a near universe of
(15:28):
jobs, so you can get like areally good picture.
You don't have to worry aboutsurveys.
And not only that, but I mean, Ican look at what the data was
doing, you know, last Friday,rather than waiting, you know,
three or four months.
Because I know one of thebiggest employment data sets,
you know, in from the governmentis the QCEW data.
And that's kind of the bestcensus overall universe of
(15:51):
employment.
But that tends to have like asix month delay, you know, in
six months when you're trying toforecast and kind of think into
the next quarter, you know, canbe so yeah.
So I think there's just so muchvalue.
In these data sets, in thesehigh frequency, you know, based
off of, you know, large amountsof data, you know, and rolling
that up, and I know for kind ofthe even more technical users,
(16:12):
you know, we also have all thisdata shared on GitHub, you know,
so we've got the job postingsindex, the Indeed wage tracker,
and all of that, so, you know,if there's anybody who is doing
that type of modeling youmentioned and wants to pull it
into whatever coding language orenvironment they're working
with, you know, we make thateasy to do as well with our
GitHub.
William Tincup (16:30):
I love that.
So now or in the future, do yousee a position for generative
AI, uh, to help folks, uh, withprompts in terms of you've given
them keys to the kingdom, theaggregate data that, uh, that is
just wonderful, but they stillmight not know.
I mean, I'm assuming youactually, you know, have a
(16:51):
degree, maybe have studied thisstuff for a while, you know,
might be pretty good at it, youknow, this, that, and the other,
but also, The averagepractitioner might not be as
skilled at like, okay, they haveaccess to this data, but then
what questions should they beasking or what pro I'm thinking
of prompts in terms ofgenerative AI, it's like, what
(17:12):
prompts should I be asking thisto get, to tease out The things
that I should really know.
Cory Stahle (17:19):
Yeah, no, those are some great questions, I think.
Unfortunately, I do have thedegrees and I do have the
skills, you know, to do theeconomics piece.
I think, you know, we're stillin an era where I don't know
that I'd feel comfortableaccounting myself as an expert
in writing prompts, you know, ingenerative AI.
Um, but I think that is one ofthe nice things about what we do
with the hiringlab.
(17:40):
org slash data.
I mean, it's really as simple askind of just changing the drop
down boxes, you know, selecting,you know, the sectors you want.
And I think that that type ofkind of democratization of data,
making it easy so that, yeah,like, I mean, AI can do some
great things, you know, it canreally help with some of the
coding and all that.
But I think also just makingsure that, like, we're telling a
(18:01):
story.
With these data visualizationsand making it so easy that, you
know, when somebody clicks on tothe site, they say, oh, I see
clearly what this data isrepresenting.
You know, that's reallysomething that we try to aim for
anytime we publish a graph,anytime we publish any of that
stuff.
So, um, I think, I think, youknow, generative AI is going to
be a part of potentially helpingpeople to Right.
(18:22):
Pull out some different trends,um, but right now, you know,
we're doing our best to, youknow, make those trends as clear
as possible, um, in a way thatyou can kind of compare and
really say, you know, if you'reworking in healthcare and you
want to see how healthcare isstacking up against retail in
terms of hiring, you know, youjust drop down that box, uh, you
know, in that, uh, Kind of inthat dashboard and you can see,
(18:44):
hey, you know, healthcare ornursing, whatever part of
healthcare you're looking at ismaybe a little stronger than
retail right now and you canmake those comparisons kind of
in the blink of an eye withouthaving to kind of go through
that iterative process with agenerative AI tool.
Right,
William Tincup (18:58):
right, right.
When people start a kind of aresearch product, uh, project,
I'm always fascinated with likethe thesis or, or kind of the,
like what we're, what's, whatare we starting with?
What do we think will happen?
And then at the end of thereporting process and the data's
back there going through it, uh,things that surprised them or
things that they kind ofinvalidated, like, eh, is that
(19:20):
Trex?
Um, was there anything like thatthere for you?
Like when you started this, I,when you started with this idea,
did you already have kind of anidea of how it would play out?
And if so, or if not, is thereanything that kind of, I
wouldn't say shocked, but wasthere anything that just kind of
came out of left field for you?
Cory Stahle (19:40):
Yeah, I think, you know, we've talked a little bit
briefly about kind of the fivetrends to watch, you know, we've
talked a lot about kind of themacroeconomic trends, but kind
of carrying forward off yourlast question with generative
AI, I think it's been a reallyinteresting thing.
So in addition to this trendsreport, we've also done some
recent work on generative AI andthe Potential exposure and
(20:02):
impact we see that having in thelabor market, um, you know, and
so what we kind of did is Ithink we approached it very,
very differently.
You know, if you want to getinto some, uh, some theory and
methodology, you know, I'm, I'mwilling to talk all day about
that stuff because, uh, graphsmethodology, you know, those are
all my, you know, happy nerdwords.
