Jason Heilman, Senior Vice President of Product, Automation, and AI at Bullhorn, and our host James Mackey, offer a masterclass in AI and Automation for recruiting organizations. Drawing from Jason’s experience as co-founder of Herefish (acquired by Bullhorn in 2019), Jason reveals how automation serves as the essential foundation for effective AI implementation in recruiting.
While many companies experiment with AI tools, Jason emphasizes focusing on core recruiter functions - particularly candidate matching and screening. These high-leverage activities not only boost efficiency but also create measurable competitive advantages. Bullhorn's data shows a striking 36% productivity increase for recruiters using automation tools compared to those who don't.
Bullhorn report: HERE
Thank you to our sponsor, SecureVision, for making this show possible!
Our host James Mackey
Follow us:
https://www.linkedin.com/company/82436841/
#1 Rated Embedded Recruitment Firm on G2!
https://www.g2.com/products/securevision/reviews
Thanks for listening!
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Hello, welcome to the Breakthrough Hiring Show.
I'm your host, James Mackey.
We got Jason Heilman with ustoday.
He is the Senior Vice Presidentof Product and Automation and
AI over at Bullhorn.
Jason, thanks for joining ustoday.
Speaker 2 (00:12):
Hey, james, excited about it, my pleasure.
Speaker 1 (00:14):
Yeah, really excited to have you on the show.
Just given your background andwhat you're doing right now, I
think a lot of people are goingto be very interested in
learning more about how Bullhornis approaching automation and
AI, just as an industry leaderand a company that everyone is
very familiar with, bothin-house teams and staffing
agencies.
And, of course, bullhorn hasbeen an incredibly successful
(00:35):
company, arguably one of themost successful, if not the most
successful, applicant trackingsystem and, I'm sure at this
point, a much larger productsuite than down the pipe
applicant tracking.
So, before we jump into some ofthe interesting use cases
related to AI and how Bullhornis approaching it, it would be
great for our audience to learna little bit more about you and
(00:57):
understand your background.
Speaker 2 (00:58):
Yeah, absolutely, and yes, we definitely excited to
offer the whole recruitingsolution space through Bullhorn,
and that's very much tied to mybackground.
So I've been in the industryalmost 20 years now, starting
off at another competitiveapplicant tracking system and
been working just in staffingfor this whole time.
But the probably most recentand relevant experience I was
(01:22):
the founder co-founder of acompany by the name of Hearfish,
which is essentiallyrecruitment automation.
It does automation in thestaffing industry.
We started that in 2014.
The idea there is very closelyrelated to where we are today,
in 2025, which is to say that instaffing, in any type of hiring
(01:44):
, you get a massive amount ofinterest and only so much time
to dedicate to responding tothat interest by human.
So let's just take the mostobvious example Whenever you
open a job and you get 400, 500,1,000 applicants, there's no
world where every one of thoseapplicants are being looked at
by a human.
We all must be honest with this.
But there definitely is, andalways should have been, a world
(02:07):
where those applicants at leastget a response back from a
system letting them know thatyou've gotten their message or
you've gotten their interest andcan respond back about if
they're not going to be a goodfit and let them move on with
their day.
So that's really where westarted is.
We knew staffing firms neededthe ability to quickly
communicate and just betterengage with their applicant pool
(02:30):
and their clients.
So that's really what BullhornAutomation now formerly HearFish
was all about candidates andtheir customers by automating
kind of some of the lower valuetouch points that, for a
staffing firm, are low value,but for the candidates and
contacts, it's very importantthat they hear from them.
(02:51):
Bullhorn acquired HearFish in2019, really built on the
baseline of continuing to dowhat we'd always done, which is
help people to create a betterexperience, but also knowing
that, as the world wascontinuing to evolve, every
staffing firm is going to needto add automation, and where
that's led us to today is thatautomation is really the
(03:12):
foundation that's necessary totake full advantage of all of
the AI capabilities that arehappening today.
Which takes us to where we aretoday, where, at Bullhorn, I
oversee our automation and AIinitiatives, and that's very
intentionally very closely heldtogether, because you can have
automation without AI and stillhave a great business even as we
continue aggressively down thisAI driven future.
(03:35):
But you can't really have AIwithout automation.
You need control, you need tobe able to scale those solutions
.
So that's what we're doing.
We're working on a lot ofexciting stuff, but I want to
turn it over to you for a littlewhile, James, because I've been
monologuing.
Speaker 1 (03:46):
Oh no, it's a really helpful introduction.
So thanks, and just so Iunderstand the lay of the land
with Bullhorn right now, are youstill primarily working with
staffing agencies or are youbuilding and breaking into other
industries as well?
Are you supporting in-houseteams at this?
Speaker 2 (04:02):
point.
No, we are very much focused ononly staffing, with one minor
exception, which is we recentlyacquired a company by the name
of TexKernel.
Probably just about anyone wholistens to this show uses one of
their solutions, likely throughparsing, so they're the global
leader in parsing.
Just about every resume thatgoes into an applicant tracking
(04:24):
system runs through a textkernel solution, so a number of
our partners on the corporateATS side use that solution.
So, with that isolated, uniquepart of the business, everywhere
else it's staffing and that'svery relevant because one of the
areas where AI is always themost valuable is whenever you've
got a unique and very largedata set and that kind of helps
(04:47):
you.
That's why you mentioned howBullhorn was large and we serve
a lot of staffing firms.
The reason that's good forpeople is that allows us to
deliver solutions on kind ofthis unique data set that's very
specific to staffing and allabout how the process works and
when we see success, how we canuse AI to replicate that.
Speaker 1 (05:07):
Yeah, absolutely.
And so for staffing companies,where do you see the highest
leverage opportunities forautomation and AI right now?
Speaker 2 (05:14):
Yes, absolutely what we did.
So you know everyone wheneverthe chat GPT moment, right when
it came out and everybody was soexcited about everything that
it can do, we share thatexcitement, of course, but we
all went through this kind oflearning process of everybody
very quickly brought out toolsthat helped you to write emails
better, helped to summarizeresumes, helped to do some of
(05:36):
the things that we would callaround the edges that a
recruiter does on a regularbasis and help them do the
things that they do a littlefaster, maybe a little better
Although now, looking back atGPT 3.5, I don't know if it was
that much better.
But where we've evolved to withour customers is, rather than
working around the edges, wereally want to go to the heart
(05:56):
of what recruiters do all day,every day that they maybe don't
need to anymore.
So kind of the biggest twoareas.
There's a bunch of nuance inactually delivering it, but the
biggest two areas are a jobopens I need to find the best
candidates right, like in mydatabase, or externally.
So that was the first big areawhere we're applying a lot of AI
(06:17):
.
And again, this isn't chat GPTstyle AI, right, you don't.
You can't have a job and domatching against a database of
millions of people with an LLM.
So this is a whole differentclass of AI which we probably
don't need to go into, butthat's a really important one
and I would say over the 20years that I've been working in
staffing, that was the first,second, third and fourth and
(06:39):
fifth reason, or the thing thatpeople wanted from AI Helped me
to better find better candidatesmore quickly.
This has been a problem that'sexisted for 20 years and AI was
always the potential solution,but I think we've really gotten
there now.
Some of that is through some ofthe recent advancements.
It's also again I described itlike we've got this large data
pool of all of our customers somuch staffing, specific
(07:00):
information.
So that's one big area ishelping recruiters to find the
best people quickly, stillgiving them the control if they
want it, or allowing that towork fully autonomously.
So, as soon as a job opens,finding the best people, sending
messages or letting therecruiter click a button to see
it.
That's one thing, and then thenext really big part of it is
(07:21):
all around screening.
Earlier, I described how youget 400 applicants, 500
applicants, whatever it might be.
Today, everyone could probablysee that actually those numbers
are way up because the jobboards have done a really good
job of making it very easy forcandidates to express interest
in a role and apply to a role.
That does not make a staffingfirm's job any easier.
That makes it much harder.
(07:41):
I read an article the other dayon Reddit.
Someone was talking about howthey applied to 13 or 1500 jobs
and their point was how theyonly got five interviews or
whatever it was.
The message they were trying totell in their story was how
little feedback they got.
What I read from that is thatperson applied to 1500 jobs in
(08:02):
two weeks.
How interested were they in anyof those jobs?
