May 8, 2023 • 50 mins
Text us your thoughts on the episode or the show!
In this episode, we talk with Sylvain Giuliani about how Marketing Ops pros can take advantage of the increasing importance of first-party cookies. Sylvain is currently Head of Growth at Census, a reverse ETL provider. He is also an active member of several communities, primarily in the revenue space. He is also an investor and mentor with experience at several startups - as a founder and leader.
Tune in to hear:
- The state of things with first-party vs third party / intent data; and, the trends/impacts on the ease of capturing and quality.
- How Sylvain approaches the challenge of the importance of identifying customers’ preferred communication channels to optimize spend/effort.
- How Marketing Ops teams can leverage this insight to strategically collaborate with other marketing teams (demand gen, ABM, etc.) - specifically as it relates to access and understanding of data.
To contact Sylvain on LinkedIn, click this link to his profile: https://www.linkedin.com/in/sylvaingiuliani/
Episode Brought to You By MO Pros
The #1 Community for Marketing Operations Professionals
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Michael Hartmann (00:00):
Hello everyone.
(00:00):
Welcome to another episode ofOpsCast, brought to you by
Marketingops.com, powered by theMO Pros.
I'm your host, Michael Hartmann.
Joined today by one co-host,Mike Rizzo, who has been just,
he just flew in, so Mike.
Yes, I just zoomed
Mike Rizzo (00:14):
right in.
Thank
you.
Michael Hartmann (00:15):
I'm back.
I was zoomed right in.
Mike Rizzo (00:17):
I just zoomed right in cause we're doing this on
Zoom today.
And uh, yeah, I was at theMOps-Apalooza venue site visit,
getting ready for a big event,so that was
Michael Hartmann (00:27):
exciting.
All right.
MOps-Apalooza.
September, October.
I think I should know this.
I know I should know this.
There's a reason why we haven'thad episodes.
Okay.
November.
Wow.
See November.
I can only keep a few more datesin my head with all the projects
I have going on, so, sorry.
Fair.
It's okay.
(00:47):
Shame on me.
All right.
Well let's, let's get this,let's get this party started
cause we've got our guest today.
Si Giuliani, uh, We are gonna betalking to him a little bit
about, um, gosh, a couple ofdifferent things, but I think it
ultimately, it kind of comesdown to the sort of pending, um,
or, or increasing need to, tohave first party cookies and how
(01:08):
that applies to marketingoperations.
So, Uh, so is currently head ofgrowth at Census, a reverse ETL
provider, which I'm, I don'teven know what that means, so he
might have to tell us what thatis.
But, uh, he is also an activemember of several communities,
primarily in the revenue space.
He's also an investor mentor andhas experience with several
startups, both as a founder anda leader.
So, Sylvain, thanks for joiningus and welcome.
(01:32):
Thank you for having me.
Yeah.
What, what the hell, what thehell is reverse etl, by the way?
I just like, just get, I mean,
Sylvain Giuliani (01:39):
yeah, yeah, yeah.
I mean, reverse CTL is the actof, uh, getting data out of your
cloud data warehouse that youusually, uh, powering like your
BI tools and getting the datainto your operational tools such
as, you know, Marketto,Salesforce, you name it.
Like we have hundred plusintegration nowadays.
Oh, okay.
And, uh, okay.
So, you know, next time someonesays like, Hey, I see this data
(02:01):
and Looker.
I wish I could do somethingabout this churn problem.
Who are the people who churn?
Well, you can use RiverUniversity to identify the
people who are churning.
Get them into market load so youcan send them emails
Michael Hartmann (02:12):
before they got.
Got it.
Okay.
All right.
That makes a little sense.
Yeah.
All right.
That was not, that was notintended to be a pitch.
I just like, I saw that term andI was like, I don't know what
that means.
I kind of know what it means,but, all right.
Um,
Mike Rizzo (02:26):
we need more on the topic.
In fact, we were talking to Siand Si and their team about, um,
helping us learn more about thetopic.
So I think there's probably moreof that coming soon.
Okay.
Because there's not enough of usthat know what reverse ETL is,
that's for sure.
