Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:00):
Today we got some
pretty cool topics to discuss,
including we're going to gothrough the roles that are most
in demand in the tech industry,talk a little bit about
recruiting for engineers andmachine learning talent.
We're going to go into a recentacquisition that the team over
at Jim has made Steve's going totell us all about that and
we're going to talk about someother things as well in terms of
(00:20):
how to engage and recruittalent.
So it's going to be a good mixof topics today and, steve, I
think where we could probablystart is going through roles
that are really in demand.
Like I had mentioned before, wehit record that we're seeing a
fair amount of demand for more,so technical talent, engineers,
machine learning talent and Iwas doing some searches on open
(00:41):
jobs just on LinkedIn in thetech industry and I was typing
in different job titles to seethe amount of open jobs.
So, like for internet softwarecompanies, I saw, I think, the
stat I saw just within that, and, of course, there are SaaS tech
companies that aren'tcategorized on their LinkedIn
page as internet.
So there's probably a lot more,but I saw 10,000 plus open jobs
for machine learning jobpostings and then for account
(01:05):
executives, I only saw athousand and for SDRs I saw a
thousand.
So it's basically a 10 to oneor five to one between revenue
roles and technical roles atthis point in time.
One of our biggest RPO contractsthat we're working on right now
is helping a growth stage SaaScompany hire machine learning
(01:26):
talent, and it's really the only.
Those types of roles are reallythe only ones where it's like a
very competitive market, likewe have product roles at other
companies, revenue roles atother, and closing or hiring is
pretty straightforward right now.
It's they're not gettingmultiple other offers at the
same time.
So we have our pick between aton of candidates and as long as
(01:47):
you make a competitive offer,they're essentially going to
accept.
I don't think we've had anoffer declined on revenue roles
or product roles in the pastquarter, but machine learning is
.
If we get to an offer, theyhave four other offers.
So I guess we wanted to talk alittle bit about that and hiring
engineers and maybe go throughsome data around that.
Where do you think we shouldstart there?
I'm just curious, based on dataand what you're seeing at Jim,
(02:10):
if you have any really helpfulinsights there related to hiring
engineering talent right now.
Speaker 2 (02:16):
Yeah, no, that makes
a lot of sense.
First off, I think that matchesmy intuition entirely when you
look at what's happened to openroles.
First of all, open roles aredown across the industry, except
for maybe in a few select spots.
But the places where they'redown the most is on the revenue
side, the sales side, comparedto before when companies had
(02:39):
more of this.
Grow at all costs mentalityback before the tech downturn.
Grow at all costs mentalityback before the tech downturn.
And so, yeah, revenue roles aredefinitely down a lot year over
year.
And then also HR and recruitingroles are pretty hard to come by
.
So, relatively speaking, I'dsay those two job families are
the lowest compared to wherethey have been, say, two years
(03:02):
ago.
But there's some roles thathave held up pretty well and
others that are, I think, ineven more high demand than ever,
the ones that I've heard aboutand that just also match my
intuition.
First off, I think execrecruiting is alive and well, so
every company needs leaders,and leaders turn over and CROs,
(03:24):
VPs, avenge, cmos, those typesof roles like companies are
still trying to hire for,replace and find great execs.
And then, as you called out,technical roles are remaining in
pretty high demand butspecifically like any talent
related to AI, given what'shappening in the industry.
So I think what you shared,what you're seeing, really
(03:45):
matches up what I'm seeing andmy intuition as well.
Speaker 1 (03:48):
Yeah, for sure.
It's just a lot of our othercustomers that aren't hiring ML
talent really don't seem to behiring much at all.
They're keeping their head downpretty low.
Yeah, we've seen some revenuetalent.
We're working with one customer.
This is an earlier stagecompany and they're hiring
somebody in product as well asin UX, and we got flooded with
(04:11):
800 applications within I don'tknow several days and it's weird
because we're turning downreally good candidates and when
they're asking for feedback,it's just look, you're really
good at what you do.
It was a positive experience.
It's just that the market'sflooded with so many candidates
that we were essentially able tofind the exact profile,
(04:33):
essentially, of what we'relooking for.
We'd love to keep in touch withthe future.
There's just so many peopleright now.
It's just wild.
Speaker 2 (04:42):
Yeah, and that's the
interesting thing is because
there are fewer open roles andthere's more candidates on the
market and there's a lot ofqualified candidates on the
market.
Actually, the biggest challengefor many of these roles the
ones that aren't as hard to fillright now is just how do I get
through all my inbound?