How much time could you havespent actually evaluating them?
Our customers need tools to helpevaluate the people who can
very quickly express interest.
So that's.
The other big part is screeningand, james, I know you can
speak to this too.
When a candidate comes into thedatabase, ai is a great
solution where they can callthem and have an actual voice
(08:25):
conversation with someone tojust validate that they've got
some of the specific keyrequirements whether that's
location, pay, whatever thatmight be but then also have an
open-ended discussion with themabout the job and about their
experience.
So I think we could probablyspend a lot of time talking
about why that's valuable, butthat's the other big area is
helping to screen out the verylarge pool of potential talent
(08:50):
and only surfacing the mostqualified and most interested to
the recruiting teams, again,just to allow them to focus more
where they're better.
Speaker 1 (08:59):
So really attacking the top of funnel, where it's
the most leverage and timeconsuming, and allowing
recruiters to essentiallyoperate more so down funnel as
much as possible?
Yeah, that's, of course, likethe use cases surrounding top of
funnel is something that I'vebeen really dialing in and
focusing on for for quite awhile, both with June and then
our guests that we're bringingon the show lately, and so I'd
(09:21):
love to dial in a little bitmore on some of those use cases.
Initially, you're saying wewere talking about what are.
The constant problem is findingpeople quickly, the right
people, good candidates, and soI would love to learn a little
bit more about the specificproducts and we can talk about
evaluation and screening in aminute, but more so, what other
products is Bullhorn offering tohelp companies, staffing firms,
(09:44):
identify talent faster?
I don't know if there's likefrom a sourcing standpoint, or
is it like inbound, or are wetalking candidate rediscovery?
Speaker 2 (09:53):
possibly it's a great call.
Yeah, it's all the above.
I think we can probably prettyquickly break it into three,
maybe four buckets.
Let's talk about the obviousones first.
The first is a new applicantapplies.
They go into the system howeverthat comes in.
Then we've got solutions whichallow teams to automatically or
more easily search externalsites and their own right.
(10:16):
I can, from a single interface,search all the different job
boards in addition to my owndatabase, and when we do that we
can also set up kind of agentsin the background to say, okay,
I regularly get projectmanagement roles, so I don't
want to have to go look all thetime for project managers.
I just want it to automaticallyrun searches and either alert
(10:37):
me whenever it finds goodmatches so you're the first one
as soon as they pop on the jobboard or just add them to my
system, just add them right in.
So I think that's what was that.
That's buckets one and two.
Then I guess the third would beautomating the process of just
staying engaged, and this wasone of the big areas that we
focused on in Herefish withautomation is just so often we
(11:00):
talk to our customers they'revery proud of the large database
that they built over time butthen whenever you go look at the
database it's neglected.
It's great that you have amillion candidates, but if you
haven't talked to any of them inthe last two years, they really
have a relationship with you.
So just continuing the processof nurturing the talent pool and
actually building these talentpools and not just trying to hit
(11:23):
them for jobs, so that'sanother big solution.
Let's just take a traditional.
Let's just say you just send amonthly newsletter, right, like
just staying engaged with yourpool and obviously now there's
more interesting things you cando than just a newsletter but
just staying engaged.
So that helps customers toreally craft their own talent
pool that is unique to the jobboards or unique to what other
(11:45):
people see in applicants.
So those are big.
I think buckets I guess it wasthree, that middle one about
internal and external search,maybe you could bucket into two.
So I think those are the bigareas.
Speaker 1 (11:55):
Yeah, for sure, and I'm just curious so the staffing
companies that you wouldconsider more so likely to be
early adopters of this type oftechnology Are there any market
segments that you're reallystanding out to?
For instance, is it SMB,mid-market enterprise?
Is it staffing agencies thatspecialize in certain industries
(12:16):
?
What are you seeing out there?
It's funny.
Speaker 2 (12:19):
I don't have an answer like that, because
everyone is obviously dedicatinga lot of resources to AI, right
?
Everyone sees this opportunityIn our largest customer base.
We've got some of the globalpublic staffing companies who
have dedicated a large amount ofresources to making this happen
, all the way down to.
We've got some of the mostinnovative companies that have
(12:40):
five internal employees thenumber one difference and, by
the way, and also acrossindustry light industrial,
healthcare, professional,commercial, however you want to
frame it all of which are seeingsuccess.
The differentiator, though, isit's really it's two things, to
be frank.
It's, right now, having anexecutive owner that has the
(13:01):
ability to to actually put thesechanges in place.
That's from looking up andmaking sure that they've got
support from their board and therest of the C-suite, and also
looking down that they'veactually got the ability within
their organization to do changemanagement, because it's a big,
has the potential to be a bigchange.
There's ways to do this in whatfeels like an evolutionary way,
(13:24):
but companies that aren't greatat rolling out new solutions.
This is nothing new about that.
This is a new piece oftechnology that your recruiting
teams should be on board with,unless you're going fully
recruitalist, which is likeanother discussion which I think
we should.
We can talk about, but let'spause that.
Let's just say that this is aworld where you do intend to
(13:46):
take an existing team and reallygrow their capabilities, rather
than completely displacing it.
So just being able to roll itout and then, if the other one
is that executive or thatexecutive's team, the ones that
actually are deeply involved inthe process, so not only having
the buy-in and getting to bothdirections, but just actively in
(14:07):
with a team and having a teamthat can iterate, because the
idea with these is they willwork.
It has been proven.
These AI solutions 100% work instaffing, but they may not work
out of the box.
You may need to go in anditerate on it and adjust your
messaging to improve yourconversion rate of people
actually taking screens, adjustyour matching criteria to make
(14:28):
sure that it's either tighter orwider.
So there's a lot of differentpotential ways that you can A
make this your own, but, bactually achieve the success
that we want.
So that makeup it's reallyabout the humans is the recipe
for success.
It's not about the industry.
It's really about having theright people and the right team
that's dedicated to seeingsuccess.
Speaker 1 (14:50):
Yeah, for sure, and we actually.
We recently had Hugo Milan onthe show.
He's the president over atKelly Services.
He runs their entireengineering division.
I think it's like a $400million division for Kelly and
we spent a lot of time talkingabout just the state of the
staffing industry and he wastalking about essentially a
contraction right in staffaugmentation, services primarily
(15:12):
, and he had a prettyinteresting take on why this is
happening.
By the way, for everybodylistening and Jason, if you're
interested, it's, I think,probably the last episode we
actually released.
It was pretty cool.
So one thing that just came topof mind is what are your
thoughts on the correlationbetween companies adopting AI
solutions and staffing and therelationship between that and
(15:36):
the staffing industry being inessentially a recession right
now?
I could see it going either way, where it's like, okay, we have
to be more efficient, right,we're already service like
staffing companies are alreadyservices companies with tight
margins.
They're not software companies,right, so they have to be
mindful of that.
So you could see, okay, they'regoing to be driving more toward
(15:57):
this type of technology,probably even faster.
Or you could say they havebudget budgetary constraints,
right, they're not reallythinking about adding on
additional software.
What correlation, if any, areyou seeing between those two
things?
Speaker 2 (16:09):
Yes.
Okay, let's first talk about, Ithink, the market as a whole
and the contraction in thecurrent state.
So I think, like last year, siaput out their study.
I think across the industrythere's pockets where it's
different, but essentially thestaffing industry took something
like around a 10% compressionPretty brutal right.
Yeah, not great percentcompression Pretty brutal, right
(16:33):
.
I guess.
One question to answer flat outis how much of that is due to AI
from their end customers.
I would say that's roughly zero.
Like I do not think.
Yeah, I don't think.
Yet at least People that usestaffing, they're not using
staffing because they've gotsuch great AI tools, because
most of those, you know that.
I don't think.
That's it.
So then we go to, okay, thestaffing suppliers how or, I
guess, the suppliers, thestaffing companies, how are they
(16:54):
implementing and what successare they seeing?
And I think what we are seeingis there are still companies
that are growing.
They're taking share from thecompanies that aren't advancing
as quickly.
It's pretty flat out, Like wecan see, for example, we have a
pretty explicit study that we'veshared widely.
Like customers that useautomation here at Fish Bullhorn
(17:14):
Automation and other automationsolutions not just ours that we
offer, they're 36% moreproductive than those that don't
.
