March 27, 2025 • 37 mins
In this episode of The Changing State of Talent Acquisition, we sit down with Craig Fisher—CEO of TalentNet Media, author of Hiring Humans, and one of the most respected voices in the HR tech space. From his early days in staffing to launching TalentNet Live, Craig has worn nearly every hat in the talent world.
We explore why friction in the hiring process is actually a good thing, how AI tools can enhance rather than replace human connection, and why most companies are missing the mark on retention and employee advocacy. Craig also shares behind-the-scenes details on rewriting 13,000 job descriptions for one of the world's largest banks—and why SEO and contextual AI optimization are now table stakes for recruiting.
Plus, he gives us a sneak peek at his upcoming book Paint Your Store, designed to help job seekers and solopreneurs level up their personal brands.
Whether you're a recruiter, TA leader, or job seeker navigating today's AI-infused hiring landscape, this is one conversation you won't want to miss.
🛠️ Topics covered:
- How social media shaped early recruitment marketing
- The case for human friction in hiring
- Why AI won't replace recruiters (but bad job descriptions might)
- Tactical tips for job SEO and optimizing for AI search
- The overlooked power of onboarding and employee advocacy
🔗 Learn more about Craig's work: hiring-humans.com New episodes every week at changestate.io
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Welcome to the Changing State of Talent
Acquisition, where your hosts,graham Thornton and Martin Kred,
share their unfiltered takes onwhat's happening in the world
of talent acquisition today.
Each week brings new guests whoshare their stories on the
tools, trends and technologiescurrently impacting the changing
state of talent acquisition.
Have feedback or want to jointhe show?
(00:21):
Head on over to changestateio.
And now on to this week'sepisode.
Speaker 2 (00:27):
All right, and we're back with another episode of the
Changing State of TalentAcquisition Podcast, super
excited for our next guest withan incredible background, craig
Fisher, the CEO of TalentNetMedia, author of multiple books,
most recently Hiring Humans,host of TalentNet Live
conferences and just a generalwell-known HR tech advisor in
(00:49):
our space.
So, craig, welcome to the show.
Speaker 3 (00:51):
Graham, thank you.
It is a great pleasure to behere with you today.
Speaker 2 (00:55):
Would love for you to maybe briefly share your
journey, from your early days inrecruitment marketing help our
audience get to know you alittle bit better to your days
starting Talent Media, and maybehow have your experiences
shaped your perspective on theworld of talent acquisition.
Speaker 3 (01:10):
Yeah.
So I came up through staffinglike a lot of people in the 90s.
Before that I was a drug repand a medical sales guy, but I
went through all the levelsrecruiter, account manager,
leader, owner of a staffingcompany starting in 2007.
(01:31):
And all the while I was doingthat, I come from an advertising
and marketing college degreebackground and when I'm doing
all this stuff, I'm attemptingto differentiate myself by
helping both the employers thatI'm working with and the
candidates that I'm working withto market themselves better.
(01:53):
And I developed some strongopinions on it and some nifty
hacks and methodologies and Istarted writing about it in blog
posts and on Twitter.
And in the early days of Twitterin the late 2000s, it was a hot
topic.
Is social media good forrecruiting?
(02:13):
I'm like, yes, it's the bestthing ever, come on.
And so I got a little Twitterfamous for my opinionated
outlook on it.
But also I started the firsthashtag chat for recruiters
called TalentNetLive.
At the time we went withhashtag TNL.
(02:35):
So I own a staffing company, I'mbuilding community.
I'm attempting to push my viewof the world, which is treat
candidates better, explainbetter about yourself as an
employer in a moderated fashion.
I'm saying listen, I think I'vegot a better idea than what's
happening right now.
(02:55):
That's very forward thinkingfor a guy that didn't really
have any platform for it yet.
But my goal was to get on stageand talk about it and get out
of the day-to-day of recruitingwith my staffing agency and
(03:16):
executive search firm.
And people did start invitingme to come speak to their teams
and talk about it.
And then the Twitter chatquickly became a request for IRL
in real life meetings and westarted the first ever Talent
Net Live conference in the fallof 2009.
(03:39):
And we had on our first boardand at our first conference were
notables like William Tencupand Chris Hoyt and Jim Schneider
, bill Borman lots ofinteresting folks from talent
acquisition were there and itreally kicked off the
unconference theme forrecruiting conferences.
(04:01):
So from there I was able topivot into full-time consulting,
doing recruitment, marketingand employer brand, starting in
2011.
So I kind of made TalentNetMedia a thing back then and
since then I've gotten to workon amazing projects with you
(04:22):
know Fortune 50 companies andsome hungry startups and
occasionally go in-house to doleadership jobs at companies
like CA Technologies and AllegisGlobal Solutions.
I was the CMO there for a whileand got to work with big
customers like Amex, hsbc and GM.
So it's been a wonderful wildride and I think to those
(04:47):
starting out now in the field,have an opinion right.
It all starts with having apoint of view on things and
don't be generic and watereddown.
Speaker 4 (04:58):
Well, thanks for sharing, craig.
We're so thrilled to have youon the podcast.
We've got a lot of guests thatjoin us, but I don't think
anyone quite has the depth ofexperience and different
experiences in the industry asyou have, and it's also
interesting that you kind ofwore a lot of the different hats
that exist in this space fromyour early days to now.
So I'd love to kind of start byjust getting kind of a long
(05:20):
view question.
You know you spend a lot oftime you've spent a lot of time
in this space thinkingparticularly about emerging
trends, cultivating communitiesyou just talked about At this
particular moment.
What are the trends or shiftsthat you're particularly
interested in or payingattention to?
You know, obviously AI is a bigtopic for everybody.
(05:40):
It could be about AI, or maybeit's not about AI.
You know what's on your mindthese days.
Speaker 3 (05:44):
Well, it's interesting.
So there's multiple factorsputting pressure on the talent
space right now.
Currently, we don't have lowunemployment.
Well, it's kind of slowlygrowing.
The numbers are a littlemisleading, but what's really
the problem is there's not muchchurn happening within
(06:05):
organizations.
There are layoffs, but peoplearen't voluntarily leaving their
jobs.
They might be quietly looking,but they are afraid of this
economy still and not leavingtheir jobs.
So talent acquisition teamshave been shrinking, partly
because of the economy, partlybecause of AI being able to help
(06:28):
us automate things.
But you know, we've seen thesetrends in the economy happen
every 10 or so years.
You know for multiple seasonsthat this too shall pass and
we're going to see a turnaround,because the trend right now is
to attempt to automateeverything in the candidate
(06:50):
process, and that's not enoughfriction.
So this frictionless idea is alittle overblown.
There has to be some humaninteraction, because if it is
too easy, candidates won'tbelieve it's real, and so when
hiring comes back, there will bea need for recruiters again.
(07:12):
There probably will be a needfor sourcers again.
We will be able to do a lot ofit, though better and easier,
because of the AI tools.
I mean, think of the thingsthat you feed into chat GPT
right now and say, okay, buildme a plan for this and it's so
easy to do.
But you can't just trust itblindly.
(07:32):
Somebody still has to say, yeah, this is good, this is not good
, this is ridiculous, this is ahallucination, and you know,
make it real.
Speaker 4 (07:43):
Sure Well, craig, earlier you were talking about
the importance of having a pointof view and you know I think
you just expressed it probablyan intentionally provocative one
, or at least it sounds like itto me this idea that we need
friction in the process, andnormally, whether you're talking
about customer experience orcandidate experience, we seem to
, as an industry or business,people try to aim for as less
(08:05):
friction as possible.
Could you unpack that a littlebit?
You know what do you mean bythat exactly.