Uh, but, you know, as we look atgenerative AI and as we started
(20:23):
kind of look, um, Understandingand saying, okay, how can we
measure the impact potentiallyof some of these tools?
What we realized was that a lotof the existing methodologies
out there were really focused onsaying you have a job that's
called a janitor.
You know, janitors use brooms,janitor of AI can't You know,
use a broom.
(20:44):
So, you know, therefore,generative AI, you know, is not
going to take that job or pieceof that job.
What we did, though, that wasdifferent in our research, is we
looked at the skills in everysingle job, you know, everything
from the janitor to the computerprogrammer.
And what we found was really,really interesting.
So this gets to your point.
I know this is kind of a longbuild up to your, to your
(21:05):
question here as far assomething that was surprising,
but what I thought was reallysurprising was kind of two
things.
First, we saw that as weanalyzed every skill across 55
million job postings, what wesaw was that every job has at
least some skills that chatGPTor generative AI is potentially
(21:25):
able to do, you know, to somedegree or another.
Um, so I thought that wasreally, really interesting.
You know, you think about how,you know, pretty much every job
now we use email, you know,email is an area where
generative AI, you know, isreally poised to potentially,
you know, revolutionize andtransform the way we're doing
that.
And so it's kind of interestingto think that, you know, it's
not necessarily a question ofcan an entire job be.
(21:48):
You know, change or transform,but really like, what are the
skills within an individual job,you know, and what we found, you
know, from this research againwas that, you know, while every
single job was exposed to somedegree or another, some were
obviously exposed more thanothers.
ChatGPT, you know, thesegenerative AI tools are better
at writing code than they areat, you know, doing, you know,
(22:09):
care, you know, personal caretype of home health tasks or
something.
Um, but I think the other thingthen, and this is kind of the
second thing I think that'sreally surprising, especially
given everything we've heardabout generative AI this year,
is as we looked across jobpostings, what we found was
that, It was about 0.
05 ish percent of job postingswere actually asking for
(22:34):
generative AI type oftechnologies just even a few
weeks ago.
You know, so I think it'sinteresting that, you know, a
lot of the coverage has been,hey, you know, generative AI is
coming, you know, it's this hugething, it's transforming
everything already.
And, and certainly like it'sgrown pretty fast this year, but
it is still like a super, supersmall fraction of the number of
jobs that we're seeing.
(22:54):
Like employers are still not, byand large, asking for people.
With these types of skills, andwe also haven't necessarily seen
a massive drop off postings of,you know, ChatGPT taking jobs or
anything away that oftentimes,you know, the media fears.
So that to me, you know, is kindof surprising, but also maybe
I'm just a little bit of anoptimist because I think that
(23:16):
these technologies are so early.
on in the game that we canreally shape them into what we
want them to be and help them tobe things that, you know,
enhance our productivity, youknow, and aren't things that,
you know, really hurt us, but weneed to be aware of, you know,
what their capabilities are tobe able to do that.
It's,
William Tincup (23:32):
it's really, it's really interesting because
it reminds me of the beginningof the internet.
So there was this whole pre, uh,I'd say pre monetization of the
internet, uh, that was going onwith message boards and whatnot.
And then there was a, there wasa time in which the internet
then basically, uh, kind of tookover.
(23:52):
And it was the buildup to thatthat's fascinating because it
feels a lot like generative AIin that people just had no idea.
They knew that it was coming.
They knew to impact, in thiscase, their life or their
business or whatever, but theydidn't know how, and I feel a
lot of the times when I talk topeople about generative AI,
they're like, I know I shouldpay attention to it, I'm
(24:17):
tinkering with it, Um, it'sinteresting, but I still don't
know if it gets to your point.
It's like, I still don't know.
Like, is that a job or is that askill within another job?
And, uh, and where do thosethings lie?
And does it, does that just kindof play out over time?
Um, I did want to ask you aquestion, uh, that's a little
bit similar, but.
(24:39):
But, but, but, uh, butdifferent.
Have, have, have y'all seen anuptick in generative eye either
on the candidates side of thingsin terms of creating
personalized resumes to fitjobs?
Or is that something you couldsee?
Or generative eye in terms ofjob postings?
(25:00):
Is that, well, you know, for me,on the outside looking at,
again, anecdotes, your data.
Okay, so we're different.
Um, it seems like if I were acandidate.
I would, I could apply to morejobs and do it more personally
using the keywords that are inthe job description, job
posting, and then mirror themwith the, my resume.
(25:24):
Um, it seems like I could dothat as a candidate.
I don't know if that's true.
The other part of that, if Iwere on the corporate side or
the hiring side, it seems like Icould create a much better job
posting.
Now I don't, I don't know ifeither of those are true or if
you have any insight into eitherof those.
Cory Stahle (25:45):
Yeah, no, I think those are both great questions.