So each recruiter half thecompany using it hires 36% more
candidates per recruiter.
So that's how we gauge theproductivity placements per
recruiter Pretty simple metriclike that doesn't take into
account margin or anything likethat.
(17:35):
So that's our first step.
So those companies that 36%wasn't due to a growth in the
industry, that 36% came fromwhere they beat other suppliers
to their mutual customers.
So that's what we're reallyseeing is like it's a little bit
of a story, a tale of twocities where those that are
(17:55):
aggressive and implementingsolutions even if it isn't the
latest AI solution they're theones who are still growing.
We've got a large contingent ofour customer base who is growing
.
They are adding seeds.
We are.
They've got their financesrunning through our system.
They're making more money.
Their margins at the top lineare increasing.
You can see it all the way downto the very bottom line.
(18:17):
So that's what we're seeing.
It will be a market of winnersand losers and there'll be a
middle period, and I don't knowif we want to go one step
farther, but where thisnaturally will go, if we look
backwards at previous cycles, isthese people that are winning
to come in and start to takebusiness by offering lower
(18:50):
markups and lower rates to thecustomer.
So then that becomes adouble-edged sword where not
only are those customers whoaren't able to deliver at the
lower cost profile are losingthe business they already have,
they're going to start to loselogos to those that can supply
at a lower rate, and then thewhole industry.
This is natural pricecompression.
(19:10):
We hope it doesn't happen, butit feels like that's probably
going to happen when ourcustomer, when the buyers of
staffing services, start to seethat there's enough suppliers
that can offer a lower costbasis.
Speaker 1 (19:22):
Because of automation and AI.
Essentially, the recruiters aremore productive and so they're
able to undercut.
That's right, and just due tostaffing, industry is just so
insanely competitive, right,there's just so many out there.
Speaker 2 (19:33):
And so think, if you're a 20-person staffing firm
, that's very motivated, youwant to grow and you have any
deal where you might lose it bydropping a margin percentage,
you're going to take that dealbecause your net at the end of
the day is still high enough tosupport that business, to
deliver at a lower rate.
And then, especially, everycompany has it where you've also
(19:57):
got the already incredibly lowmargin business and then so much
business that you aren'tservicing today Our best in
class customers maybe have callit a 25 or 30% fill rate.
That's 70% of the roles thatyou aren't hiring and it's
likely, because you don't have,you can't effectively deliver to
those customers, given yourcurrent cost profile.
(20:17):
Maybe you work it, but notreally you don't put your best
on it.
Maybe you outsource it.
That business can now befurther driven down your own
cost basis.
I don't know.
I was going to try to keep avery cohesive answer, I feel
like I've gone a little far.
Speaker 1 (20:31):
All the extra context is it's really helpful?
Yeah, it's.
It's really interesting tothink about how it's going to
impact the competitive landscapeof staffing and I think you you
made a really good point, right, it's the in order to grow.
Particularly, it's always it'shard to grow any staffing
company and essentially anymarket because there's so much
competition, right.
But beyond that, now whenyou're looking at a market
(20:51):
contraction, right, it probablywill force early adoption by
agencies if they do want tocompete, and I think there's a
lot of companies and staffingthat are really old school and
probably will not change andthey're probably going to get
left behind, right.
Speaker 2 (21:06):
Yeah, it's going to be tough for them.
I will say though I will hedgeit a little bit with this isn't
happening tomorrow.
So to throw a little bit ofmaybe numbers into the context,
we recently did a survey of alarge our grid survey and 52% of
our customers are evaluatingand testing AI solutions.
Only 15% have purchased andimplemented and seeing benefits
(21:30):
from AI solutions, so prettyearly.
And, by the way, that 15%source and match.
So like the ability toautomatically run searches and
find candidates.
Of that 15%, that was like 70%of it.
So it's also not even all thisnew stuff like the screening,
like the hyper-personalizedmessaging.
It's things that have beenaround for a little bit longer.
(21:52):
So this isn't gonna happentomorrow, but there is gonna be.
Let's pretend like it's two,three, five years, whatever that
window is, before the seriouscompression starts.
Some people are gonna make haylike there's gonna be some
companies that grow like crazyor reap massive margin benefits
during that time.
But I don't know In my mind,given the previous technological
(22:14):
changes and all the humansinvolved in staffing, this isn't
going to happen overnight.
I don't know if maybe fiveyears is overnight, but it's not
like next year.
It's not like this year where,all of a sudden, the whole
market dynamics shift.
Speaker 1 (22:25):
Yeah, it's going to take time.
I think you briefly touched onthis earlier.
It's the first couple of yearspeople after ChatGPT came out,
people are just essentiallyexperimenting, trying to figure
out which use cases essentiallyare actually creating real value
versus maybe not quite as much.
I think sometimes some of thetools or products that have been
created were really justshifting where you're doing the
(22:45):
work.
I don't know how much they wereactually driving productivity
per se, but I think, like we're,you know at least folks that
are leading the charge right,like people such as yourself.
At this point we have a prettygood pulse on what's driving
real productivity and the usecases, that customers are
becoming a little bit moreeducated right and starting to
really figure out where they can, where they could, drive
(23:07):
outcomes.
You shared that stat about 50%of staffing companies, I think
you said experimenting with AI,and then 15% actually driving
value or having found itimplemented, a use case that's
really like just they're payingfor it and rolled it out
enterprise Okay.
I got it I was reading in theWall Street Journal is something
like overall in the US economy,61% companies are experimenting
(23:28):
with AI, but I think it waslike less than 6% have actually
rolled out an application that'sgenerating like real ROI,
tangible value for theorganization, which is pretty
wild.
It's a very small, so there'sstill a lot of education in the
market.
That needs to happen too.
Yeah, it's going to take time.
Speaker 2 (23:44):
I want to pull one thread from what you just
described about how, when peopleadd these solutions and it just
moves the work, tie that backto earlier, what we talked about
how, where we're seeing success, and it's all about the
leadership in the actualcompanies that are driving it
One of the number one thingsthat must be done.
So let's take the old worldversus the new world.
So let's pretend like arecruiter has 10 actual live
(24:05):
conversations a day.
Hopefully it's more than that,but let's pretend like it's 10.
Today those conversations areat least half of them are
talking to people that aren'tqualified, that we just throw in
the trash, right, we re-engagethem and we find out right, but
we don't use them for that dayor that month.
And then we talk to a few thatare okay and then two that are
awesome.
Now imagine the world wheneverevery candidate you talk to has
(24:26):
been pre-screened, already foundtime on your calendar.
So let's pretend like theystill only have 10 conversations
.
Every one of those should besomeone that's a good candidate
and qualified, but not all forone job.
So in the old world, I'mtalking to 10 people to fill one
role.
In the new world, I should betalking to 10 people to fill
three or four roles.
Speaker 1 (24:43):
You need fewer candidates, because they're
already qualified and good.
Speaker 2 (24:47):
So then, what does that mean?
So I, as a staffing firm, inorder to achieve that outcome,
had to spend money to invest insolutions to get that.
You're investing so you cannothave the recruiter's metrics
remain the same.
You can't still expect them tobuild the same margin, because
it's going to be way If you makeit so that they have to have
the same criteria to make acomfortable lifestyle.
(25:07):
Sure, there's going to be somethat go above and beyond and
absolutely crush it and make amillion dollars a year or
whatever, but the norm isthey're going to just go back to
a comfortable place and let'spretend like you need to add
five, three new starts everymonth.
Even they have all these tools,they're just going to keep
adding three new starts, unlessyou expect them to add six.
So that's the other part.
To drive change and to actuallydrive it, you need to have
(25:29):
confidence in actually thatthese tools will work and change
their outcomes and change themetrics that drive it, and
that's a hard change Recruitersdo not want to hear, yeah, your
requirements are doubling.
To make the same amount of money, you have to have a lot of
proof points.
There's a lot of changemanagement that goes into it.
So that's the part of thetransition.
That's where it can get bumpyand where another area that we
(25:50):
have to watch to make sure youget the outcomes you want.
Speaker 1 (25:53):
Yeah, for sure you should be evaluating and
tracking exactly how much thisis truly impacting revenue.
I'm curious so your customers,are they primarily looking at
these technologies more from aboosting revenue perspective, or
efficiency or margin bottomline?
What is driving them to thepurchasing decision?