Why is friction a good thing,and how should organizations be
thinking about it?
Speaker 3 (08:15):
Yeah.
So my friend Jim D'Amico and Italk about this and he agrees
with me that if there is nofriction, the process seems like
you're going to get ghosted100% if there's no human
intervention at all, and sothere has to be some human touch
(08:36):
points.
A little bit of friction isimportant.
So I feel that we're going togo over our skis with our
attempt to automate everythingand take recruiters and humans
out of the mix, because itleaves you with a process that
just feels vacant.
Speaker 2 (08:58):
So you know you've written a book recently, you
know, Hiring Humans, and I thinkyou know, really, one of the
points emphasized, you know, inthat book is that hiring is
really a human endeavor and youknow there is some balance
between technology and the humantouch playing out.
So you know, let's dive intoHiring Humans a bit, craig.
(09:19):
So you know what drove you towrite that book and let's the
high-level theme that is mostapplicable in today's market my
philosophy, graham, as you mayknow, is that people want to
work with people and we haveautomation.
Speaker 3 (09:39):
We have all these great tools and I'm a techie guy
myself great tools, and I'm atechie guy myself and I like
nothing more than to be alone,put my phone down and crank away
on a project, write some code,whatever it is right, edit a
video, but at some point I needpeople.
So even the hardcore coder thatin theory works in a basement
(10:07):
is apt to occasionally have aquestion about their benefits or
their quarterly reviews orarrays or something like that.
And you can't automate all ofthose things.
You have to have people in theorganization.
And if there's going to beinternal mobility, if there's
(10:27):
going to be culture of any kind,if there's going to be what I
think is going to be an attemptto keep people in organizations
longer because right now we'rein this sort of gig economy
where things are very temporaryand people don't have any very
good reason to stay at a companylong term.
(10:47):
I think without more of a humantouch in that process also,
we're not going to be verysuccessful.
So, yes, automation is great.
Yes, tools are good, butthey're just tools.
If you remember when the iPhonewas introduced right 2008, that
was going to change everything.
We won't need recruitersanymore.
(11:08):
Same thing when the internethappened.
I mean all these greatinventions are going to do away
with the need for humans, but itturns out no, they're just
tools and they need people touse them.
Speaker 4 (11:20):
Yeah, I think that's a really interesting point,
craig, and it sounds when yousay it out loud, it sounds
almost obvious, but it is asubtle point and I don't think
it's a point that is.
I think it's a point that'slost on a lot of people, whether
it's recruiting or any otherdomain that's going to be
impacted by AI.
There's a lot of fear.
Of course, people are worriedthat it's going to take all the
(11:40):
jobs, but it seems to miss thisfundamental point, that why are
we doing this in the first place?
This is a human creation, firstof all, and human beings really
cannot be reduced to a machine.
Sure, a machine may be able todo a lot of the things that a
human can do, and do them morequickly or better or more
(12:01):
efficiently, but it doesn't havethose intangibles, and that's
what makes life and work andcareers meaningful.
Is that a fair encapsulation ofyour point of view?
Speaker 3 (12:11):
It is, and so there's an REM record called Life's
Rich Pageant and it always singsto me.
In respect to the humanexperience, yeah, we created
these machines for our benefit.
(12:33):
Computers are finally doing whatwe always wanted them or hoped
for them to do, but I read andheard on a podcast recently that
AI models are already toppingout with their consumption of
knowledge and that we don't haveenough knowledge for them to
grow and expand at the rate thatthey have.
Up to now, they've basicallyconsumed everything that there
(12:53):
is to know as it stands today,and so, while they are good at
summarizing things and creatingthings quickly and giving you
suggestions for things, they'realso still relatively not great
and kind of dumb in a lot ofrespects.
So I'm glad that I can build acustom GPT to help me write a
(13:14):
book that learns my voice andthat suggests chapter titles and
agendas and speaker lineups forconferences.
All these things are fantasticand I have so many ideas If
anybody wants to reach out to meon LinkedIn, for business
development or candidatesourcing all of these wonderful
(13:35):
things you can do with thesetools, but they're just tools
and people still got to bepeople.
Speaker 2 (13:43):
Yeah, they're just tools and people still got to be
people.
Yeah, well, I think, like alldata, all ai models are only as
good as the data that it's sortof trained on.
I think you kind of you know,if you're using or poking around
with chat, gpt or any of these,you can see, you can see the
holes, right, and I think thatyou know that's why you talk to
anyone.
It's, hey, what are the skillsthat are going to be most
important?
As you know, know, ai evolvesand it's, you know, critically
(14:04):
thinking, critical thinkingskills, right, and you know.
So you could pop into JetGPTand say, hey, like you know,
write me an article, you know,or a framework for a blog post
on this and find 10, you know 10sources that back it up.
And I think you know what you dofind is you find is a lot of
these AI tools are pulling fromthat same repository of data,
(14:26):
and I think that ends up being achallenge too, because it's
easy to get articles orquotations from a Harvard
Business Review or an MIT Sloanarticle that go out.
It's not easy to get out andmake sure that, hey, you're
looking at the most relevantsources or pulling the most
relevant data, and so I wouldsay, like, arguably, chat, gpt,
(14:50):
ai is going to be great, but youknow you have to be.
That's why you see promptengineers and all of these roles
that are growing that are, youknow, really critical thinking
type skill sets.
You know that help you get themost out of AI.
Otherwise, you know you'regoing from you know zero to the
40 yard line, but like there's awhole lot of real estate left
to go across the board.
(15:12):
Does that resonate, craig?
Or like is that what you'reseeing too?
Speaker 3 (15:15):
I'm absolutely seeing that, and you do have to be
specific, right, if you wantsources from specific authors or
specific places that aren'tsort of the top of the search
results.
And you grateful for and wejust saw it at the Talent Net
(15:45):
live conference in Austin onMarch 7th was that the
recruiting community and thetalent acquisition community are
embracing the tools, notnecessarily fighting the tools
and using them in very originaland interesting ways, and so I
(16:05):
really get inspired when I seemy colleagues being creative and
that's what it takes.
That's also what it takes towrite a good Boolean search
string or to search any databaseLinkedIn or ats, or even google
.
Uh, when you're sourcing foreither customers or candidates,
(16:29):
and it's not much different, Imean you, you learn the tool,
you learn the platform, youlearn the language and, right,
you get creative with it yeah,well, I think that's great and,
like you know, I know channelapp live and austin.
Speaker 2 (16:43):
You know it was earlier this month.
You know, I think a lot ofconversations are focused on,
you know, ai, but also combiningAI with a human element, which
is really, I think you know,central to part of your thesis
and hiring humans.
I'm curious, you know, comingout of the last week conference,
can you think of any goodexamples where you know
practitioners or companies are,you know, just doing a great job
(17:06):
of blending AI, you know, withthat human?
You know intuition.
Any examples of hiring outcomesyou know are tied to TA, you
know that really do a good jobof blending, you know, ai with
humans.
Speaker 3 (17:18):
Yeah.
So my friend, jason Roberts,who runs some technology
functions at an RPO I guess wewon't name names here, but we
could he's doing someinteresting things with AI and
building sort of the.
At Allegis, we built some kindof Frankenstein software that
(17:39):
pulled data from all yoursources software that pulled
data from all your sources.
So if you were to ask a hiringmanager or anyone in a large
organization how many people dowe currently employ on a
full-time basis and a contractbasis and all the other ways we
might employ people, it's almostimpossible to get it exactly
right, and so what you reallywant in a talent intelligence
(18:03):
platform is all your talent allin one place, to know all the
things that have to happen toget business done, including
churn.
How many people do we have tohire this month?