Michael Hartmann (02:40):
So we're excited about it.
You know, this, this whole ideaof like, Well, the, well, we're
now recording.
It's, it's Cinco de Mayo, 2023as we're recording.
And, um, there's been the, the,the latest MarTech landscape
just came out, what, earlierthis week, maybe late last week.
Yeah.
And what's the number?
(03:01):
14,000?
Is that what I saw?
I haven't had a chance to lookat it.
It's over 10,000.
I know that, but yeah.
Yeah, it's 11,000 something.
Now this whole need to sort ofstitch together data, I think
is, is a big one.
So yeah, definitely a topic weprobably need to, we need to dig
deeper into at some point.
So, Just, yeah.
You know, Sylvain, we'll haveto, uh, we'll have to bring you
(03:22):
back on for that.
Maybe.
I love it.
But one of those things isprobably, so first party
cookies, right?
So, um, you and I had aconversation quite a while ago,
but um, and things maybe havechanged a little bit, but, um,
kinda what's like, I love yourtake on kinda where things are
with first party versus thirdparty.
It's been a while since, I thinkwe actually had a guest on, like
(03:44):
over a year ago talking about.
This a little bit, but more,more so from the sort of
deprecation of third party dataor intent data.
But what are you seeing?
What are the trends, impacts?
Is there any equality stuff likejust a general overview?
Sylvain Giuliani (03:59):
Yeah, no, I mean, you know, the cookie
apocalypse happened.
I, I feel like in, in our space,I think, uh, you know, if you
think back a couple of yearsback, like as a marketer, you
were using one of those 10,000tools out there and to get data
in, you know, your hotspot,Marketo.
It was very simple.
You just slap some JavaScriptcode in your app, like on your
website or some code SDK in yourmobile app, right?
(04:21):
And it will do all the trackingfor you.
Facebook will do all theattribution for you, will tell
you like the conversion rate ofall of your campaigns, things
like that.
So didn't require you any kindof like data, uh, skillsets or
data transformation, skillset.
Then create happen and all thatdata basically.
Went away or you are very lowconfidence about it because it's
(04:41):
getting blocked and a lot andyou know, it's just not possible
to track it basically.
Right.
And so what we are seeing, youknow, as census as someone who
like took to a lot of like datateam and marketing team who are
using data is like a huge if tofirst party data, right?
And so what is first party datareally?
I mean, it's kind of like thesame thing, but you are the one
generating it and not a thirdparty vendor.
(05:01):
So it's like, you know, if yougo on census website today, we
are gonna track your click,right?
But it's like, Our own libraryis tracking your click.
So nobody's blocking it at,doesn't block it.
Right?
So, or it's event that might behappening in your, on the server
side.
Right.
So you know, when you create adocument in Figma, like that
event is generated on the serverside, so nothing can block it.
Right.
So we can know, oh, si create adocument in Figma and you shared
(05:24):
it with Michael.
Right?
So that's again, like that's anevent.
That is first party data.
And then you want to again, As amarketer, you wanna consume that
data because you wanna buildlike audiences segmentation.
Maybe you wanna do sometransactional email or push
notification based on like, uh,a trigger that they achieved.
Like, hey, you said like 10, uh,10.
10 documents in Figma, somethinglike that, right?
And so you, that data goesthrough the warehouse, that's
(05:46):
first party data goes throughthe W and then you wanna sync it
again to a tool like Marketo.
That's how way Senten tool likesensors comes in handy, right?
There's other ways to do it.
You can just write customscript, get you type the data
team to write code for you.
But most, most importantly, it'slike how do we, what we see is
like a lot of marketing peoplewanting to have.
Self-serve, access, themythical, like, you know, I
(06:06):
don't wanna talk to your datateam to self serve the data.
Like, kinda like what you usedto have with like the third
party data.
Uh, that's what we've seen.
That's kinda like the shift thatis happening.
Uh, you know, we can talk alittle about, also about like
CDPs, you know, the rise anddemise of CDPs.
It's also powered by this kind,like trend I would say.