(26:15):
That's an excellent question.
Speaker 2 (26:17):
So the short answer is that we think most want to
ultimately increase the top line, drive more revenue, expand
their business.
Right now, though, given themarket conditions and everyone's
internal cost cutting thatthey're forced to go through,
sometimes they go into itlooking at it for that as an
(26:38):
opportunity.
So it's a little bit of a blend.
There's a little bit ofshort-term pain that people are
feeling right now that maybeforces them to look a little bit
more inward and look at waysthey can cut costs, but
ultimately, everyone knowsthey're actually investing in
these solutions to more rapidlygrow and better serve their
customers, and that's a reallybig part of it.
Even if you aren't using thesetools to add new logos, wouldn't
(26:58):
it be nice if you were actuallyjust making your customers that
already exist today muchhappier with you being the
platinum supplier, right Likegetting all the awards and value
that accrues from just doing abetter job, but with the jobs
you've already got today?
Speaker 1 (27:12):
Yeah, for sure.
Look, Jason, we're coming up ontime here and I wanted to see
if there are any other topics oranything else you want to
mention about what you're doingat Bullhorn that you want to
share with the audience.
Speaker 2 (27:21):
I don't think so.
The only other thing that Ithink is always very important
we just talked about how we canmake customers happy by
backfilling.
The only other thing that Ithink is important I think you
can speak to as well is justwhat about the candidate's
experience?
Is this better, worse, the same?
We've got data that talks abouthow candidates are very much
ready to adopt AI.
(27:42):
We don't live in this utopiawhere every candidate gets
responded to.
If we did, AI might make it alittle more rocky, but we know
today they want faster responsetimes, they want to get to work
faster, and AI can actually helpthem to achieve those.
So I guess that's the otherinteresting part is, it's not
about putting your work on thecandidate and sacrificing that.
(28:03):
It's actually this helps you todeliver a better candidate
experience.
Speaker 1 (28:06):
Yeah, it's a much better candidate experience and
it's honestly speaking withfolks even in staffing.
A lot of people do think it'sgoing to hurt candidate
experience and it surprises me alittle bit because it's like
getting into the evaluationtools that we're both building
pre-screening conversationsFolks come inbound to provide
recruiters with qualifiedcandidate shortlists and speak
(28:28):
to folks that are actuallyqualified.
A lot of people are worriedthat candidates are not going to
want to engage with thistechnology, but to me it's like
it's so valuable to you apply toa job and you immediately get a
call or a text to engage you inthe interview process.
I'm in the headspace, I justapplied to the job, I'm in front
of my phone, I got my phone onme and my computer.
(28:49):
I'm available essentially andyou can move a lot faster.
But the other thing that'scurious is and it does make
sense to be concerned about thiswe need to be very careful.
But the ideas of AI and bias,ai discrimination and I think
generally there's just aneducation that needs to occur on
how generative AI actuallyworks versus AI models.
Potentially in the past thatmore so operated in a black box
(29:12):
where it was harder to see whyit was making certain evaluation
decisions, but I see the toolscoming out right now.
Certain evaluation decisions,but I see the tools coming out
right now, like the ones we'rebuilding, as something to
actually create less biasprocesses.
I think people are a big issuewhen it comes to bias right, so
I think these technologiesactually can level the playing
(29:33):
field and create a much betterand more fair and equitable
process for people.
Speaker 2 (29:38):
Absolutely.
It's very hard to change howmillions of people go about
their Absolutely.
It's very hard to change howmillions of people go about
their day.
It's very easy to tell acomputer how to operate.
Even the challenges thateveryone's heard in the news
stories in the past, it'sbecause the way those models
were built is.
They replicate previous humanbehaviors, so all that they're
doing is continuing to push outmore and yeah, and then the
(29:59):
other part.
I think that's really important.
We talk about the candidateexperience is, today all the
decisions are made based on apiece of paper and everyone
knows on a resume, right?
Everyone knows a resume cannever reflect A you as a person
more holistically or B yourspecific qualifications for a
given role.
These screening solutions justallow them to tell more of the
(30:23):
story, talk about why they're agood fit, and I'm sure you've
seen this.
There's been very many timeswhen someone's been a relatively
low resume fit but then havegone through a screen and
they've been a perfect match forthat role, because the maids
are not effective at actuallytelling a story about how good
of a fit you'll be for a role.
Speaker 1 (30:40):
And particularly, I think, maybe some industries
more than others too.
Folks may not really have greatresumes, so it's just having a
level of clarity you could getfrom a screen is, I think, a lot
more valuable.
So we're definitely on the samepage there.
Yeah, for sure.
Speaker 2 (30:54):
And again, just maybe .
One very last thing.
We've got some stats like thosescreening ages.
Again, is it weird to talk toan AI person?
I don't know, Maybe, but it'sgoing to be the new future and
it isn't.
If you've ever done one of these, Maybe in this thing I can put
a link to your screener orsomebody's screener so people
can experience these.
They're actually really goodand it's funny when people are
having these discussions theytalk like they're talking to a
(31:16):
person.
They're like, oh, that's a goodquestion, or sorry, I didn't
hear that, Can you repeat it?
And this is the human talkingto the AI.
And then whenever I'm sure youdo this we do surveys after
every single screen to say howtheir experience was, and
they're overwhelmingly positive82% give an 8, 9, or a 10.
And I think this is important,but that's 82%.
(31:36):
There are some that give a oneand a two and it's actually
pretty polarizing.
The ones that give the lowscores.
It's usually.
I don't think AI should be doinghiring.
I don't want to have my future.
They've got very opinionated,dogmatic concerns and I want to
bring that up.
Whether that's valid or not,every customer that implements
one of these is going to have arecruiter that tells them a
story about a candidate whoreached out with this opinion.
(31:59):
But we just have to weigh inthe counterbalance of that's a
very small subset that's 2%Whereas the vast majority really
actually like the experience.
So there are going to be somethat don't love it, but that's
anything in life.
Speaker 1 (32:12):
What it reminds me of is back in the day, when
there's a lot of industries,right Even like when it comes to
buying stocks where you used tohave to essentially speak with
a salesperson early in theprocess, and remember when a lot
of that went online and becameautomated, a lot of stockbrokers
and people would say, oh, theyneed to speak to us, right,
they're not going to beinterested in working through
(32:32):
automation.
Think about just how manyindustries have made that shift,
have made that shift.
And it's the same thing.
Like, people don't necessarilywant to speak with somebody
early on.
If they could just do thescreen and get a little down
funnel, let's see if it'sactually worth the time.
Scheduling an interview andthen also these screen products
are also good for candidates tobe able to ask questions too and
(32:53):
learn more about theopportunity.
So the idea that people want tospeak with people early on,
it's not necessarily people wantto speak with people early on.
It's not necessarily they wantto speak with somebody if they
know it's a relevant opportunity.
Nobody wants to waste theirtime and you get scheduling and
rescheduling and it's just, butthat's also making the
assumption that ever has thatthat's happening today.
Like that is the difference.
Speaker 2 (33:15):
Like the stockbrokers , had a very vested interest in
calling a bunch of people andtrying to sell them.
But we don't have.
Let's pretend like only 100people apply to a job.
95% of them are never going tohave a conversation.
They don't even get the chanceto have a conversation.
Speaker 1 (33:24):
Yeah.
So it's yes it levels theplaying field and you're right,
it gives folks the opportunityto talk about their skill set
that might otherwise have notbeen reached.
So I think people are going tocome around and it's from a
product perspective.
It's important for us to bebuilding products of the future
and clearly things are going togo in this direction and people
are going to become morecomfortable with it over time.
(33:45):
There's always going to bepeople that, regardless what you
do to the hiring process,they're not going to like it.
People don't like it if youdon't like getting turned down
for jobs.
So there's always going to besome negative feedback,
regardless if it's people or AIor anything else.
Yeah, yeah, going to be somenegative feedback, regardless if
it's people or AI or anythingelse.
Yeah, yeah, jason, thank you somuch for joining us today.
This has been a lot of fun.
You've shared a lot of greatinsights with our audience today
(34:06):
, so I know everybody will havelearned a lot.
I really appreciate yourcontribution to the show.
Speaker 2 (34:11):
Absolutely.
Thank you so much for having meand excited to talk again.
Speaker 1 (34:14):
Yeah, absolutely.