How many widgets have wepromised to prospective
customers or actual customers?
All these business questionsthat you know talent acquisition
should be asking are nowstarting to be able to get
(18:26):
pulled into easy-to-accessplatforms, because we can feed
the data into these languagemodels and they can really
summarize it for us in real time.
If you build the web hooks intoyour databases, you can get
(18:46):
real time answers for things,and so, taking this to the next
level, you can actually start todo capacity planning, and it's
a mythical thing that we'vetalked about forever, but it's
becoming more and more real.
And so then humans take thatand do the human thing that we
do with it and say, okay, sowe're going to need to hire this
(19:09):
many people.
We can plan this far out, wecan build these kinds of
campaigns with the help of theseGPT models and actually have a
better experience for employees,candidates, hiring managers,
sales leaders, because we'reable to track that path much
(19:29):
better.
Speaker 4 (19:30):
Yeah, that's a very helpful real example.
I mean, I think some of thechallenges we see with these
conversations about AI is thatthey're just so high level and
it's really hard to get past AIas automation and actually get
specific about ways that canreally disrupt, in a very good
way, what we're doing in thespace.
So thanks for that.
And I think the other thing isthat you about what you just
(19:52):
said.
The example is, you know,because this is such a new
technology, I think peoplesuffer from and I'm included
from a lack of imagination aboutwhat's possible, and so we
think of AI just doing thethings that humans are already
doing better.
But you shared an example herewhere it's like even the
smartest, most genius person isnot going to be able to ingest
(20:14):
all of these different datasources and come up with a
cogent, accurate prediction orsuggestion.
You know that's right, that's.
There's no human that can dothat and no human that would
probably want to do that.
It's not a human activity.
So I think that's just a reallygreat marriage of those two
ideas.
Speaker 3 (20:29):
Yeah, you have to look in 30 different places to
try to pull that informationtogether and by the time you got
done doing it, the informationwouldn't be accurate anymore.
Speaker 4 (20:38):
Right, and I think it's worth just pausing here on
this topic a little bit becausepeople are so scared that AI is
going to take over.
But a counterintuitive point Idon't remember which economist I
just read a sub-stack aboutthis but a limiting factor of
the proliferation of these largelanguage models is the
(21:00):
bandwidth of human beings tointeract with them.
Language models is thebandwidth of human beings to
interact with them.
So, yes, you can have somethingthat could do a year's worth of
work that a human might do inan afternoon, but at the end of
that there's a human on theother side of that that has to
interact with those outputs,that has to at least understand
them on some level, maybe notthis super intelligent level,
but just make sense of them andthen make a decision and apply
(21:22):
it within the organization level.
But just make sense of them andthen make a decision and apply
it within the organization.
So it's easy to think of AIs asjust running amok, unchecked,
doing things and, you know,continually making progress, and
I think people just don'tnecessarily pause and think that
humans have a.
There's a natural limitingfunction in there, I think, or
at least arguably.
Speaker 3 (21:39):
There is it's taxing, by the way right, if you've
ever built a webpage or writtena long-form article or done a
research project.
You can only go so far for solong.
So your hands get tired oftyping, your eyes get tired of
lights and screens and yourbrain just has fatigue.
And trying to get the rightequation into these language
(22:04):
models to get the desiredoutcome takes a lot of uh of of
practice and a lot of testing,and so it you.
You really can only go for solong until you have to take a
break, and so you're very rightabout this.
We did a project for theworld's largest bank Again, I'm
not going to name names, butit's the largest financial
(22:25):
institution in the world and wehad to rewrite 13,000 job
descriptions for them.
So they use an ATS thatdisplays a little blurb of each
job across the careers page ontheir website, and so they can't
all be exactly the same.
They have to have somedifferentiating language up
(22:46):
front in order for them not tolook just completely generic
across the board.
So we had to rewrite thetemplates for all of these jobs
and I created a custom GPT forit to get the syntax just right,
and we created a large databaseof bias language for it to look
for and omit, and then a modelfor it to write some distinct
(23:09):
and unique marketing languageupfront that speaks to the
candidate first.
You're this type of person.
You like this kind of challenge.
We're looking for people likeyou.
Imagine what we can do together.
Some variation on that 13,000times what we can do together.
Some variation on that 13,000times.
Now it's all great to say, yeah, the computer can just do it,
but it can't.
You have to still human eyeballevery one of them because
(23:32):
there's risk involved.
Right?
You can't just leave that up toa machine.
And so imagine the taxingnature of trying to human
eyeball test 13,000 jobs, evenif you've got a really great
custom GPT for it.
Yeah, yeah.
So the project was supposed tostart with three to six months.
It took two years.
Speaker 2 (23:52):
How did that project come about, craig?
I'm curious like what's theimpetus for, like you know,
large bank companies?
Is like sees, hey, we've got13,000 jobs.
Like was it driven by?
Hey, we don't like the way thislooks on our career site.
Or like publicly to job seekers.
Or was it, like you know, we'renot getting the right people.
Like I'm just curious, likewhat's you know what's the
(24:12):
impetus for?
Like, hey, we need to do thisdifferently.
Speaker 3 (24:15):
So these were just the tech jobs, by the way.
So there's a lot more jobs inthis organization.
But there are several thingsthat employers have to be aware
of and watch out for, and mostjob descriptions for companies
are very outdated.
Their templates get cloned andreused over and over again in
the applicant tracking system.
(24:36):
That has problems of its ownright.
Those templates, if notrecreated properly, can carry
over old JSON metadata thatmakes your jobs look old in
search results on places likeIndeed, and that's a big, costly
problem.
And so we fix things like riskmitigation in jobs because of
(24:57):
bias words or pay transparency.
If your wages aren't in linewith what the market says and
you're way off, you could getsued for that now.
So there's all kinds of reasonsto want to look at and update
your job descriptions on aregular basis, and that's one of
the actually pretty fun thingsthat we do on a at scale.
Speaker 2 (25:16):
Yeah, well, you know, I think that's funny, right, so
you're talking about a reallybig organization.
Like you know, I think we gointo projects, you know,
probably for similar asks, right, and like you know, we call it
the basic blockading tacklingright of of recruitment.
And yeah, like, hey, I, youknow we were working with a a
new client recently and you know, let's just say you know
(25:40):
they're recruiting for want tomake it up, how can I make up uh
, or let's just say they'rerecruiting for software
engineers and you know the jobdescription, recruiting for want
to make it up.
How can I make up a word?
Let's just say they'rerecruiting for software
engineers and you know, in thejob description that they've
been using for months, youdidn't have a software engineer
in any of the jobs and like youknow, it's not rocket yeah.
Right.
And so, you know, a lot oftimes I think we get so excited
(26:00):
to, you know, jump to the nextgreat thing and like, hey, how
are we going to automate, howare we going to move people
through the process?
And like, hey, we got to talkabout all of our benefits and
you forget the basics.
And I think what you're saying,I think, is like, hey, we can
trust the machines to do so much, but sometimes you need what
(26:20):
machines might be.
Lacking is a little bit ofcommon sense, that critical
thinking aspect.
You don't want to run 13,000jobs through some great new AI
tool that might bolt onto Oracleand have that output be.
Yeah, we're hiring forengineers, but we're not going
to put that in the jobdescription.
That's how people get in thedoor.
Speaker 3 (26:42):
It's interesting because you're alluding to basic
SEO, which is one of myfavorite subjects of all time,
and that's accurate, right?
How do you come up in searchresults, search for your own
jobs right on Google and Indeedand see what happens?
If you're getting clobbered byyour competitors, there's a
problem.