But, uh, I will stop that to seeif that's, Answer your question.
Michael Hartmann (06:24):
The rise and demise of CDPs.
That's, it was a quick one ifit's demise already, but, um, so
there's two things that aregoing on.
I think two things that strikeme about this.
One is this sort of move awayfrom third party data.
It reliance on that.
Feels, feels like it's like adown, like a down negative
thing, but there's a potentialupside to first party data, but
(06:47):
it also comes with extra.
You need different additionalskill sets maybe, and tools to
capture even more data thanyou're already capturing in, in.
I don't know about you, but mostmarketing ops teams I know are
already swimming in data but notable to do much with it.
So how do we over like, yeah.
Feels like two sort ofchallenges at the same time.
Sylvain Giuliani (07:09):
Yeah, and I think that's like the, I think
there's.
Like next to the cookie agra,uh, I would say of third party
data.
Acas, right?
Like there was a big trend, uh,in the back, which was like the
rise of the warehouse, like, youknow, so the past few years, you
know, storing data, transformingdata has never been that easy.
Like, you know, it's like you UKit's very cheap to deploy.
It's very cheap to implement.
(07:30):
And so on the other side of yourbusiness, an IT team, a data
team, are probably investedtremendous amount of time and
money into building a warehouseand ultimately a bi.
Solution.
Right.
You know, like there's aexecutive team or board meeting
that is being powered by adashboard somewhere, right?
And so all that data is beingclean, transformed because big
decisions are being made on thatdata, right?
(07:50):
And so that's the, that's thebig shift that happened in the
back.
And so now it's like asmarketing team, that's great
because we can just build on topof that work.
Like we don't have to reinventthe wheel.
We don't have to.
That's why I say like the demiseof CDP, right?
Because the CDP is like you haveto redo all that work from
scratch again, as a marketingteam, like nowadays.
Solution plug into the west.
I'm not even talking aboutsensors, but you see, like if
(08:12):
you're a price customer, houseplus customer, you will see
them.
They're talking about like, Hey,we can connect to the warehouse,
we can send data to the west.
Like, you know, they're signed,everybody's trying to make a
warehouse, the first classcitizen, right?
And so as a marketing team, wecan leverage the work that has
been done by IT and data team.
And then just consume data thatwe can trust.
Right?
And I think that's like also athing that is accelerating the
(08:32):
adoption of first party databecause you don't have to start
from scratch essentially.
Right?
So all of those event that wetalked about that is happening
in the background, like numberof document created, who created
those documents, like anythinglike that, all of that is
already exists.
Like it's powering BI report,you can just reuse it as a
marketing team, right?
And I think that's, uh, one ofthe biggest, uh, you know,
accelerator of that trend.
(08:53):
Yeah, I
think,
Mike Rizzo (08:54):
I mean, I think a lot of this layers into there's,
you know, the PLG motion for us,for the SaaS companies out there
where they're trying tounderstand, you know, you were
still, you were hinting at, youknow, oh, you shared, uh, 10,
uh, links from Figma, right?
So like that, this idea thatyou're collecting usage behavior
from your, from your first partysort of interactions.
(09:17):
Um, and I think it's important,like, you know, for, for the
audience out there to understandthat like, The death of the
cookie, so to speak, isn't likearound this concept of, you
know, your HubSpot and munchkincodes and all that stuff, like
that's not going away.
Like, not like you've been doingthat and it's not going
anywhere.
(09:37):
Those are first party cookies.
That is your data.
It is collected, it is not gonnago anywhere.
People can turn off the cookielike consent banner and yes,
that part will stop flowing justlike any other cookie.
Michael Hartmann (09:51):
But that's already happening today.
Mike Rizzo (09:52):
Well like that.
Like that's already happeningtoday.
So in large part, like I wouldlove, and I'm gonna interject
here and this is where Michaelloves me cause I just jump in on
things and despite the outlineI'm sure we were probably
getting here.
Uh, I would love.
For you to say, you know, to, tojust share your thoughts on, you
know, really like, is the B2Bmarketer at the end of the day,
(10:17):
small to medium size, even the,even the enterprise.