You're welcome to come back onanytime.
Speaker 2 (34:16):
Nice.
Hello, welcome to the Breakthrough Hiring Show.
I'm your host, James Mackey.
We got Jason Heilman with ustoday.
He is the Senior Vice Presidentof Product and Automation and
AI over at Bullhorn.
Jason, thanks for joining ustoday.
Speaker 2 (00:12):
Hey, james, excited about it, my pleasure.
Speaker 1 (00:14):
Yeah, really excited to have you on the show.
Just given your background andwhat you're doing right now, I
think a lot of people are goingto be very interested in
learning more about how Bullhornis approaching automation and
AI, just as an industry leaderand a company that everyone is
very familiar with, bothin-house teams and staffing
agencies.
And, of course, bullhorn hasbeen an incredibly successful
(00:35):
company, arguably one of themost successful, if not the most
successful, applicant trackingsystem and, I'm sure at this
point, a much larger productsuite than down the pipe
applicant tracking.
So, before we jump into some ofthe interesting use cases
related to AI and how Bullhornis approaching it, it would be
great for our audience to learna little bit more about you and
(00:57):
understand your background.
Speaker 2 (00:58):
Yeah, absolutely, and yes, we definitely excited to
offer the whole recruitingsolution space through Bullhorn,
and that's very much tied to mybackground.
So I've been in the industryalmost 20 years now, starting
off at another competitiveapplicant tracking system and
been working just in staffingfor this whole time.
But the probably most recentand relevant experience I was
(01:22):
the founder co-founder of acompany by the name of Hearfish,
which is essentiallyrecruitment automation.
It does automation in thestaffing industry.
We started that in 2014.
The idea there is very closelyrelated to where we are today,
in 2025, which is to say that instaffing, in any type of hiring
(01:44):
, you get a massive amount ofinterest and only so much time
to dedicate to responding tothat interest by human.
So let's just take the mostobvious example Whenever you
open a job and you get 400, 500,1,000 applicants, there's no
world where every one of thoseapplicants are being looked at
by a human.
We all must be honest with this.
But there definitely is, andalways should have been, a world
(02:07):
where those applicants at leastget a response back from a
system letting them know thatyou've gotten their message or
you've gotten their interest andcan respond back about if
they're not going to be a goodfit and let them move on with
their day.
So that's really where westarted is.
We knew staffing firms neededthe ability to quickly
communicate and just betterengage with their applicant pool
(02:30):
and their clients.
So that's really what BullhornAutomation now formerly HearFish
was all about candidates andtheir customers by automating
kind of some of the lower valuetouch points that, for a
staffing firm, are low value,but for the candidates and
contacts, it's very importantthat they hear from them.
(02:51):
Bullhorn acquired HearFish in2019, really built on the
baseline of continuing to dowhat we'd always done, which is
help people to create a betterexperience, but also knowing
that, as the world wascontinuing to evolve, every
staffing firm is going to needto add automation, and where
that's led us to today is thatautomation is really the
(03:12):
foundation that's necessary totake full advantage of all of
the AI capabilities that arehappening today.
Which takes us to where we aretoday, where, at Bullhorn, I
oversee our automation and AIinitiatives, and that's very
intentionally very closely heldtogether, because you can have
automation without AI and stillhave a great business even as we
continue aggressively down thisAI driven future.
(03:35):
But you can't really have AIwithout automation.
You need control, you need tobe able to scale those solutions
.
So that's what we're doing.
We're working on a lot ofexciting stuff, but I want to
turn it over to you for a littlewhile, James, because I've been
monologuing.
Speaker 1 (03:46):
Oh no, it's a really helpful introduction.
So thanks, and just so Iunderstand the lay of the land
with Bullhorn right now, are youstill primarily working with
staffing agencies or are youbuilding and breaking into other
industries as well?
Are you supporting in-houseteams at this?
Speaker 2 (04:02):
point.
No, we are very much focused ononly staffing, with one minor
exception, which is we recentlyacquired a company by the name
of TexKernel.
Probably just about anyone wholistens to this show uses one of
their solutions, likely throughparsing, so they're the global
leader in parsing.
Just about every resume thatgoes into an applicant tracking
(04:24):
system runs through a textkernel solution, so a number of
our partners on the corporateATS side use that solution.
So, with that isolated, uniquepart of the business, everywhere
else it's staffing and that'svery relevant because one of the
areas where AI is always themost valuable is whenever you've
got a unique and very largedata set and that kind of helps
(04:47):
you.
That's why you mentioned howBullhorn was large and we serve
a lot of staffing firms.
The reason that's good forpeople is that allows us to
deliver solutions on kind ofthis unique data set that's very
specific to staffing and allabout how the process works and
when we see success, how we canuse AI to replicate that.
Speaker 1 (05:07):
Yeah, absolutely.
And so for staffing companies,where do you see the highest
leverage opportunities forautomation and AI right now?
Speaker 2 (05:14):
Yes, absolutely what we did.
So you know everyone wheneverthe chat GPT moment, right when
it came out and everybody was soexcited about everything that
it can do, we share thatexcitement, of course, but we
all went through this kind oflearning process of everybody
very quickly brought out toolsthat helped you to write emails
better, helped to summarizeresumes, helped to do some of
(05:36):
the things that we would callaround the edges that a
recruiter does on a regularbasis and help them do the
things that they do a littlefaster, maybe a little better
Although now, looking back atGPT 3.5, I don't know if it was
that much better.
But where we've evolved to withour customers is, rather than
working around the edges, wereally want to go to the heart
(05:56):
of what recruiters do all day,every day that they maybe don't
need to anymore.
So kind of the biggest twoareas.
There's a bunch of nuance inactually delivering it, but the
biggest two areas are a jobopens I need to find the best
candidates right, like in mydatabase, or externally.
So that was the first big areawhere we're applying a lot of AI
(06:17):
.
And again, this isn't chat GPTstyle AI, right, you don't.
You can't have a job and domatching against a database of
millions of people with an LLM.
So this is a whole differentclass of AI which we probably
don't need to go into, butthat's a really important one
and I would say over the 20years that I've been working in
staffing, that was the first,second, third and fourth and
(06:39):
fifth reason, or the thing thatpeople wanted from AI Helped me
to better find better candidatesmore quickly.
This has been a problem that'sexisted for 20 years and AI was
always the potential solution,but I think we've really gotten
there now.
Some of that is through some ofthe recent advancements.
It's also again I described itlike we've got this large data
pool of all of our customers somuch staffing, specific
(07:00):
information.
So that's one big area ishelping recruiters to find the
best people quickly, stillgiving them the control if they
want it, or allowing that towork fully autonomously.
So, as soon as a job opens,finding the best people, sending
messages or letting therecruiter click a button to see
it.
That's one thing, and then thenext really big part of it is
(07:21):
all around screening.
Earlier, I described how youget 400 applicants, 500
applicants, whatever it might be.
Today, everyone could probablysee that actually those numbers
are way up because the jobboards have done a really good
job of making it very easy forcandidates to express interest
in a role and apply to a role.
That does not make a staffingfirm's job any easier.
That makes it much harder.
(07:41):
I read an article the other dayon Reddit.
Someone was talking about howthey applied to 13 or 1500 jobs
and their point was how theyonly got five interviews or
whatever it was.
The message they were trying totell in their story was how
little feedback they got.
What I read from that is thatperson applied to 1500 jobs in
(08:02):
two weeks.
How interested were they in anyof those jobs?
How much time could you havespent actually evaluating them?
Our customers need tools to helpevaluate the people who can
very quickly express interest.
So that's.
The other big part is screeningand, james, I know you can
speak to this too.
When a candidate comes into thedatabase, ai is a great
solution where they can callthem and have an actual voice
(08:25):
conversation with someone tojust validate that they've got
some of the specific keyrequirements whether that's
location, pay, whatever thatmight be but then also have an
open-ended discussion with themabout the job and about their
experience.
So I think we could probablyspend a lot of time talking
about why that's valuable, butthat's the other big area is
helping to screen out the verylarge pool of potential talent
(08:50):
and only surfacing the mostqualified and most interested to
the recruiting teams, again,just to allow them to focus more
where they're better.
Speaker 1 (08:59):
So really attacking the top of funnel, where it's
the most leverage and timeconsuming, and allowing
recruiters to essentiallyoperate more so down funnel as
much as possible?