But now there's another problemthat involves AI, because job
(27:05):
candidates are now using AI tohyper-elevate their profiles and
resumes to match all thesedifferent kinds of jobs, and if
your job description isn't AIready, it's not going to go
through the matching softwarevery well and bring up the right
candidates, and so employersall over the place are getting
(27:28):
too many of the wrong candidatesbecause their job descriptions
aren't up to par.
Speaker 2 (27:32):
Yeah, well, a hundred percent agree.
And I'd also say, before we leantoo heavy on AI to do
everything, I think I saw a statyesterday that the percentage
of jobs everyone's worried aboutwriting their job description,
so AI tools are going torecognize it or it's going to
come up in chat GPT searches andall this stuff.
And so I think that the stat Isaw was last year the percentage
(27:57):
of search volume share onGoogle compared to chat GPT and
it's, you know something of, youknow some obscene number like
you know, 99 point.
You know, uh, you know 90, 99%of all searches done in the U S
in 2024, I guess, like 93%, Ijust pulled it up 93.57% of all
(28:17):
searches done, you know, acrosssearch platforms in 2024, we're
done on Google.
0.25% we're done on chat GPT,so it's 373 times more searches
are still being done on Google.
So like, hey, maybe weshouldn't be optimizing for
things to come up in chat GPTjust yet.
You know, should we payattention to it?
Sure, but like, let's not.
(28:37):
You know, let's not gloss overthe basics.
Speaker 3 (28:40):
Well, yes, and I think the key thing right there
is chat GPT isn't necessarilythe only answer right.
So now with Google, becauseGoogle is not a great search
engine anymore.
Let's just face it right.
You get, you do a search andyou get video, video, video
video, sponsored ad.
Sponsored ad, sponsored ad andit's not giving you the thing
(29:03):
you're looking for.
But now you have add-ons likeGemini that do summarize the
results and do give you thething you're looking for.
So maybe optimizing for chatGPT isn't quite right, but
optimizing for AI in general isstill a very good idea.
I built a whole part of ourwebsite about it and I wrote
(29:28):
about it, and you can see it atemployerseocom.
This is a fascinating topic tome.
I'm a huge fan.
Speaker 4 (29:35):
Yeah Well, I'm sure most folks in the audience have
at least had the experience ofperforming a Google search these
days and seeing the AI-poweredresults at the top.
So you know, obviously we'repointing towards a future where
that's probably the norm.
Really fascinating to thinkabout.
I'm sure you've thought a lotmore than I have, Craig, but I
(29:56):
don't even know how you wouldthink of an optimized name for
different large language models,for example.
I mean, it's been such an armsrace over the last 15 years with
SEO, and part of that is Googlefamously has a secret algorithm
and people can sort oftriangulate how to game it if
(30:17):
you will, or optimize for it,but there's not a clear rule
book and it strikes me that ifyou think about optimizing for a
chat, GPT or any othergenerative AI, there's an even
bigger black box there.
But I don't know what are yourthoughts.
Do you have any perspective onthat?
Speaker 3 (30:33):
100%, and also I hate the term 100%.
I'm sure I didn't just say that.
So if you think aboutcontextual search, which is what
large language models do, thatmeans that you have to have some
storyline in your jobdescriptions in order for them
(30:54):
to do well.
Also, search engines, whichLLMs rely on for their results,
also really hyper focus onlocality.
So think about it.
If you search for something,the first thing Google's going
to try to give you is localresults.
Can we access your currentlocation?
(31:16):
And so very often employers arenot doing a good enough job of
that and there's some nuance toit.
There's too much to go intoright now but that could be a
whole other podcast if we wantedto do it.
But there's certainly somenuance to optimizing for
locality, for contextual search,and there are 50 other things
(31:38):
that ChatGPT and Gemini and BARDand all these Claude they all
look for that job descriptionsdon't have.
So it's a puzzle and it's noteasily solved.
I mean there's a little bit ofsome study to it.
Speaker 2 (32:00):
Yeah, I think that's great and I'm going to selfishly
try and get one more questionin that.
I want to talk about Craigbefore we wrap, and that's like
(32:20):
we've been talking a lot aboutattraction.
Really love is this concept ofthe nurture wheel outlining
stages of candidate attraction?
I'm just curious.
Well, first let's set the stage.
I think the six stages areattraction, engagement,
education, conversion, retentionand then employee advocacy.
I'm just curious with AI toolscoming into place, with
(32:42):
automation, which of thesestages, if any you know, do you
kind of see, you know, companiessort of neglecting you know the
most, or AI sort of shaking upthe most in this candidate life
cycle?
Speaker 3 (32:54):
Yeah.
So I think retention andadvocacy are the most overlooked
.
We're starting to get it right.
Like you know, I said on thefront end of things and possibly
automating too much, and we'llcircle back to a more human
experience at some point.
But we could use these tools toreally have a fantastic
(33:18):
onboarding experience.
We're not doing that.
You can set up the algorithms tomessage a manager to say, hey,
now would be a good time to tapin with your brand new employee
with this kind of message andreally make employees feel
wanted, comfortable.
Summarize the massive amounts ofinformation we're giving them,
(33:40):
rather than make them read andwatch hours and days of videos
them, rather than make them readand watch hours and days of
videos.
There's plenty of opportunitythere.
And then regular touch pointsyou know, set alerts, set.
You can now there's a chat GPTthat will do repetitive tasks
and go back and, you know,update things and you can have a
(34:00):
much richer experience ifyou're doing your touch points
correctly.
And then the advocacy pieceright Survey your people as soon
as they get hired that's thetime they're the happiest and
ask them how they felt about theexperience and then you know if
they have a rich answer for you.
(34:22):
Give them a button, ask them ifthey want to write something
about that on LinkedIn andpossibly even Glassdoor or
Indeed reviews, and you know,you can change the trajectory of
your review status very quicklyby doing something like this.
And all of this could be easilyautomated with with AI.
Speaker 2 (34:41):
Yeah, I think that.
I think that's great and I'llsay, you know, retention,
attrition, you know, are two ofthe not two of the two highest
measurements that TA leaders arepaying attention to, you know,
when we look at our TA trendstudy that we just published
earlier this year.
So the good news is I think,hey, if that's a gap that you
see or an opportunity you see, Ithink TA leaders are starting
(35:03):
to recognize it too, or at leastyou know the people that are,
we like to think the smartestpeople are paying attention to
our surveys and insights.
So nice to hear that resonate,craig.
Speaker 3 (35:12):
Yeah, can I plug one more thing real quick?
Yeah, of course.
My next book is called Paintyour Store, and a lot of the
concepts we talked about todayare in Paint your Store.
But it's, whereas hiring humansis for employers, hiring
managers, recruiters, paint yourStore is for the job seeker or
(35:35):
the person looking for theirnext career move, or the
salesperson looking to gathermore clients, or the
organization looking to bring onmore customers clients or the
organization looking to bring onmore customers.
If you're a local independentconsultant and you want to just
attract more gigs, this book'sfor you.
It comes out March 25th and itwill be on Amazon.
(35:56):
How to Network your Way to yourNext Career Move Paint your
Store by Craig Fisher.
Speaker 2 (36:01):
That's awesome.
Well, this podcast is going tocome out on the 25th, so it's
perfect timing.
It's like we almost planned itthat way.
Craig, I'll ask the easiestquestion of all.
We'll certainly link everythingin the show notes.
Where else can listeners learna little bit more about you and
your work online?
Speaker 3 (36:17):
Yeah, the easiest place is hiring-humanscom it's
kind of a catch-all for my booksand my work, and dash humanscom
it's kind of a catch all for mybooks and my work, and it's at
the talent net live website, uh,and you can find out all about
great ways to engage with me andI'm also easily searchable.