We've seen, I think, somethinglike a 40%, maybe even higher,
increase in inbound like yearover year and it was already
(05:05):
increased last year compared tothe years prior.
And really cool GEM stat, Ithink 20% of our customers have
1,000 plus inbound applicantsper role, which is pretty wild
and you just think about thesheer amount of inbound that you
(05:25):
have to get through.
That's a lot of inbound andhopefully a lot of qualified
people applying to these productroles and also to these revenue
roles and open recruitingheadcount and things like that.
Speaker 1 (05:35):
Yeah, for sure, For
sure.
But getting back to ML andengineering talent, I was having
a conversation earlier todaywith our probably our best
technical ML recruiter here atSecure Vision.
She shared some prettyinteresting insights with me,
but basically I asked questionsaround how she goes about
sourcing and engaging talentwhere she's finding them and, as
(05:59):
you can probably imagine, shewas finding a lot of the talent
on LinkedIn.
But a lot of these candidatesare getting a ton of outbound
recruiters reaching them viaoutbound and she is sending
LinkedIn and mails, but she'salso leveraging Jim to send
emails as well and she'sessentially it's a combination
(06:19):
of those things.
She is sending follow-ups, butshe's trying to be thoughtful.
She's not spamming them withtoo many messages.
We actually didn't get intoexactly how many follow-ups
she's doing.
I'm sure there's a bit of adiminishing returns aspect, but
also I think you have some statson like how many follow-ups are
actually ideal for getting thebest results.
I guess like we could startthere, like in terms of
(06:41):
multi-channel and the amount offollow-ups for engineering
talent specifically, do you haveany insights there on what
companies and recruiters shouldbe doing?
Speaker 2 (06:48):
Yeah, a bunch.
This is our bread and butterbecause we got our start with
sourcing and, of course, gem'splatform's a lot broader than
that now, but we're experts onthe sourcing side and the best
practices to reach hard to filltalent and roles.
Yeah, on the sourcing side,especially for things like ML
engineers.
Of course it applies to likeleadership roles, any hard to
(07:09):
fill role.
First off, actually, forengineering talent especially, a
lot of engineering talent,including ML engineers, have
their email notifications off.
They're savvy to know that theycan actually change that
setting in their LinkedInsettings.
Most of us don't even know thatit's an option, but yeah,
(07:30):
they've turned that off because,historically, the last 10 plus
years they've just beeninundated with emails and a lot
of technical talent feels likeit's a noisy channel for them
and for a long time it's been alot better and more effective to
reach technical talent,especially like ML talent, over
(07:51):
email.
But I think the omni-channelapproach is even better.
But something like I thinkclose to half of, I think,
technical talent, especially forthese hard to fill roles, might
even just have their emailnotifications off, which is a
lot higher than like most otherroles, and there's certain roles
where you couldn't even turnyour notifications off like
(08:11):
revenue roles, recruiting rolesbecause you need those to be
able to do your job right.
So that's the first thing.
Taking an omni-channel approachis really effective.
So it's actually cool to hearthat your recruiter is maybe
sending an email, but alsosending a few emails and
follow-ups, because then she'sgot a much higher chance of
reaching people in theirpreferred inbox, whether that's
(08:32):
LinkedIn or email.
In terms of follow-ups, our datahas shown for a long time that
you get about half of thepositive responses from emails
two through four that you send.
So if you're only sending onemessage, you're leaving about
half of the positive replies onthe table right.
So it's really important tofollow up.
(08:54):
Historically, there might'vebeen a question as to is that
worth it?
Because following up takes alot of time and energy.
You have to remember to do it,you have to set calendar
reminders, all that good stuff.
But now with tools like Jam andother email automation, like
sourcing automation tools, it'salmost a no-brainer, because you
can tee up a four-stage emailsequence with three automated
follow-ups on the exact days andtime a week that you want with
(09:17):
no extra effort, and so you'reno longer leaving those 50% of
the positive replies on thetable.
Speaker 1 (09:24):
Yeah, for sure.
And for engineering talentspecifically, do you have I
don't know if it's justaggregate statistics across
several roles, or do you guyshave the stats on for
engineering talent?
How many follow-ups is ideal?
Speaker 2 (09:39):
You still want to
send four messages total in that
initial outreach.
It's, I think, the same acrossmost roles the stat that I
shared about 50% of yourpositive replies being left on
the table if you're not sendingthose follow-ups.
That's the average across allroles.