Yeah, that's, of course, likethe use cases surrounding top of
funnel is something that I'vebeen really dialing in and
focusing on for for quite awhile, both with June and then
our guests that we're bringingon the show lately, and so I'd
(09:21):
love to dial in a little bitmore on some of those use cases.
Initially, you're saying wewere talking about what are.
The constant problem is findingpeople quickly, the right
people, good candidates, and soI would love to learn a little
bit more about the specificproducts and we can talk about
evaluation and screening in aminute, but more so, what other
products is Bullhorn offering tohelp companies, staffing firms,
(09:44):
identify talent faster?
I don't know if there's likefrom a sourcing standpoint, or
is it like inbound, or are wetalking candidate rediscovery?
Speaker 2 (09:53):
possibly it's a great call.
Yeah, it's all the above.
I think we can probably prettyquickly break it into three,
maybe four buckets.
Let's talk about the obviousones first.
The first is a new applicantapplies.
They go into the system howeverthat comes in.
Then we've got solutions whichallow teams to automatically or
more easily search externalsites and their own right.
(10:16):
I can, from a single interface,search all the different job
boards in addition to my owndatabase, and when we do that we
can also set up kind of agentsin the background to say, okay,
I regularly get projectmanagement roles, so I don't
want to have to go look all thetime for project managers.
I just want it to automaticallyrun searches and either alert
(10:37):
me whenever it finds goodmatches so you're the first one
as soon as they pop on the jobboard or just add them to my
system, just add them right in.
So I think that's what was that.
That's buckets one and two.
Then I guess the third would beautomating the process of just
staying engaged, and this wasone of the big areas that we
focused on in Herefish withautomation is just so often we
(11:00):
talk to our customers they'revery proud of the large database
that they built over time butthen whenever you go look at the
database it's neglected.
It's great that you have amillion candidates, but if you
haven't talked to any of them inthe last two years, they really
have a relationship with you.
So just continuing the processof nurturing the talent pool and
actually building these talentpools and not just trying to hit
(11:23):
them for jobs, so that'sanother big solution.
Let's just take a traditional.
Let's just say you just send amonthly newsletter, right, like
just staying engaged with yourpool and obviously now there's
more interesting things you cando than just a newsletter but
just staying engaged.
So that helps customers toreally craft their own talent
pool that is unique to the jobboards or unique to what other
(11:45):
people see in applicants.
So those are big.
I think buckets I guess it wasthree, that middle one about
internal and external search,maybe you could bucket into two.
So I think those are the bigareas.
Speaker 1 (11:55):
Yeah, for sure, and I'm just curious so the staffing
companies that you wouldconsider more so likely to be
early adopters of this type oftechnology Are there any market
segments that you're reallystanding out to?
For instance, is it SMB,mid-market enterprise?
Is it staffing agencies thatspecialize in certain industries
(12:16):
?
What are you seeing out there?
It's funny.
Speaker 2 (12:19):
I don't have an answer like that, because
everyone is obviously dedicatinga lot of resources to AI, right
?
Everyone sees this opportunityIn our largest customer base.
We've got some of the globalpublic staffing companies who
have dedicated a large amount ofresources to making this happen
, all the way down to.
We've got some of the mostinnovative companies that have
(12:40):
five internal employees thenumber one difference and, by
the way, and also acrossindustry light industrial,
healthcare, professional,commercial, however you want to
frame it all of which are seeingsuccess.
The differentiator, though, isit's really it's two things, to
be frank.
It's, right now, having anexecutive owner that has the
(13:01):
ability to to actually put thesechanges in place.
That's from looking up andmaking sure that they've got
support from their board and therest of the C-suite, and also
looking down that they'veactually got the ability within
their organization to do changemanagement, because it's a big,
has the potential to be a bigchange.
There's ways to do this in whatfeels like an evolutionary way,
(13:24):
but companies that aren't greatat rolling out new solutions.
This is nothing new about that.
This is a new piece oftechnology that your recruiting
teams should be on board with,unless you're going fully
recruitalist, which is likeanother discussion which I think
we should.
We can talk about, but let'spause that.
Let's just say that this is aworld where you do intend to
(13:46):
take an existing team and reallygrow their capabilities, rather
than completely displacing it.
So just being able to roll itout and then, if the other one
is that executive or thatexecutive's team, the ones that
actually are deeply involved inthe process, so not only having
the buy-in and getting to bothdirections, but just actively in
(14:07):
with a team and having a teamthat can iterate, because the
idea with these is they willwork.
It has been proven.
These AI solutions 100% work instaffing, but they may not work
out of the box.
You may need to go in anditerate on it and adjust your
messaging to improve yourconversion rate of people
actually taking screens, adjustyour matching criteria to make
(14:28):
sure that it's either tighter orwider.
So there's a lot of differentpotential ways that you can A
make this your own, but, bactually achieve the success
that we want.
So that makeup it's reallyabout the humans is the recipe
for success.
It's not about the industry.
It's really about having theright people and the right team
that's dedicated to seeingsuccess.
Speaker 1 (14:50):
Yeah, for sure, and we actually.
We recently had Hugo Milan onthe show.
He's the president over atKelly Services.
He runs their entireengineering division.
I think it's like a $400million division for Kelly and
we spent a lot of time talkingabout just the state of the
staffing industry and he wastalking about essentially a
contraction right in staffaugmentation, services primarily
(15:12):
, and he had a prettyinteresting take on why this is
happening.
By the way, for everybodylistening and Jason, if you're
interested, it's, I think,probably the last episode we
actually released.
It was pretty cool.
So one thing that just came topof mind is what are your
thoughts on the correlationbetween companies adopting AI
solutions and staffing and therelationship between that and
(15:36):
the staffing industry being inessentially a recession right
now?
I could see it going either way, where it's like, okay, we have
to be more efficient, right,we're already service like
staffing companies are alreadyservices companies with tight
margins.
They're not software companies,right, so they have to be
mindful of that.
So you could see, okay, they'regoing to be driving more toward
(15:57):
this type of technology,probably even faster.
Or you could say they havebudget budgetary constraints,
right, they're not reallythinking about adding on
additional software.
What correlation, if any, areyou seeing between those two
things?
Speaker 2 (16:09):
Yes.
Okay, let's first talk about, Ithink, the market as a whole
and the contraction in thecurrent state.
So I think, like last year, siaput out their study.
I think across the industrythere's pockets where it's
different, but essentially thestaffing industry took something
like around a 10% compressionPretty brutal right.
Yeah, not great percentcompression Pretty brutal, right
(16:33):
.
I guess.
One question to answer flat outis how much of that is due to AI
from their end customers.
I would say that's roughly zero.
Like I do not think.
Yeah, I don't think.
Yet at least People that usestaffing, they're not using
staffing because they've gotsuch great AI tools, because
most of those, you know that.
I don't think.
That's it.
So then we go to, okay, thestaffing suppliers how or, I
guess, the suppliers, thestaffing companies, how are they
(16:54):
implementing and what successare they seeing?
And I think what we are seeingis there are still companies
that are growing.
They're taking share from thecompanies that aren't advancing
as quickly.
It's pretty flat out, Like wecan see, for example, we have a
pretty explicit study that we'veshared widely.
Like customers that useautomation here at Fish Bullhorn
(17:14):
Automation and other automationsolutions not just ours that we
offer, they're 36% moreproductive than those that don't
.
So each recruiter half thecompany using it hires 36% more
candidates per recruiter.
So that's how we gauge theproductivity placements per
recruiter Pretty simple metriclike that doesn't take into
account margin or anything likethat.
(17:35):
So that's our first step.
So those companies that 36%wasn't due to a growth in the
industry, that 36% came fromwhere they beat other suppliers
to their mutual customers.
So that's what we're reallyseeing is like it's a little bit
of a story, a tale of twocities where those that are
(17:55):
aggressive and implementingsolutions even if it isn't the
latest AI solution they're theones who are still growing.
We've got a large contingent ofour customer base who is growing
.
They are adding seeds.
We are.
They've got their financesrunning through our system.
They're making more money.
Their margins at the top lineare increasing.
You can see it all the way downto the very bottom line.
(18:17):
So that's what we're seeing.