Like, are they really impactedby this idea of like the cookie
being retired and like goingaway?
Like it feels like it's more ofthe e-com side of things and
like the consumer, the B2Cconsumer than anything else.
Sylvain Giuliani (10:35):
I think for sure like the small companies
reverse resource or econs, likethings like that.
Like they're getting destroyedbecause again, it was so easy to
just inject the Facebook likeyou know, tracker, right?
Like that's the thing's like,yeah, you don't have to be an A
genius.
You copy paste this thing.
Or even worse, sometimes youhave a direct integration.
You don't even know what'shappening, but you's saying to
track, right?
Yeah.
They're getting impacted.
(10:56):
I think like that.
Because that's going away.
It makes it harder for even likelarger company to do quickly
deployment.
Like, you know.
Now if I, like, I mean, let'stake an example.
We, the three of us, we work ata normal size, like B2B company,
300 people, right?
And we're like, Hey, you knowwhat?
We are gonna start doingadvertising on TikTok, right?
Like, so.
(11:16):
Now we can't just put the, thetracker, right?
Like now it's like, okay, weneed to sync our audience
somehow.
Like, you know, so, because wewanna retarget, so how do we do
that?
Okay, we can do a CSV uploadonce in a while.
Okay, that sucks, but I can comelike do it, but it sucks, right?
And then it's like, but really Ineed to send like conversion
event to, to, to feed themachine, the algorithm
technology.
(11:36):
Hey, this is good audience.
Like keep, keep showing the adsto the right people to train
their algorithm, right?
And then to do that, you need tosend event again.
You can do the Monday morningCSV upload, right?
Like it's okay, but like youdon't have that ease ness of
like just, yeah, just plug thesystem together and you just
work.
And then you're like, oh, whereis my conversion even coming
from?
Right?
Like, okay, some of them arecoming from spo, like you said,
(11:59):
those trackers are not goingaway.
If you're like, Hey, a demo formrequest, like, you know, SPO.
Credit tracking them form, likefor the mission, like nobody's
gonna tell you to take thataway.
But again, if you wanna say,Hey, I'll give you some example
of our customers, right?
It's like, hey, they, they showawareness ads.
So B2B plg company, right?
They show awareness ads or likefeature that you are not using.
So they're like, Hey Mike, weknow that you are notion.
(12:22):
User, right?
Where you're our user audienceand, but you're not haven tried
like our new AI feature, right?
For example.
And so they're gonna retargetyou aggressively everywhere
until you stop, uh, you startusing the ai, right?
And so that's kinda like the,those dynamic audience, this is
where it's get more complicated.
And before it was stillrelatively easy to do with the
third party tracker.
(12:42):
Now it's just impossible withoutlike, Some infrastructure or
help from data and engineering,right?
So that's, that's, I thinkthat's how they get impacted.
It's not like it's impossible,but it's just like an extra hoop
that they're to jump for thatthey didn't
Michael Hartmann (12:54):
have to do it before.
And I, and I don't, I actually,since I've, I actually am the
one here who's works at, hasworked at larger companies
currently working at a largecompany.
Like I don't think there'sreally much differentiating
their, you, you were headingwhere I expected you to go.
So, which was when Mike askedthat question, I suspected the
issue was actually how do youpush that?
First party data to thoseadvertising platforms now to do
(13:18):
the targeting in different ways.
And it sounds like that'sreally, that's gonna be the new
challenge because we're, they'regonna be dependent on general
stuff that they will probablystill try to package and keep to
the, you know, but it's gonna berelatively, if I remember right,
right, it's gonna be chunks of afew thousand people as opposed
to, you know, hundreds or dozensor, or, you know, whatever is
(13:40):
the right appropriate amount foryour, your business.
Mm-hmm.
Sylvain Giuliani (13:43):
Yeah.
And, and to me there's like thiskind of like duality, right?
It's like as marketer, like, youknow, you go to any marketer in
the world and you say, Hey,we're gonna build like agencies.
Like, nobody's gonna be like, ohmy God, this is a novel concept,
right?
It's more like, how do we build.