I don't have the exact specificstats for EngTalent in terms of
, like, how many people respondto messages two through four,
(10:03):
but it's important.
I think.
That's where the balance,though, comes in, because as you
send more and more follow-ups,as you can imagine, you start to
get diminishing returns.
Because, as you send more andmore follow-ups, as you can
imagine, you start to getdiminishing returns, and our
data showed and I forget exactlywhat the data is because it's
been a while since I've lookedat it but when you send message
5, 6, 7, 8, which some people doin that initial outreach, it's
(10:24):
just not really worth it.
You get a little bit of a boost, but it really falls off after
message 4.
And at that point you have tostart worrying about am I
hurting my talent brand?
Speaker 1 (10:34):
Yeah, you probably at
that point.
Also, if you're not, the otheraspect I would say is, if you're
not getting follow-ups, at thatpoint maybe look at your
messaging again.
Yeah, I was asking.
So I was asking my lead techrecruiter, like, how do you go
about messaging?
And it was actually prettystraightforward and simple.
She just said she was sendingpretty short messages but she
would really focus on what shewas relevant about their
(10:58):
background and speaking to theprojects that they would be
working on.
So she wouldn't necessarilyjust say high-level stuff like
hey, I'm working with a growthstage SaaS company that's
building out their ML team, forinstance.
She might say something likethat, but it was more of hey, I,
really I noticed that you havethis experience in distributed
systems on this and that, andwe're, we're.
(11:18):
I'm working at a company rightnow who's also in the growth
stage, similar size and scale,who needs ML engineers to work
on X, y, z project.
Cool, because she was justsaying that a lot of tech
recruiters that are just readingrequirements in terms of
technologies.
(11:39):
They don't really understand howto communicate with machine
learning engineers or justtechnical talent in general in
terms of speaking to specificprojects that they're likely
working on, even if they don'thave it completely written out
Understanding the industry, thetype of company they're working
for, the types of projectsthey're probably working on,
context cues on their profiles,and then, speaking to that
(12:03):
experience, along with how italigns with the, when you're
able to basically articulate theprojects that they're working
on, not just the technologiesbut the types of projects and
the types of projects they wouldbe doing with your company,
then that's really how youmaximize the response rate,
(12:24):
because otherwise they're goingto think this person doesn't
really understand my backgroundor the type of work that I do.
Just because they understand atechnology I work with doesn't
really mean they understand,like, my role within the company
and how I make an impact inthese types of things.
Pretty straightforward, but Ithought it was pretty helpful as
well.
Speaker 2 (12:40):
Yeah, that makes a
lot of sense, I think, in terms
of the outreach itself forreally hard to fill roles like
hyper personalization starts tomake a lot more sense.
Even personalization in generalcan be really effective.
Now, I'm not talking about heyname at company Saw, you've been
working on title right, likethat kind of stuff gets you a
tiny boost.
(13:00):
But it's not like the same aslike a recruiter actually
sitting down looking atsomeone's profile and, like you
said, like mapping that to whythey'd be excited to work on
these very specific initiativesor projects.
But even using AI topersonalize, we find that
customers that use gem AI intheir draft creation get
something like a 40% boost ontheir response rate, which is
(13:21):
pretty cool.
Speaker 1 (13:23):
That's insane.
Speaker 2 (13:25):
Yeah, it's awesome.
Maybe the last best practice Iwould offer up in terms of folks
that are looking to breakthrough the noise and reach hard
to fill talent, is sending onbehalf of a hiring manager can
be super effective when youthink about it.
For a machine learning engineer,somebody that's in really high
demand like, who would theyrather hear from?
(13:47):
Would they rather hear from arecruiter or maybe the hiring
manager on the team that they'dbe joining, or, better yet, like
the VP of engineering, right?
And so we find that sending onbehalf of a hiring manager or an
exec I've seen this canincrease response rates by one
and a half X to up to three X,depending on the type of role
(14:09):
that you're hiring for, how hardto fill it is and how much it
would resonate to hear from anexecutive on the team or a
hiring manager.
Now I think the key is you wantto make sure that those folks
are actually invested in thehiring process and especially
your hiring managers, that theywould be excited to hop on a
pre-sale call with reallystellar candidates.
I think it can feel a littlebit bad to be bait and switched
(14:34):
in terms of getting reached outto by an exec but then having
them pass you right back to therecruiter.
That can still be effective.
Both approaches can work.