I know on the, the research sideof things and what we've done
within my team of economists, Iknow we've focused a lot to this
point, you know, really on thejob posting side of things, you
know, really looking to see kindof what employers are, what
their reaction is, you know, ouremployers are adopting and
that's where that, you know, 0.
05 ish percent number comesfrom.
(26:06):
Um, you know, so we look atthat, but I, but I do think that
you also, you know, you bring upa good point that there are a
lot of other things beyond just.
Having it mentioned as a skilland other areas, you know, where
it could potentially have animpact.
And I know that, you know,indeed, um, you know, is doing a
lot, you know, we've usedartificial intelligence tools
now for years as part of kind ofour matching and trying to, you
(26:28):
know, align people together.
I can't necessarily, I'm not theperson to speak to the specifics
exactly of what we're doing withall that.
But I know that, um, I've seen alot of things that we're doing
to you.
Make the job posting better tomake it easier for people who
are posting jobs, becausecertainly, you know, generative
AI has a lot of promise for, youknow, as you talked about, you
know, creating better alignedresumes, creating job postings
(26:51):
easier, and maybe job postingsthat, you know, better attract
the type of worker that, uh, theemployer actually is trying to
attract, you know, so I thinkthat those are going to be
really interesting things Toreally watch over the coming
years, I mean, as you kind ofsaid, you know, nobody really
knows like we're so early inthis, but I think you've pointed
out two things that are reallygoing to be interesting, um, as
(27:12):
we kind of move forward withthis is to see how people adopt
this and people, how people usethese types of technologies.
So
William Tincup (27:19):
last question is what's the, what's the next
thing you're researching?
What do you get a littledramatic foreshadowing?
What are you, what are youlooking
Cory Stahle (27:26):
into next?
Yeah.
So, I mean, so again, we have ateam of economists kind of all
over the world.
Um, you know, so we haveeconomists and, you know,
Germany and England and, um, youknow, just kind of everywhere,
even in Japan.
And so we are constantlyresearching a bunch of different
topics.
Uh, for me in the U S though, Iwould say I can speak probably
(27:46):
personally to what I'm workingon right now.
I think really looking at someof the kind of skills.
pieces of job postings.
You know what we've done withthe generative AI piece and
really unpacking the skills issomething that we had kind of
done a little bit and we've kindof looked at in the past.
Um, but it's something that Ithink we still feel like there's
(28:07):
more to look at and to reallyhelp people to understand, you
know, what skills look like overtime, you know, and especially,
um, how.
The strength of skills haschanged over time, you know,
just because an employer puts askill in a job posting doesn't
necessarily mean that it's, youknow, a must have skill.
And, you know, this is like, Ineed to have this, right?
(28:28):
And so like, trying to unpack,you know, it's like, what are
kind of the mandatory skills,you know, versus kind of the
nice to have skills, you know,so we're kind of looking into
that, also looking into likeeducational requirements, you
know, trying to understand howthose are changing over time.
We've, we've seen a lot Uh,discussion around states and
other companies, you know,dropping the requirements for
(28:48):
education, um, in their jobpostings and for their jobs.
And that's something that we'relooking to see, um, if that
bears out in the Indeed data aswell, so that we can kind of
track how those are changing.
Because I think, ultimately, allof these are good.
Measures for, you know, kind ofthe recruiting intensity in the
labor market and where thingsare and where, you know, HR and
(29:08):
talent attraction is right now.
And so I think those are just afew of the things that are kind
of at the top of my mind.
Yeah, it's,
William Tincup (29:14):
it's interesting because it's opening the funnel.
You know, if you ban the box,uh, with, with felons, you're
going to have 70 million peoplethat, that, okay, so you've,
you've, you've opened the funnelthere.
If you've, again, take degreesaway, you're opening the funnel
there.
Uh, and if you add compensation,To those things, you're going to
(29:36):
get a higher uptick of peoplebecause it's not a, you know,
guessing game in terms of whatthe, what the job pays.
So it's like opening up thosefunnels.
What are all the differentlevers we can pull to open the
funnel so that you get morecandid flow and then you can
qualify them as you go throughthe process.
So it indeed does a wonderfuljob of being able and letting
(29:57):
their customers have the abilityto have some qualifying
questions.
Being able to push peoplethrough, et cetera, but, uh,
this has been fantastic, Corey.
Thank you so much for your time.
Cory Stahle (30:09):
Yeah.
Thanks for having me.
Anytime I can get on and talkabout methodology and graphs and
stuff, you know, I'm going to behappy.
So.
William Tincup (30:16):
And for the audience, it's a hiringlab.
org.
Please go there because there'sjust some great information
there.
And uh, thanks again forlistening until next time.
Popular Podcasts
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Therapy Gecko
An unlicensed lizard psychologist travels the universe talking to strangers about absolutely nothing. TO CALL THE GECKO: follow me on https://www.twitch.tv/lyleforever to get a notification for when I am taking calls. I am usually live Mondays, Wednesdays, and Fridays but lately a lot of other times too. I am a gecko.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.