Better audiences.
Like, you know, I already have ahundred, but like I need to go
even further.
I need to press more.
(14:03):
And so it's like, I want, I knowI wanna do it.
What's stopping me is likeaccess to that data.
Right.
And so you took some of some ofthe things away from me by
having like this kind of thirdparty coming away or like, Like
that, kinda like third partyside of the house.
And so this is also why you seea lot more first party data
being used, right?
Because like, oh, I want to, Iwanna do a, no, I wanna do a, a
(14:24):
targeting that is like, youknow, location based, usage
base, like pre, like a likecalled purchase base, like yada
yada.
And so it's like you, youstopped talking to people.
I mean, I, I was talking tolike, uh, a large gaming
company, right?
And.
For them.
It's like we have audiences inthe order of magnitude of like
hundred thousands of audiences,right?
And it's just like everything islike programmatically done and
(14:45):
it's all powered by first partydata, right?
Like, and that's the, and youknow, and I asked them, yeah, I
mean that sounds like a lot ofwork.
Is it even worth it?
Like to go that big?
Right.
Uh, like in terms of like, forthem it's worth it because they
have large user base and it'snot about.
Reaching to people, they canreach anyone, but it's like
reaching to people to rightplace with the right message.
Right.
And that's the only thing, like,that's the power of first party
(15:06):
data is like, they know thatseal plays on, uh, Xbox and like
School of Duty, right.
And Mike, you more like aPlayStation person or a PC
person.
And so, and you like in like,you know, AARD game or something
like that of them, and they'llshow you the right ad in the
right place at the right timeinstead of like spending the ad
dollars showing ads toeverybody.
It was like, I keep seeing it.
A for wall of white, I'm a callperson.
(15:27):
Right?
Like that.
So that's the
Michael Hartmann (15:31):
how, how, how like, so my understanding, cause
I haven't had the opportunity toreally work with them a whole
lot.
But these intent data platforms,right?
Are they assumed they're pretty.
Relying on third party cookies.
Is, is that true or is it like,are they gonna be like, affected
by this as well?
So, so in other words then our,the marketers who are relying on
(15:51):
intent data for, you know,whether, whether they're using
for lead scoring or whateverthey're using it for, to.
Prioritize targets or whateveryou think they're gonna be
affected.
I think it depends
Mike Rizzo (16:01):
on where your intent data is coming from.
Total.
Like it's a total like reverseIP lookup play.
And for people from working fromhome, that's a challenge.
Um, but people going back to theoffice, that's helping.
And then I think it's also like,like G2 for example, you know,
the review platform.
They've got a crap ton of intentdata that they can signal out
(16:21):
back to people, but that's theirfirst party intent data, right?
So like there's value in thatdata.
Sylvain Giuliani (16:27):
Um, yeah, there's not too much issue
there, like intern data.
I mean, we can go into the pageslike, is it even worth investing
in this?
But like, uh, maybe it's a bigfan of a day.
Mike Rizzo (16:38):
Forget the same subject, throw like ro, like,
you know.
Yeah.
Uh, on those folks, but maybe Icould get some intent people to
come on and tell me about why itworks.
Sylvain Giuliani (16:49):
Yeah, yeah.
I think it works well for a verythin slice of the market.
Yeah.
And it works extremely well forthem, but everybody else, it's
kind of like us.
A money pit that doesn't bringyou anything.
Uh, that, that's my, that's myhot take on it.
Uh, but I think from anidentification point of view,
there's no issue cause you justneed an identity key, right?
(17:09):
Like if you use G2 as anexample, Here's, like G is gonna
be like, Hey, this companycalled uh, accent.com is looking
to buy like a warehouse, right?
Like, and so the only thing thatyou need is like this identity,
like the account, website ordomain, sorry, campaign.
Like that's the thing.
And it'll be the same forpeople.
It'd be like, Hey, this silk isdoing something right.
(17:32):
And so as long as you can matchthose two identity, you know,
that's, that's okay.
That's okay.
And that doesn't really requirethird party data.
I mean, you could like, youknow, using like, Cookies.