Speaker 1 (14:41):
Yeah, definitely,
that's a really good point, and
I guess that's probably onlypossible through not only
possible through Jim, but I'veactually done that sometimes on
LinkedIn as well, where I'llhave recruiters send messages
through my profile Totally,which I don't know if that's a
best practice people shouldfollow.
(15:02):
I don't know if I could get.
I could see myself gettingflagged by LinkedIn for doing
that.
Hey, we saw a recruiter loggingin from a different city or
from LATAM or whatever.
We saw that IP address and theycould shut down my account or
something.
So maybe that's not.
It hasn't happened.
I've been doing it off and onfor almost six months at this
point.
Speaker 2 (15:23):
So nope, people do it
all the time and actually it's.
We've done this at Gem and, yeah, we've never run into any
troubles with sharing logins forLinkedIn, especially when you
want to be sending on behalf ofsomebody.
But I think it gets a lot easierif you can automate the process
.
All these things whether it'ssending follow-ups,
(15:43):
personalization, stobo, as wecall it, send on behalf of a
hiring manager or an execomni-channel approach and all of
these things are certainlypossible without GEM and
regardless of whether you'reusing GEM or not, you should
definitely try them out, measurethem, see the impact for
yourself.
But they get a whole lot easierwith Gem because we automate
the whole thing and put it onautopilot.
(16:04):
Actually, james, I didn't knowif you know this, but even for
folks that are collaboratingwith in-house teams, you can
even set up a cross domain.
Your recruiter could be usingyour account and then you could
get an alias from one of yourclients that says it basically
looks like the hiring manager orit looks like the VP.
They could off that with yourgem account and then you could
(16:26):
now be sending on behalf of thatemail address for your client.
Speaker 1 (16:31):
Oh nice, yeah, that's
pretty cool.
That's awesome.
I'll definitely I'll have tocheck that out.
I don't know, I'll talk to mytech recruiter to see exactly
how we're doing it now, but yeahfor sure.
And I also I wanted to talk toyou a little bit about the
recent acquisition that yourteam made.
If you could do, you want togive us an overview of the type
of solutions.
I don't know if you can namethe company or whatever details
(16:53):
you feel like you want to share,but yeah, I guess, just like
why your team made theacquisition, what, but yeah, I
guess, just like why your teammade the acquisition, what the
technology does and why you felt, based maybe on, like customer
conversations or whatnot, itmade sense to make this
acquisition.
Speaker 2 (17:08):
Yeah, 100%.
We actually announced theacquisition a few weeks ago, so
I'm really excited to share more, get the word out.
So JEM acquired a companycalled Interview Planner and
we're really excited to welcomethem to the team.
Interview Planner doesautomated interview loop
scheduling, so similar to goodtime modern loop prelude.
Of course, like we continue tohave partnerships with all of
(17:32):
those folks because ourphilosophy is that it's really
important for customers to beable to choose whatever solution
is best for their needs.
With that said, we're reallyexcited to be able to offer
interview scheduling nativewithin the GEM platform now, and
so we're actually for any ATScustomer of GEM, we're including
(17:52):
that as part of the applicanttracking system package all in
one.
But one of the really coolthings about Interview Planner
is not only is it going to befully integrated with Jam ATS,
but it also integrates withGreenhouse and Lever and, very
soon, workday and later thisyear, a lot of the enterprise
applicant tracking systems, andso, whether you use Jam ATS or a
(18:17):
different ATS, you can stillleverage the full power of
interview planner, and that'sjust one of our core ethos when
it comes to building gem is allof our products work whether you
use a gem ATS or an ATS of yourchoice.
Speaker 1 (18:33):
Yeah, for sure, and
so this makes sense because I
think also part of your strategytoo is be able to offer like a
full suite of products, like allin one to an extent, or to be
able to provide all thefunctionality to your customers,
so there isn't a need tonecessarily go to other
providers or vendors.
That's essentially thedirection that you're going in,
right To some extent.
(18:55):
Yeah, that's right.
Speaker 2 (18:56):
Yeah, we got our
start by building like just the
best of breed sourcing, crmanalytics, but over the last few
years we've really broadenedour mandate and now we're trying
to build the best all in onehiring platforms.
That spans sourcing, crm, fullfunnel analytics, recently
applicant tracking system,talent marketing, recruitment
(19:18):
events, career site which wejust announced interview planner
and interview scheduling as ofa few weeks ago, and in just a
month or two we're now going tohave AI sourcing bots as well as
part of that all-in-one bundlesTruly the most comprehensive
all-in-one recruiting platformfor pretty much any company out
(19:39):
there.