So find me on LinkedIn andabsolutely connect.
Speaker 2 (36:36):
Yeah, and I love the uh fish dogs.
Uh to you know, marty and I alot of us, especially over at
G's day all big fans of dogs.
So, uh, anyone else can checkout that story in your free time
.
On how Craig came up with thatwebsite domain too, I love it,
thank you.
Awesome.
Well, thanks, Craig.
All right, Thanks for tuning in.
As always, head on over tochangestateio or shoot us a note
(37:00):
on all the social media.
We'd love to hear from you andwe'll check you guys next week.
Welcome to the Changing State of Talent
Acquisition, where your hosts,graham Thornton and Martin Kred,
share their unfiltered takes onwhat's happening in the world
of talent acquisition today.
Each week brings new guests whoshare their stories on the
tools, trends and technologiescurrently impacting the changing
state of talent acquisition.
Have feedback or want to jointhe show?
(00:21):
Head on over to changestateio.
And now on to this week'sepisode.
Speaker 2 (00:27):
All right, and we're back with another episode of the
Changing State of TalentAcquisition Podcast, super
excited for our next guest withan incredible background, craig
Fisher, the CEO of TalentNetMedia, author of multiple books,
most recently Hiring Humans,host of TalentNet Live
conferences and just a generalwell-known HR tech advisor in
(00:49):
our space.
So, craig, welcome to the show.
Speaker 3 (00:51):
Graham, thank you.
It is a great pleasure to behere with you today.
Speaker 2 (00:55):
Would love for you to maybe briefly share your
journey, from your early days inrecruitment marketing help our
audience get to know you alittle bit better to your days
starting Talent Media, and maybehow have your experiences
shaped your perspective on theworld of talent acquisition.
Speaker 3 (01:10):
Yeah.
So I came up through staffinglike a lot of people in the 90s.
Before that I was a drug repand a medical sales guy, but I
went through all the levelsrecruiter, account manager,
leader, owner of a staffingcompany starting in 2007.
(01:31):
And all the while I was doingthat, I come from an advertising
and marketing college degreebackground and when I'm doing
all this stuff, I'm attemptingto differentiate myself by
helping both the employers thatI'm working with and the
candidates that I'm working withto market themselves better.
(01:53):
And I developed some strongopinions on it and some nifty
hacks and methodologies and Istarted writing about it in blog
posts and on Twitter.
And in the early days of Twitterin the late 2000s, it was a hot
topic.
Is social media good forrecruiting?
(02:13):
I'm like, yes, it's the bestthing ever, come on.
And so I got a little Twitterfamous for my opinionated
outlook on it.
But also I started the firsthashtag chat for recruiters
called TalentNetLive.
At the time we went withhashtag TNL.
(02:35):
So I own a staffing company, I'mbuilding community.
I'm attempting to push my viewof the world, which is treat
candidates better, explainbetter about yourself as an
employer in a moderated fashion.
I'm saying listen, I think I'vegot a better idea than what's
happening right now.
(02:55):
That's very forward thinkingfor a guy that didn't really
have any platform for it yet.
But my goal was to get on stageand talk about it and get out
of the day-to-day of recruitingwith my staffing agency and
(03:16):
executive search firm.
And people did start invitingme to come speak to their teams
and talk about it.
And then the Twitter chatquickly became a request for IRL
in real life meetings and westarted the first ever Talent
Net Live conference in the fallof 2009.
(03:39):
And we had on our first boardand at our first conference were
notables like William Tencupand Chris Hoyt and Jim Schneider
, bill Borman lots ofinteresting folks from talent
acquisition were there and itreally kicked off the
unconference theme forrecruiting conferences.
(04:01):
So from there I was able topivot into full-time consulting,
doing recruitment, marketingand employer brand, starting in
2011.
So I kind of made TalentNetMedia a thing back then and
since then I've gotten to workon amazing projects with you
(04:22):
know Fortune 50 companies andsome hungry startups and
occasionally go in-house to doleadership jobs at companies
like CA Technologies and AllegisGlobal Solutions.
I was the CMO there for a whileand got to work with big
customers like Amex, hsbc and GM.
So it's been a wonderful wildride and I think to those
(04:47):
starting out now in the field,have an opinion right.
It all starts with having apoint of view on things and
don't be generic and watereddown.
Speaker 4 (04:58):
Well, thanks for sharing, craig.
We're so thrilled to have youon the podcast.
We've got a lot of guests thatjoin us, but I don't think
anyone quite has the depth ofexperience and different
experiences in the industry asyou have, and it's also
interesting that you kind ofwore a lot of the different hats
that exist in this space fromyour early days to now.
So I'd love to kind of start byjust getting kind of a long
(05:20):
view question.
You know you spend a lot oftime you've spent a lot of time
in this space thinkingparticularly about emerging
trends, cultivating communitiesyou just talked about At this
particular moment.
What are the trends or shiftsthat you're particularly
interested in or payingattention to?
You know, obviously AI is a bigtopic for everybody.
(05:40):
It could be about AI, or maybeit's not about AI.
You know what's on your mindthese days.
Speaker 3 (05:44):
Well, it's interesting.
So there's multiple factorsputting pressure on the talent
space right now.
Currently, we don't have lowunemployment.
Well, it's kind of slowlygrowing.
The numbers are a littlemisleading, but what's really
the problem is there's not muchchurn happening within
(06:05):
organizations.
There are layoffs, but peoplearen't voluntarily leaving their
jobs.
They might be quietly looking,but they are afraid of this
economy still and not leavingtheir jobs.
So talent acquisition teamshave been shrinking, partly
because of the economy, partlybecause of AI being able to help
(06:28):
us automate things.
But you know, we've seen thesetrends in the economy happen
every 10 or so years.
You know for multiple seasonsthat this too shall pass and
we're going to see a turnaround,because the trend right now is
to attempt to automateeverything in the candidate
(06:50):
process, and that's not enoughfriction.
So this frictionless idea is alittle overblown.
There has to be some humaninteraction, because if it is
too easy, candidates won'tbelieve it's real, and so when
hiring comes back, there will bea need for recruiters again.
(07:12):
There probably will be a needfor sourcers again.
We will be able to do a lot ofit, though better and easier,
because of the AI tools.
I mean, think of the thingsthat you feed into chat GPT
right now and say, okay, buildme a plan for this and it's so
easy to do.
But you can't just trust itblindly.
(07:32):
Somebody still has to say, yeah, this is good, this is not good
, this is ridiculous, this is ahallucination, and you know,
make it real.
Speaker 4 (07:43):
Sure Well, craig, earlier you were talking about
the importance of having a pointof view and you know I think
you just expressed it probablyan intentionally provocative one
, or at least it sounds like itto me this idea that we need
friction in the process, andnormally, whether you're talking
about customer experience orcandidate experience, we seem to
, as an industry or business,people try to aim for as less
(08:05):
friction as possible.
Could you unpack that a littlebit?
You know what do you mean bythat exactly.
Why is friction a good thing,and how should organizations be
thinking about it?
Speaker 3 (08:15):
Yeah.
So my friend Jim D'Amico and Italk about this and he agrees
with me that if there is nofriction, the process seems like
you're going to get ghosted100% if there's no human
intervention at all, and sothere has to be some human touch
(08:36):
points.
A little bit of friction isimportant.
So I feel that we're going togo over our skis with our
attempt to automate everythingand take recruiters and humans
out of the mix, because itleaves you with a process that
just feels vacant.
Speaker 2 (08:58):
So you know you've written a book recently, you
know, Hiring Humans, and I thinkyou know, really, one of the
points emphasized, you know, inthat book is that hiring is
really a human endeavor and youknow there is some balance
between technology and the humantouch playing out.