Like CrossFit cookie, but like,that's going away.
But like I think nobody reallyrelies on that anymore.
I mean, as far as as
Mike Rizzo (17:49):
I am.
Yeah.
Oh, I think what's interestingtoo, and I know we've got some
other things we want to diveinto here, but around this, like
the innovation that's gonna takeplace on, on the platform side,
right?
Like not, not the marketingautomation, technology side and
the 11,000 vendors there, butmore on the ad tech side.
Right.
So like you see there's likeLinkedIn lead forms, right?
(18:09):
Like yeah, they're going to usethat as your conversion signal
now.
And they may even, like, there'sa very, I could see a business
case where platforms likeLinkedIn and Facebook and all
those other providers are gonnaoptimize for you to have, uh,
potentially lower cost perconversion to go use those ad
platforms that hook up to theirown lead conversion metrics.
(18:30):
Versus trying to send you offplatform to go track a
conversion and do the whole backand forth thing, they're gonna
be like, no, you know what,just.
Just get your lead form here,right?
Yeah.
And then they're probably gonnabe like, you know what, just
upload your list.
In fact, take your data, uploadyour list over here, and run
your email campaign out of ourplatform.
And it's like, so like I, I feellike there's gonna be this big
(18:52):
shift in MarTech around theoutbound marketing capabilities
on the platform side.
Sylvain Giuliani (18:58):
I mean, like you're a hundred percent right?
I mean, if you look at LinkedIn,Add performance, right?
Like, I mean, you run exactlythe same ads.
Sames one is a leading lead genform, right?
So form on LinkedIn, you neverleave the website.
The other one is the same thing,but you click, you have to go to
an earning page.
Like the leading legend, likeobliterates, like the conversion
rate of the, of site conversion.
(19:19):
Right.
And, and you know, like LinkedIntell you straight up, like you
should do it like the, the ui,like they make it on purpose
that the non formm converseless.
The buttons are smaller, theclick zones are smaller.
Right.
Like it's, it's made through foryou to direct you to.
No, just keep people on.
It's fine.
Yeah.
It's gonna be totally, you know,like of course.
And
Mike Rizzo (19:39):
then forward tracking is there, right?
Like Yeah.
It's already happening.
And, and all the conversiontracking is happening on their
platform, which means it's likethere's this like, uh, false
sense of security that like, oh,cool, like I've got all the
things that I need.
But to your point earlier, whenyou want to start to layer in
the deeper segments of like,well, I really need to target
somebody that hasn't used X, Y,Z feature.
(20:02):
And as in a particular regionmm-hmm.
There's no way to be able to dothat connection anymore, which
is why, you know, marketers needto be paying attention.
Right.
Michael Hartmann (20:10):
So, okay.
So how, how will, how will the,and this, this may be, I know,
so, so you and I were talkingbefore we started about like
Aiche P t and how that, I thinkthere may be a tie in here, but
like one of the things we allwant to get to with our
marketing teams and marketingops teams is to.
Better understand like what'sthe preferred channel that
(20:31):
people want to be communicatedthrough, right?
Uh, or that performs best interms of whatever it is we're
trying to optimize.
How, how, how is this?
I guess two parts of it.
How is the, the cookie situationgonna affect that, do you think?
And then second, is there, like,you're describing all this sort
of intelligence, but it's builtinto the, to the, uh, the
(20:51):
platforms like LinkedIn orFacebook or whatever, or the,
you know, us having data andthen trying to deploy it in the
right way back out to them.
Like how, like is there a playfor AI in this somehow?
I don't, like, I don't know whatthat looks
Sylvain Giuliani (21:02):
like.
I think there's, I think youdecomposed like the problems in
like free, in free bucket,right?
It's like one is like, How do Icollect as much data as
possible, which is first partydata.
Like, so even like, you know,identifying prefer channel, like
it's, you need to know whichchannel works well for each
individual, right?
So again, if you think in termof first party data, it's like,
well, I know SI is a human.
(21:23):
Okay.
He does his email address.
Is this Id.
On different platforms.
So I've, I've sold identityresolution across as many
channel as possible.