It will be a market of winnersand losers and there'll be a
middle period, and I don't knowif we want to go one step
farther, but where thisnaturally will go, if we look
backwards at previous cycles, isthese people that are winning
to come in and start to takebusiness by offering lower
(18:50):
markups and lower rates to thecustomer.
So then that becomes adouble-edged sword where not
only are those customers whoaren't able to deliver at the
lower cost profile are losingthe business they already have,
they're going to start to loselogos to those that can supply
at a lower rate, and then thewhole industry.
This is natural pricecompression.
(19:10):
We hope it doesn't happen, butit feels like that's probably
going to happen when ourcustomer, when the buyers of
staffing services, start to seethat there's enough suppliers
that can offer a lower costbasis.
Speaker 1 (19:22):
Because of automation and AI.
Essentially, the recruiters aremore productive and so they're
able to undercut.
That's right, and just due tostaffing, industry is just so
insanely competitive, right,there's just so many out there.
Speaker 2 (19:33):
And so think, if you're a 20-person staffing firm
, that's very motivated, youwant to grow and you have any
deal where you might lose it bydropping a margin percentage,
you're going to take that dealbecause your net at the end of
the day is still high enough tosupport that business, to
deliver at a lower rate.
And then, especially, everycompany has it where you've also
(19:57):
got the already incredibly lowmargin business and then so much
business that you aren'tservicing today Our best in
class customers maybe have callit a 25 or 30% fill rate.
That's 70% of the roles thatyou aren't hiring and it's
likely, because you don't have,you can't effectively deliver to
those customers, given yourcurrent cost profile.
(20:17):
Maybe you work it, but notreally you don't put your best
on it.
Maybe you outsource it.
That business can now befurther driven down your own
cost basis.
I don't know.
I was going to try to keep avery cohesive answer, I feel
like I've gone a little far.
Speaker 1 (20:31):
All the extra context is it's really helpful?
Yeah, it's.
It's really interesting tothink about how it's going to
impact the competitive landscapeof staffing and I think you you
made a really good point, right, it's the in order to grow.
Particularly, it's always it'shard to grow any staffing
company and essentially anymarket because there's so much
competition, right.
But beyond that, now whenyou're looking at a market
(20:51):
contraction, right, it probablywill force early adoption by
agencies if they do want tocompete, and I think there's a
lot of companies and staffingthat are really old school and
probably will not change andthey're probably going to get
left behind, right.
Speaker 2 (21:06):
Yeah, it's going to be tough for them.
I will say though I will hedgeit a little bit with this isn't
happening tomorrow.
So to throw a little bit ofmaybe numbers into the context,
we recently did a survey of alarge our grid survey and 52% of
our customers are evaluatingand testing AI solutions.
Only 15% have purchased andimplemented and seeing benefits
(21:30):
from AI solutions, so prettyearly.
And, by the way, that 15%source and match.
So like the ability toautomatically run searches and
find candidates.
Of that 15%, that was like 70%of it.
So it's also not even all thisnew stuff like the screening,
like the hyper-personalizedmessaging.
It's things that have beenaround for a little bit longer.
(21:52):
So this isn't gonna happentomorrow, but there is gonna be.
Let's pretend like it's two,three, five years, whatever that
window is, before the seriouscompression starts.
Some people are gonna make haylike there's gonna be some
companies that grow like crazyor reap massive margin benefits
during that time.
But I don't know In my mind,given the previous technological
(22:14):
changes and all the humansinvolved in staffing, this isn't
going to happen overnight.
I don't know if maybe fiveyears is overnight, but it's not
like next year.
It's not like this year where,all of a sudden, the whole
market dynamics shift.
Speaker 1 (22:25):
Yeah, it's going to take time.
I think you briefly touched onthis earlier.
It's the first couple of yearspeople after ChatGPT came out,
people are just essentiallyexperimenting, trying to figure
out which use cases essentiallyare actually creating real value
versus maybe not quite as much.
I think sometimes some of thetools or products that have been
created were really justshifting where you're doing the
(22:45):
work.
I don't know how much they wereactually driving productivity
per se, but I think, like we're,you know at least folks that
are leading the charge right,like people such as yourself.
At this point we have a prettygood pulse on what's driving
real productivity and the usecases, that customers are
becoming a little bit moreeducated right and starting to
really figure out where they can, where they could, drive
(23:07):
outcomes.
You shared that stat about 50%of staffing companies, I think
you said experimenting with AI,and then 15% actually driving
value or having found itimplemented, a use case that's
really like just they're payingfor it and rolled it out
enterprise Okay.
I got it I was reading in theWall Street Journal is something
like overall in the US economy,61% companies are experimenting
(23:28):
with AI, but I think it waslike less than 6% have actually
rolled out an application that'sgenerating like real ROI,
tangible value for theorganization, which is pretty
wild.
It's a very small, so there'sstill a lot of education in the
market.
That needs to happen too.
Yeah, it's going to take time.
Speaker 2 (23:44):
I want to pull one thread from what you just
described about how, when peopleadd these solutions and it just
moves the work, tie that backto earlier, what we talked about
how, where we're seeing success, and it's all about the
leadership in the actualcompanies that are driving it
One of the number one thingsthat must be done.
So let's take the old worldversus the new world.
So let's pretend like arecruiter has 10 actual live
(24:05):
conversations a day.
Hopefully it's more than that,but let's pretend like it's 10.
Today those conversations areat least half of them are
talking to people that aren'tqualified, that we just throw in
the trash, right, we re-engagethem and we find out right, but
we don't use them for that dayor that month.
And then we talk to a few thatare okay and then two that are
awesome.
Now imagine the world wheneverevery candidate you talk to has
(24:26):
been pre-screened, already foundtime on your calendar.
So let's pretend like theystill only have 10 conversations
.
Every one of those should besomeone that's a good candidate
and qualified, but not all forone job.
So in the old world, I'mtalking to 10 people to fill one
role.
In the new world, I should betalking to 10 people to fill
three or four roles.
Speaker 1 (24:43):
You need fewer candidates, because they're
already qualified and good.
Speaker 2 (24:47):
So then, what does that mean?
So I, as a staffing firm, inorder to achieve that outcome,
had to spend money to invest insolutions to get that.
You're investing so you cannothave the recruiter's metrics
remain the same.
You can't still expect them tobuild the same margin, because
it's going to be way If you makeit so that they have to have
the same criteria to make acomfortable lifestyle.
(25:07):
Sure, there's going to be somethat go above and beyond and
absolutely crush it and make amillion dollars a year or
whatever, but the norm isthey're going to just go back to
a comfortable place and let'spretend like you need to add
five, three new starts everymonth.
Even they have all these tools,they're just going to keep
adding three new starts, unlessyou expect them to add six.
So that's the other part.
To drive change and to actuallydrive it, you need to have
(25:29):
confidence in actually thatthese tools will work and change
their outcomes and change themetrics that drive it, and
that's a hard change Recruitersdo not want to hear, yeah, your
requirements are doubling.
To make the same amount of money, you have to have a lot of
proof points.
There's a lot of changemanagement that goes into it.
So that's the part of thetransition.
That's where it can get bumpyand where another area that we
(25:50):
have to watch to make sure youget the outcomes you want.
Speaker 1 (25:53):
Yeah, for sure you should be evaluating and
tracking exactly how much thisis truly impacting revenue.
I'm curious so your customers,are they primarily looking at
these technologies more from aboosting revenue perspective, or
efficiency or margin bottomline?
What is driving them to thepurchasing decision?
(26:15):
That's an excellent question.
Speaker 2 (26:17):
So the short answer is that we think most want to
ultimately increase the top line, drive more revenue, expand
their business.
Right now, though, given themarket conditions and everyone's
internal cost cutting thatthey're forced to go through,
sometimes they go into itlooking at it for that as an
(26:38):
opportunity.
So it's a little bit of a blend.
There's a little bit ofshort-term pain that people are
feeling right now that maybeforces them to look a little bit
more inward and look at waysthey can cut costs, but
ultimately, everyone knowsthey're actually investing in
these solutions to more rapidlygrow and better serve their
customers, and that's a reallybig part of it.
Even if you aren't using thesetools to add new logos, wouldn't
(26:58):
it be nice if you were actuallyjust making your customers that
already exist today muchhappier with you being the
platinum supplier, right Likegetting all the awards and value
that accrues from just doing abetter job, but with the jobs
you've already got today?