But the great thing about howall this works is you can use
every single one of thoseproducts, whether you use Jam
ATS or a different ATS, so it'svery bespoke in terms of being
able to plug in any ATS you wantinto that bundle.
Speaker 1 (19:56):
Yeah, that's smart,
makes a lot of sense and, yeah,
you're able to providefunctionality that other firms
can, and then hopefully they'llend up moving to your ATS as
well, doing everything throughyou guys.
Speaker 2 (20:09):
Totally.
The way I like to think aboutit and talk about it is
certainly you can use any ATSyou want.
Everything will work slightlybetter together if you use all
gem products.
It's like how, with Apple, youcan mix and match right, you can
have your AirPods, but they'regoing to work slightly better if
you're using an Apple computerand an iPhone, just like the
(20:29):
seamlessness between theBluetooth and the connectivity
is going to work a tiny bitbetter than mix and matching
with a different device.
Speaker 1 (20:36):
Yeah, for sure that
makes a lot of sense.
Also, just like from the dataperspective reporting and data,
I would assume.
Like that, when you integrateall these different systems,
it's probably like nobody wantsto log into.
Like multiple dashboards acrossmultiple systems.
Speaker 2 (20:52):
It's probably like
nobody wants to log into like
multiple dashboards acrossmultiple systems.
I don't like to.
Oh yeah, no, totally.
I think, like the big reasonswhy you'd want to use an all in
one platform and why thesethings would be better together.
First is like you get a trulysingle source of truth because
everything's fully integrated.
Every single talent,relationship and touch point is
going to sync back to the ATSand the CRM in a way where, when
(21:14):
you're trying to integrate abunch of different tools in the
stack, you're bound to lose somedata or for that integration to
not be 100% reliable, and soyou get that single source of
truth.
That's important for a fewreasons, one of which is better
data, because then you can havea single place to report on all
that data.
You can see how all of thethings at the top of the funnel
whether it's sourcing your CRMefforts, your talent marketing,
(21:37):
your events how all of that isconnected to and driving actual
results with your hiring funnel.
And then the other thing that alot of our customers are
excited about is having oneplace to configure all of their
admin functionality, their usergroups, their permissioning,
their compliance requirements.
Speaker 1 (21:56):
I didn't even think
about that, but yeah.
Speaker 2 (21:58):
Yeah, it's huge
because you don't want to set up
different data retentionpolicies across eight different
tools and have them be out ofsync for GDPR.
But yeah, user groups likeconfiguring all that once, like
all of your user permissions inone place.
User groups like configuringall that once, like all of your
user permissions in one place.
And then, for folks that areinterested in AI solutions and
like AI sourcing, for example,which I'd say is a lot of people
right now, it's reallyimportant to have that single
(22:20):
source of truth where you havefull context about every talent,
relationship and every touchpoint, cause you don't want your
AI sourcing bots spamming themarket without any knowledge of
who's who and who's talked toyour company recently.
Speaker 1 (22:33):
Yeah, that's a really
good point.
Actually, I hadn't thoughtabout that one.
But yeah, you'd want there tobe a consistency and you'd want
to not if you're going to beusing AI to not basically spam
people over and over again orthrough multiple tools or
anything like that.
It makes a lot of sense.
Yeah, it's really cool.
I'm excited to see how you guyscontinue to change Evolve.
We use as an RPO.
(22:54):
We're actually using GEM.
Right now.
We haven't made the transitionto the ATS, but I think we'd
actually like to do that as well, so we should maybe talk about
that some other time.
I'm curious.
I'd like to get started on that.
But yeah, we're definitelyusing I.
I gotta talk and figure out,make sure we're using the ai
functionality as much aspossible, but when the bots go
(23:16):
live, definitely let me know andwe can give it that that a shot
, and then maybe we could talkabout that functionality as well
and just talk about, like, howit's working and what types of
roles it's working with, andthat kind of thing.
Speaker 2 (23:27):
That'd be really
interesting yeah, sounds like a
great future conversation.
Speaker 1 (23:32):
Yeah, for sure.
Hey, Steve, thank you so muchfor joining again today and for
everybody tuning in.
We're going to be doing thisonce a quarter, so make sure to
tune in.
But yeah, Steve, this is alwayssuper helpful, so I appreciate
you coming on.
Speaker 2 (23:44):
Likewise.
Speaker 1 (23:44):
Thanks.
Speaker 2 (23:45):
James.
Speaker 1 (23:45):
Awesome, thank you.