So you know, let's dive intoHiring Humans a bit, craig.
(09:19):
So you know what drove you towrite that book and let's the
high-level theme that is mostapplicable in today's market my
philosophy, graham, as you mayknow, is that people want to
work with people and we haveautomation.
Speaker 3 (09:39):
We have all these great tools and I'm a techie guy
myself great tools, and I'm atechie guy myself and I like
nothing more than to be alone,put my phone down and crank away
on a project, write some code,whatever it is right, edit a
video, but at some point I needpeople.
So even the hardcore coder thatin theory works in a basement
(10:07):
is apt to occasionally have aquestion about their benefits or
their quarterly reviews orarrays or something like that.
And you can't automate all ofthose things.
You have to have people in theorganization.
And if there's going to beinternal mobility, if there's
(10:27):
going to be culture of any kind,if there's going to be what I
think is going to be an attemptto keep people in organizations
longer because right now we'rein this sort of gig economy
where things are very temporaryand people don't have any very
good reason to stay at a companylong term.
(10:47):
I think without more of a humantouch in that process also,
we're not going to be verysuccessful.
So, yes, automation is great.
Yes, tools are good, butthey're just tools.
If you remember when the iPhonewas introduced right 2008, that
was going to change everything.
We won't need recruitersanymore.
(11:08):
Same thing when the internethappened.
I mean all these greatinventions are going to do away
with the need for humans, but itturns out no, they're just
tools and they need people touse them.
Speaker 4 (11:20):
Yeah, I think that's a really interesting point,
craig, and it sounds when yousay it out loud, it sounds
almost obvious, but it is asubtle point and I don't think
it's a point that is.
I think it's a point that'slost on a lot of people, whether
it's recruiting or any otherdomain that's going to be
impacted by AI.
There's a lot of fear.
Of course, people are worriedthat it's going to take all the
(11:40):
jobs, but it seems to miss thisfundamental point, that why are
we doing this in the first place?
This is a human creation, firstof all, and human beings really
cannot be reduced to a machine.
Sure, a machine may be able todo a lot of the things that a
human can do, and do them morequickly or better or more
(12:01):
efficiently, but it doesn't havethose intangibles, and that's
what makes life and work andcareers meaningful.
Is that a fair encapsulation ofyour point of view?
Speaker 3 (12:11):
It is, and so there's an REM record called Life's
Rich Pageant and it always singsto me.
In respect to the humanexperience, yeah, we created
these machines for our benefit.
(12:33):
Computers are finally doing whatwe always wanted them or hoped
for them to do, but I read andheard on a podcast recently that
AI models are already toppingout with their consumption of
knowledge and that we don't haveenough knowledge for them to
grow and expand at the rate thatthey have.
Up to now, they've basicallyconsumed everything that there
(12:53):
is to know as it stands today,and so, while they are good at
summarizing things and creatingthings quickly and giving you
suggestions for things, they'realso still relatively not great
and kind of dumb in a lot ofrespects.
So I'm glad that I can build acustom GPT to help me write a
(13:14):
book that learns my voice andthat suggests chapter titles and
agendas and speaker lineups forconferences.
All these things are fantasticand I have so many ideas If
anybody wants to reach out to meon LinkedIn, for business
development or candidatesourcing all of these wonderful
(13:35):
things you can do with thesetools, but they're just tools
and people still got to bepeople.
Speaker 2 (13:43):
Yeah, they're just tools and people still got to be
people.
Yeah, well, I think, like alldata, all ai models are only as
good as the data that it's sortof trained on.
I think you kind of you know,if you're using or poking around
with chat, gpt or any of these,you can see, you can see the
holes, right, and I think thatyou know that's why you talk to
anyone.
It's, hey, what are the skillsthat are going to be most
important?
As you know, know, ai evolvesand it's, you know, critically
(14:04):
thinking, critical thinkingskills, right, and you know.
So you could pop into JetGPTand say, hey, like you know,
write me an article, you know,or a framework for a blog post
on this and find 10, you know 10sources that back it up.
And I think you know what you dofind is you find is a lot of
these AI tools are pulling fromthat same repository of data,
(14:26):
and I think that ends up being achallenge too, because it's
easy to get articles orquotations from a Harvard
Business Review or an MIT Sloanarticle that go out.
It's not easy to get out andmake sure that, hey, you're
looking at the most relevantsources or pulling the most
relevant data, and so I wouldsay, like, arguably, chat, gpt,
(14:50):
ai is going to be great, but youknow you have to be.
That's why you see promptengineers and all of these roles
that are growing that are, youknow, really critical thinking
type skill sets.
You know that help you get themost out of AI.
Otherwise, you know you'regoing from you know zero to the
40 yard line, but like there's awhole lot of real estate left
to go across the board.
(15:12):
Does that resonate, craig?
Or like is that what you'reseeing too?
Speaker 3 (15:15):
I'm absolutely seeing that, and you do have to be
specific, right, if you wantsources from specific authors or
specific places that aren'tsort of the top of the search
results.
And you grateful for and wejust saw it at the Talent Net
(15:45):
live conference in Austin onMarch 7th was that the
recruiting community and thetalent acquisition community are
embracing the tools, notnecessarily fighting the tools
and using them in very originaland interesting ways, and so I
(16:05):
really get inspired when I seemy colleagues being creative and
that's what it takes.
That's also what it takes towrite a good Boolean search
string or to search any databaseLinkedIn or ats, or even google
.
Uh, when you're sourcing foreither customers or candidates,
(16:29):
and it's not much different, Imean you, you learn the tool,
you learn the platform, youlearn the language and, right,
you get creative with it yeah,well, I think that's great and,
like you know, I know channelapp live and austin.
Speaker 2 (16:43):
You know it was earlier this month.
You know, I think a lot ofconversations are focused on,
you know, ai, but also combiningAI with a human element, which
is really, I think you know,central to part of your thesis
and hiring humans.
I'm curious, you know, comingout of the last week conference,
can you think of any goodexamples where you know
practitioners or companies are,you know, just doing a great job
(17:06):
of blending AI, you know, withthat human?
You know intuition.
Any examples of hiring outcomesyou know are tied to TA, you
know that really do a good jobof blending, you know, ai with
humans.
Speaker 3 (17:18):
Yeah.
So my friend, jason Roberts,who runs some technology
functions at an RPO I guess wewon't name names here, but we
could he's doing someinteresting things with AI and
building sort of the.
At Allegis, we built some kindof Frankenstein software that
(17:39):
pulled data from all yoursources software that pulled
data from all your sources.
So if you were to ask a hiringmanager or anyone in a large
organization how many people dowe currently employ on a
full-time basis and a contractbasis and all the other ways we
might employ people, it's almostimpossible to get it exactly
right, and so what you reallywant in a talent intelligence
(18:03):
platform is all your talent allin one place, to know all the
things that have to happen toget business done, including
churn.
How many people do we have tohire this month?
How many widgets have wepromised to prospective
customers or actual customers?
All these business questionsthat you know talent acquisition
should be asking are nowstarting to be able to get
(18:26):
pulled into easy-to-accessplatforms, because we can feed
the data into these languagemodels and they can really
summarize it for us in real time.
If you build the web hooks intoyour databases, you can get
(18:46):
real time answers for things,and so, taking this to the next
level, you can actually start todo capacity planning, and it's
a mythical thing that we'vetalked about forever, but it's
becoming more and more real.
And so then humans take thatand do the human thing that we
do with it and say, okay, sowe're going to need to hire this
(19:09):
many people.