And now I correct.
Data was like, oh, he consumeemail, he opens email, he
replies to email, he click onemail, right?
So a lot of first party datastream coming from that, and
it's like, well, I know email'sthe cheapest channel I have, so
I'm gonna use that first partydata to actually exclude sale
(21:44):
from all of the profe, like overchannel.
So I'm not gonna show you ads onLinkedIn because, I can get to
him for free Via emails.
Right.
Or via app messaging.
Right.
Or or whatever platform youhave.
Right.
Uh, then if you think aboutlike, Uh, third party platform,
ad platform or even like socialplatform, right?
Like lot of committee, like lotof things happen in Slack
committee now.
(22:04):
Like, you know, I think likethere's something called
marketing ops mobs, PLAcommittee, something like that
out there, right?
Uh, and it's like how do youtrack the, the engagement where
it's like you have to do.
Deep link, like UTM trackers,things like that.
So again, like you can collect,oh, this link is unique to seal
because he has like his ownlike, like, uh, anonymized ID
and he clicked on TikTok, right?
For example, on the dm on TikTokor WhatsApp, message on TikTok,
(22:26):
right?
So it's like I can get that databack and be like, oh, looks like
seals a lot on WhatsApp orwhatever platform you took
about, right?
And so, so I think that's one,like it's collecting that data.
Second point is like, how do wemake sense of that data, right?
Like, so I think this is wherehistorically the best
Michael Hartmann (22:40):
companies that, that, that is the big
question in my mind.
Sylvain Giuliani (22:43):
Yeah, historically it was gotta go
hire a bunch of data scientistpeople, essentially.
Right?
Or like, you know, during thekoolaid of one of the many,
mark, mark tech, advi, uh, youknow, like, uh, marketing
message.
It's like, no, just give us yourdata.
We put in the black box, insightcomes out.
You don't know how it works, buttrust us.
That's how it works.
(23:03):
Like, you know, that'sbasically, but again, it costs
you money.
I think right now what'sinteresting is like, you know,
GPT is trying to be very good.
I mean, as a technology,something that is very good at
like pattern matching.
Uh, identify things like I'mrunning like myself right now.
Just experiment like using G P Tto do identity resolution, for
example.
And that works extremely well.
Like, you know, like it can,like, it can detect that these
(23:25):
two humans like are almost thesame as like, Hey, we should do
something with it.
We should match them with a highlevel of confidence.
Right?
And so I think again, likethere's, there's gonna be a way
to repackage this kind of like GP T technology to kind of like
make it like, As, as a featureof a product.
I'm not saying like you're gonnago and charge PT and be like,
Hey, what's the differencebetween these two mics?
Kind of thing.
But like, it's gonna be veryeasy to build this feature, uh,
(23:46):
to make marketing life easier,right?
And then the third, the thirdbucket, right, is like, cool, I
got my data, I got my insights.
What do I do with it, right?
Because, uh, like how do Ideploy, how do I activate my
data in marketing speak, right?
Uh, and I think that's the,that's the next thing.
And so, again, like these freebuckets, like you need to be
exceptional at these threethings to be best of written
(24:09):
marketers nowadays, like, youknow, you like if you can be
good at one.
And that gets you ahead of thepack.
But like the best in classmarketing team I've, I've met
are, you know, they are worldclass at these things.
Yeah.
All of the data.
Super accessible, highconfidence.
They generate tons of this sideof it and they know what to do
with the data and they deploy itin like super tactical
advertising campaign, emailnurture campaign, like, you
(24:30):
know, empowering your sales teamwith insight to be like, Hey,
this account is, you know, thisaccount is at risk, this account
is ready to up sell.
Things like that.
Like that's, that's what thebest marketing team team do,
team do nowadays.
Sorry.
Um, but that's my take.
But that's how AI could fit intoit.
Michael Hartmann (24:45):
Um, wow.
So I, I bet there's a lot ofpeople out there listening like
me, where, where we know we havegaps or challenges with our, our
data.
Either it's not complete, notnormalized, right.
On key things that we typicallywould use.