Speaker 1 (27:12):
Yeah, for sure.
Look, Jason, we're coming up ontime here and I wanted to see
if there are any other topics oranything else you want to
mention about what you're doingat Bullhorn that you want to
share with the audience.
Speaker 2 (27:21):
I don't think so.
The only other thing that Ithink is always very important
we just talked about how we canmake customers happy by
backfilling.
The only other thing that Ithink is important I think you
can speak to as well is justwhat about the candidate's
experience?
Is this better, worse, the same?
We've got data that talks abouthow candidates are very much
ready to adopt AI.
(27:42):
We don't live in this utopiawhere every candidate gets
responded to.
If we did, AI might make it alittle more rocky, but we know
today they want faster responsetimes, they want to get to work
faster, and AI can actually helpthem to achieve those.
So I guess that's the otherinteresting part is, it's not
about putting your work on thecandidate and sacrificing that.
(28:03):
It's actually this helps you todeliver a better candidate
experience.
Speaker 1 (28:06):
Yeah, it's a much better candidate experience and
it's honestly speaking withfolks even in staffing.
A lot of people do think it'sgoing to hurt candidate
experience and it surprises me alittle bit because it's like
getting into the evaluationtools that we're both building
pre-screening conversationsFolks come inbound to provide
recruiters with qualifiedcandidate shortlists and speak
(28:28):
to folks that are actuallyqualified.
A lot of people are worriedthat candidates are not going to
want to engage with thistechnology, but to me it's like
it's so valuable to you apply toa job and you immediately get a
call or a text to engage you inthe interview process.
I'm in the headspace, I justapplied to the job, I'm in front
of my phone, I got my phone onme and my computer.
(28:49):
I'm available essentially andyou can move a lot faster.
But the other thing that'scurious is and it does make
sense to be concerned about thiswe need to be very careful.
But the ideas of AI and bias,ai discrimination and I think
generally there's just aneducation that needs to occur on
how generative AI actuallyworks versus AI models.
Potentially in the past thatmore so operated in a black box
(29:12):
where it was harder to see whyit was making certain evaluation
decisions, but I see the toolscoming out right now.
Certain evaluation decisions,but I see the tools coming out
right now, like the ones we'rebuilding, as something to
actually create less biasprocesses.
I think people are a big issuewhen it comes to bias right, so
I think these technologiesactually can level the playing
(29:33):
field and create a much betterand more fair and equitable
process for people.
Speaker 2 (29:38):
Absolutely.
It's very hard to change howmillions of people go about
their Absolutely.
It's very hard to change howmillions of people go about
their day.
It's very easy to tell acomputer how to operate.
Even the challenges thateveryone's heard in the news
stories in the past, it'sbecause the way those models
were built is.
They replicate previous humanbehaviors, so all that they're
doing is continuing to push outmore and yeah, and then the
(29:59):
other part.
I think that's really important.
We talk about the candidateexperience is, today all the
decisions are made based on apiece of paper and everyone
knows on a resume, right?
Everyone knows a resume cannever reflect A you as a person
more holistically or B yourspecific qualifications for a
given role.
These screening solutions justallow them to tell more of the
(30:23):
story, talk about why they're agood fit, and I'm sure you've
seen this.
There's been very many timeswhen someone's been a relatively
low resume fit but then havegone through a screen and
they've been a perfect match forthat role, because the maids
are not effective at actuallytelling a story about how good
of a fit you'll be for a role.
Speaker 1 (30:40):
And particularly, I think, maybe some industries
more than others too.
Folks may not really have greatresumes, so it's just having a
level of clarity you could getfrom a screen is, I think, a lot
more valuable.
So we're definitely on the samepage there.
Yeah, for sure.
Speaker 2 (30:54):
And again, just maybe .
One very last thing.
We've got some stats like thosescreening ages.
Again, is it weird to talk toan AI person?
I don't know, Maybe, but it'sgoing to be the new future and
it isn't.
If you've ever done one of these, Maybe in this thing I can put
a link to your screener orsomebody's screener so people
can experience these.
They're actually really goodand it's funny when people are
having these discussions theytalk like they're talking to a
(31:16):
person.
They're like, oh, that's a goodquestion, or sorry, I didn't
hear that, Can you repeat it?
And this is the human talkingto the AI.
And then whenever I'm sure youdo this we do surveys after
every single screen to say howtheir experience was, and
they're overwhelmingly positive82% give an 8, 9, or a 10.
And I think this is important,but that's 82%.
(31:36):
There are some that give a oneand a two and it's actually
pretty polarizing.
The ones that give the lowscores.
It's usually.
I don't think AI should be doinghiring.
I don't want to have my future.
They've got very opinionated,dogmatic concerns and I want to
bring that up.
Whether that's valid or not,every customer that implements
one of these is going to have arecruiter that tells them a
story about a candidate whoreached out with this opinion.
(31:59):
But we just have to weigh inthe counterbalance of that's a
very small subset that's 2%Whereas the vast majority really
actually like the experience.
So there are going to be somethat don't love it, but that's
anything in life.
Speaker 1 (32:12):
What it reminds me of is back in the day, when
there's a lot of industries,right Even like when it comes to
buying stocks where you used tohave to essentially speak with
a salesperson early in theprocess, and remember when a lot
of that went online and becameautomated, a lot of stockbrokers
and people would say, oh, theyneed to speak to us, right,
they're not going to beinterested in working through
(32:32):
automation.
Think about just how manyindustries have made that shift,
have made that shift.
And it's the same thing.
Like, people don't necessarilywant to speak with somebody
early on.
If they could just do thescreen and get a little down
funnel, let's see if it'sactually worth the time.
Scheduling an interview andthen also these screen products
are also good for candidates tobe able to ask questions too and
(32:53):
learn more about theopportunity.
So the idea that people want tospeak with people early on,
it's not necessarily people wantto speak with people early on.
It's not necessarily they wantto speak with somebody if they
know it's a relevant opportunity.
Nobody wants to waste theirtime and you get scheduling and
rescheduling and it's just, butthat's also making the
assumption that ever has thatthat's happening today.
Like that is the difference.
Speaker 2 (33:15):
Like the stockbrokers , had a very vested interest in
calling a bunch of people andtrying to sell them.
But we don't have.
Let's pretend like only 100people apply to a job.
95% of them are never going tohave a conversation.
They don't even get the chanceto have a conversation.
Speaker 1 (33:24):
Yeah.
So it's yes it levels theplaying field and you're right,
it gives folks the opportunityto talk about their skill set
that might otherwise have notbeen reached.
So I think people are going tocome around and it's from a
product perspective.
It's important for us to bebuilding products of the future
and clearly things are going togo in this direction and people
are going to become morecomfortable with it over time.
(33:45):
There's always going to bepeople that, regardless what you
do to the hiring process,they're not going to like it.
People don't like it if youdon't like getting turned down
for jobs.
So there's always going to besome negative feedback,
regardless if it's people or AIor anything else.
Yeah, yeah, going to be somenegative feedback, regardless if
it's people or AI or anythingelse.
Yeah, yeah, jason, thank you somuch for joining us today.
This has been a lot of fun.
You've shared a lot of greatinsights with our audience today
(34:06):
, so I know everybody will havelearned a lot.
I really appreciate yourcontribution to the show.
Speaker 2 (34:11):
Absolutely.
Thank you so much for having meand excited to talk again.
Speaker 1 (34:14):
Yeah, absolutely.
You're welcome to come back onanytime.
Speaker 2 (34:16):
Nice.
Popular Podcasts
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!
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Crime Junkie
Does hearing about a true crime case always leave you scouring the internet for the truth behind the story? Dive into your next mystery with Crime Junkie. Every Monday, join your host Ashley Flowers as she unravels all the details of infamous and underreported true crime cases with her best friend Brit Prawat. From cold cases to missing persons and heroes in our community who seek justice, Crime Junkie is your destination for theories and stories you won’t hear anywhere else. Whether you're a seasoned true crime enthusiast or new to the genre, you'll find yourself on the edge of your seat awaiting a new episode every Monday. If you can never get enough true crime... Congratulations, you’ve found your people. Follow to join a community of Crime Junkies! Crime Junkie is presented by audiochuck Media Company.