We can plan this far out, wecan build these kinds of
campaigns with the help of theseGPT models and actually have a
better experience for employees,candidates, hiring managers,
sales leaders, because we'reable to track that path much
(19:29):
better.
Speaker 4 (19:30):
Yeah, that's a very helpful real example.
I mean, I think some of thechallenges we see with these
conversations about AI is thatthey're just so high level and
it's really hard to get past AIas automation and actually get
specific about ways that canreally disrupt, in a very good
way, what we're doing in thespace.
So thanks for that.
And I think the other thing isthat you about what you just
(19:52):
said.
The example is, you know,because this is such a new
technology, I think peoplesuffer from and I'm included
from a lack of imagination aboutwhat's possible, and so we
think of AI just doing thethings that humans are already
doing better.
But you shared an example herewhere it's like even the
smartest, most genius person isnot going to be able to ingest
(20:14):
all of these different datasources and come up with a
cogent, accurate prediction orsuggestion.
You know that's right, that's.
There's no human that can dothat and no human that would
probably want to do that.
It's not a human activity.
So I think that's just a reallygreat marriage of those two
ideas.
Speaker 3 (20:29):
Yeah, you have to look in 30 different places to
try to pull that informationtogether and by the time you got
done doing it, the informationwouldn't be accurate anymore.
Speaker 4 (20:38):
Right, and I think it's worth just pausing here on
this topic a little bit becausepeople are so scared that AI is
going to take over.
But a counterintuitive point Idon't remember which economist I
just read a sub-stack aboutthis but a limiting factor of
the proliferation of these largelanguage models is the
(21:00):
bandwidth of human beings tointeract with them.
Language models is thebandwidth of human beings to
interact with them.
So, yes, you can have somethingthat could do a year's worth of
work that a human might do inan afternoon, but at the end of
that there's a human on theother side of that that has to
interact with those outputs,that has to at least understand
them on some level, maybe notthis super intelligent level,
but just make sense of them andthen make a decision and apply
(21:22):
it within the organization level.
But just make sense of them andthen make a decision and apply
it within the organization.
So it's easy to think of AIs asjust running amok, unchecked,
doing things and, you know,continually making progress, and
I think people just don'tnecessarily pause and think that
humans have a.
There's a natural limitingfunction in there, I think, or
at least arguably.
Speaker 3 (21:39):
There is it's taxing, by the way right, if you've
ever built a webpage or writtena long-form article or done a
research project.
You can only go so far for solong.
So your hands get tired oftyping, your eyes get tired of
lights and screens and yourbrain just has fatigue.
And trying to get the rightequation into these language
(22:04):
models to get the desiredoutcome takes a lot of uh of of
practice and a lot of testing,and so it you.
You really can only go for solong until you have to take a
break, and so you're very rightabout this.
We did a project for theworld's largest bank Again, I'm
not going to name names, butit's the largest financial
(22:25):
institution in the world and wehad to rewrite 13,000 job
descriptions for them.
So they use an ATS thatdisplays a little blurb of each
job across the careers page ontheir website, and so they can't
all be exactly the same.
They have to have somedifferentiating language up
(22:46):
front in order for them not tolook just completely generic
across the board.
So we had to rewrite thetemplates for all of these jobs
and I created a custom GPT forit to get the syntax just right,
and we created a large databaseof bias language for it to look
for and omit, and then a modelfor it to write some distinct
(23:09):
and unique marketing languageupfront that speaks to the
candidate first.
You're this type of person.
You like this kind of challenge.
We're looking for people likeyou.
Imagine what we can do together.
Some variation on that 13,000times what we can do together.
Some variation on that 13,000times.
Now it's all great to say, yeah, the computer can just do it,
but it can't.
You have to still human eyeballevery one of them because
(23:32):
there's risk involved.
Right?
You can't just leave that up toa machine.
And so imagine the taxingnature of trying to human
eyeball test 13,000 jobs, evenif you've got a really great
custom GPT for it.
Yeah, yeah.
So the project was supposed tostart with three to six months.
It took two years.
Speaker 2 (23:52):
How did that project come about, craig?
I'm curious like what's theimpetus for, like you know,
large bank companies?
Is like sees, hey, we've got13,000 jobs.
Like was it driven by?
Hey, we don't like the way thislooks on our career site.
Or like publicly to job seekers.
Or was it, like you know, we'renot getting the right people.
Like I'm just curious, likewhat's you know what's the
(24:12):
impetus for?
Like, hey, we need to do thisdifferently.
Speaker 3 (24:15):
So these were just the tech jobs, by the way.
So there's a lot more jobs inthis organization.
But there are several thingsthat employers have to be aware
of and watch out for, and mostjob descriptions for companies
are very outdated.
Their templates get cloned andreused over and over again in
the applicant tracking system.
(24:36):
That has problems of its ownright.
Those templates, if notrecreated properly, can carry
over old JSON metadata thatmakes your jobs look old in
search results on places likeIndeed, and that's a big, costly
problem.
And so we fix things like riskmitigation in jobs because of
(24:57):
bias words or pay transparency.
If your wages aren't in linewith what the market says and
you're way off, you could getsued for that now.
So there's all kinds of reasonsto want to look at and update
your job descriptions on aregular basis, and that's one of
the actually pretty fun thingsthat we do on a at scale.
Speaker 2 (25:16):
Yeah, well, you know, I think that's funny, right, so
you're talking about a reallybig organization.
Like you know, I think we gointo projects, you know,
probably for similar asks, right, and like you know, we call it
the basic blockading tacklingright of of recruitment.
And yeah, like, hey, I, youknow we were working with a a
new client recently and you know, let's just say you know
(25:40):
they're recruiting for want tomake it up, how can I make up uh
, or let's just say they'rerecruiting for software
engineers and you know the jobdescription, recruiting for want
to make it up.
How can I make up a word?
Let's just say they'rerecruiting for software
engineers and you know, in thejob description that they've
been using for months, youdidn't have a software engineer
in any of the jobs and like youknow, it's not rocket yeah.
Right.
And so, you know, a lot oftimes I think we get so excited
(26:00):
to, you know, jump to the nextgreat thing and like, hey, how
are we going to automate, howare we going to move people
through the process?
And like, hey, we got to talkabout all of our benefits and
you forget the basics.
And I think what you're saying,I think, is like, hey, we can
trust the machines to do so much, but sometimes you need what
(26:20):
machines might be.
Lacking is a little bit ofcommon sense, that critical
thinking aspect.
You don't want to run 13,000jobs through some great new AI
tool that might bolt onto Oracleand have that output be.
Yeah, we're hiring forengineers, but we're not going
to put that in the jobdescription.
That's how people get in thedoor.
Speaker 3 (26:42):
It's interesting because you're alluding to basic
SEO, which is one of myfavorite subjects of all time,
and that's accurate, right?
How do you come up in searchresults, search for your own
jobs right on Google and Indeedand see what happens?
If you're getting clobbered byyour competitors, there's a
problem.
But now there's another problemthat involves AI, because job
(27:05):
candidates are now using AI tohyper-elevate their profiles and
resumes to match all thesedifferent kinds of jobs, and if
your job description isn't AIready, it's not going to go
through the matching softwarevery well and bring up the right
candidates, and so employersall over the place are getting
(27:28):
too many of the wrong candidatesbecause their job descriptions
aren't up to par.
Speaker 2 (27:32):
Yeah, well, a hundred percent agree.
And I'd also say, before we leantoo heavy on AI to do
everything, I think I saw a statyesterday that the percentage
of jobs everyone's worried aboutwriting their job description,
so AI tools are going torecognize it or it's going to
come up in chat GPT searches andall this stuff.
And so I think that the stat Isaw was last year the percentage
(27:57):
of search volume share onGoogle compared to chat GPT and
it's, you know something of, youknow some obscene number like
you know, 99 point.
You know, uh, you know 90, 99%of all searches done in the U S
in 2024, I guess, like 93%, Ijust pulled it up 93.57% of all
(28:17):
searches done, you know, acrosssearch platforms in 2024, we're
done on Google.
0.25% we're done on chat GPT,so it's 373 times more searches
are still being done on Google.
So like, hey, maybe weshouldn't be optimizing for
things to come up in chat GPTjust yet.
You know, should we payattention to it?
Sure, but like, let's not.
(28:37):
You know, let's not gloss overthe basics.
Speaker 3 (28:40):
Well, yes, and I think the key thing right there
is chat GPT isn't necessarilythe only answer right.
So now with Google, becauseGoogle is not a great search
engine anymore.
Let's just face it right.
You get, you do a search andyou get video, video, video
video, sponsored ad.
Sponsored ad, sponsored ad andit's not giving you the thing
(29:03):
you're looking for.
But now you have add-ons likeGemini that do summarize the
results and do give you thething you're looking for.
So maybe optimizing for chatGPT isn't quite right, but
optimizing for AI in general isstill a very good idea.
I built a whole part of ourwebsite about it and I wrote
(29:28):
about it, and you can see it atemployerseocom.
This is a fascinating topic tome.
I'm a huge fan.
Speaker 4 (29:35):
Yeah Well, I'm sure most folks in the audience have
at least had the experience ofperforming a Google search these
days and seeing the AI-poweredresults at the top.
So you know, obviously we'repointing towards a future where
that's probably the norm.
Really fascinating to thinkabout.
I'm sure you've thought a lotmore than I have, Craig, but I
(29:56):
don't even know how you wouldthink of an optimized name for
different large language models,for example.
I mean, it's been such an armsrace over the last 15 years with
SEO, and part of that is Googlefamously has a secret algorithm
and people can sort oftriangulate how to game it if
(30:17):
you will, or optimize for it,but there's not a clear rule
book and it strikes me that ifyou think about optimizing for a
chat, GPT or any othergenerative AI, there's an even
bigger black box there.
But I don't know what are yourthoughts.
Do you have any perspective onthat?
Speaker 3 (30:33):
100%, and also I hate the term 100%.
I'm sure I didn't just say that.
So if you think aboutcontextual search, which is what
large language models do, thatmeans that you have to have some
storyline in your jobdescriptions in order for them
(30:54):
to do well.
Also, search engines, whichLLMs rely on for their results,
also really hyper focus onlocality.
So think about it.
If you search for something,the first thing Google's going
to try to give you is localresults.
Can we access your currentlocation?
(31:16):
And so very often employers arenot doing a good enough job of
that and there's some nuance toit.
There's too much to go intoright now but that could be a
whole other podcast if we wantedto do it.
But there's certainly somenuance to optimizing for
locality, for contextual search,and there are 50 other things
(31:38):
that ChatGPT and Gemini and BARDand all these Claude they all
look for that job descriptionsdon't have.
So it's a puzzle and it's noteasily solved.
I mean there's a little bit ofsome study to it.
Speaker 2 (32:00):
Yeah, I think that's great and I'm going to selfishly
try and get one more questionin that.
I want to talk about Craigbefore we wrap, and that's like
(32:20):
we've been talking a lot aboutattraction.
Really love is this concept ofthe nurture wheel outlining
stages of candidate attraction?
I'm just curious.
Well, first let's set the stage.
I think the six stages areattraction, engagement,
education, conversion, retentionand then employee advocacy.
I'm just curious with AI toolscoming into place, with
(32:42):
automation, which of thesestages, if any you know, do you
kind of see, you know, companiessort of neglecting you know the
most, or AI sort of shaking upthe most in this candidate life
cycle?
Speaker 3 (32:54):
Yeah.
So I think retention andadvocacy are the most overlooked
.
We're starting to get it right.
Like you know, I said on thefront end of things and possibly
automating too much, and we'llcircle back to a more human
experience at some point.
But we could use these tools toreally have a fantastic
(33:18):
onboarding experience.
We're not doing that.
You can set up the algorithms tomessage a manager to say, hey,
now would be a good time to tapin with your brand new employee
with this kind of message andreally make employees feel
wanted, comfortable.
Summarize the massive amounts ofinformation we're giving them,
(33:40):
rather than make them read andwatch hours and days of videos
them, rather than make them readand watch hours and days of
videos.
There's plenty of opportunitythere.
And then regular touch pointsyou know, set alerts, set.
You can now there's a chat GPTthat will do repetitive tasks
and go back and, you know,update things and you can have a
(34:00):
much richer experience ifyou're doing your touch points
correctly.
And then the advocacy pieceright Survey your people as soon
as they get hired that's thetime they're the happiest and
ask them how they felt about theexperience and then you know if
they have a rich answer for you.
(34:22):
Give them a button, ask them ifthey want to write something
about that on LinkedIn andpossibly even Glassdoor or
Indeed reviews, and you know,you can change the trajectory of
your review status very quicklyby doing something like this.
And all of this could be easilyautomated with with AI.
Speaker 2 (34:41):
Yeah, I think that.
I think that's great and I'llsay, you know, retention,
attrition, you know, are two ofthe not two of the two highest
measurements that TA leaders arepaying attention to, you know,
when we look at our TA trendstudy that we just published
earlier this year.
So the good news is I think,hey, if that's a gap that you
see or an opportunity you see, Ithink TA leaders are starting
(35:03):
to recognize it too, or at leastyou know the people that are,
we like to think the smartestpeople are paying attention to
our surveys and insights.
So nice to hear that resonate,craig.
Speaker 3 (35:12):
Yeah, can I plug one more thing real quick?
Yeah, of course.
My next book is called Paintyour Store, and a lot of the
concepts we talked about todayare in Paint your Store.
But it's, whereas hiring humansis for employers, hiring
managers, recruiters, paint yourStore is for the job seeker or
(35:35):
the person looking for theirnext career move, or the
salesperson looking to gathermore clients, or the
organization looking to bring onmore customers clients or the
organization looking to bring onmore customers.
If you're a local independentconsultant and you want to just
attract more gigs, this book'sfor you.
It comes out March 25th and itwill be on Amazon.
(35:56):
How to Network your Way to yourNext Career Move Paint your
Store by Craig Fisher.
Speaker 2 (36:01):
That's awesome.
Well, this podcast is going tocome out on the 25th, so it's
perfect timing.
It's like we almost planned itthat way.
Craig, I'll ask the easiestquestion of all.
We'll certainly link everythingin the show notes.
Where else can listeners learna little bit more about you and
your work online?
Speaker 3 (36:17):
Yeah, the easiest place is hiring-humanscom it's
kind of a catch-all for my booksand my work, and dash humanscom
it's kind of a catch all for mybooks and my work, and it's at
the talent net live website, uh,and you can find out all about
great ways to engage with me andI'm also easily searchable.
So find me on LinkedIn andabsolutely connect.
Speaker 2 (36:36):
Yeah, and I love the uh fish dogs.
Uh to you know, marty and I alot of us, especially over at
G's day all big fans of dogs.
So, uh, anyone else can checkout that story in your free time
.
On how Craig came up with thatwebsite domain too, I love it,
thank you.
Awesome.
Well, thanks, Craig.
All right, Thanks for tuning in.
As always, head on over tochangestateio or shoot us a note
(37:00):
on all the social media.
We'd love to hear from you andwe'll check you guys next week.
Popular Podcasts
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!