Could, could something like thathelp with data cleansing kind of
(25:06):
stuff as well?
Sylvain Giuliani (25:08):
Yeah.
Well I think that like, this iswhere it depends what type of
company you have.
Like, you know, I thinkhistorically humans are still
like, you know, like if you tookyour data engineer or data
analyst, like they're gonna tellyou 60% of their job is like
data transformation is a niceway to say it.
Right?
Cleaning, right?
Yeah.
Yep.
And, uh, I think that's notgonna go away anytime soon.
(25:28):
I think, you know, like AI,things like, that's gonna make
that life a lot easier.
Like, I mean, I mean, go to J GB T right now, past like 24
numbers with completelydifferent format and you say,
put all of that into likeintentional formats, you know,
ease, o standard, whatever theZO standard.
It just does it right?
Like, you know, same thing pastlike the company name of like
(25:49):
one with ink, one with Ltd, onewith.com and say, just put them
into normal company name, humanreadable.
It just does it right?
Like, so I think that's going tohelp you tremendously, like
cleaning your data.
Uh, I don't think it's, it isstill ready to like, transform
your data and make sense of it.
Michael Hartmann (26:06):
It's not gonna, it's not gonna do it in
an automated way.
You still have to prompt it and,Give it the clue.
No, you can,
Sylvain Giuliani (26:10):
you No, you can, you can prompt it.
Like, I mean, you can, you canjust be like, this is my role,
you know, company name.
And then, you know, the otherone is like, you know, use G B T
to clean it by making like humanridable and it just kind like,
just keep doing it over andover.
Getting your data set, likethat's like, like these tools
already does it.
It's, and more like, this isjust gonna be a new normal, like
(26:32):
for most of the tools.
Sure.
Because it's so cheap andapproachable to
Michael Hartmann (26:34):
build things in this Yeah.
And the other, like the otherthing that, so I bet.
The, the, the first thing thatcame to mind when you talked
about like matching twoidentities from different places
that look like they're a matchwas, um, one of the challenges I
think everybody runs into isyou've got a really good client
at your, at your company.
They move to a similar role ora, you know, to another company.
Like how do you follow them soyou can go pursue them as a new,
(26:57):
new cli a client at thatcompany?
Uh, that may actually bepossible, if I'm understanding
what you're describing.
Sylvain Giuliani (27:05):
Yeah, I mean it's, it's possible today,
right?
Like, I mean, there's a bunch oftools out there, gem, you know,
APU does it like, I mean, youknow, it is at the end of the
day, it's like you are gonna belate when I think there's two
things, like knowing that personjoin, right?
So it's like, You know, it goesinto product building.
But like, you know, LinkedIn isa good place you can spare, plug
in, you know, you can, uh, Iwait for an email to bounce
(27:26):
like, you know, I send an emailto Mike user, right?
And like, oh, bounce, I guessit's not there so I can at least
understand it's not at thiscompany anymore.
Right?
Uh, so I think that's gonna be,that's a different set of
product.
Like I think this is not whereAI gonna be able to help you,
like magically, cause it needs adata set.
I mean, if you wanna checksomething, it's called people
GBT io.
It's got announced yesterday andit's like, A chat G P T
(27:50):
interface that you can create alot of people and can be like,
Hey, show me all the marketing pof people.
That looks like Mike Riso.
And I also have like thousandsof followers on Twitter.
And it's like, or like changejob.
Recently I just give you overlist.
Right?
It's like, it's like, oh, humanreadable or like NLP to, uh,
query a database of people andprospects.
Mike Rizzo (28:12):
Wow.
Crazy.
Michael Hartmann (28:14):
That's amazing.
Like I'm, I am actually sort ofspeechless right now.
I, I, I can't, I can't decide ifI'm, I'm amazed or I'm, or if
I'm frightened.
So
Mike Rizzo (28:26):
I don't know.
I mean, having come from theworld of, like, I started my
career in ad tech, um, and likeonce I figured out how all this
stuff worked, and of coursewe're like, we're literally in
the midst of a conversationabout how that whole world is
changing right now.
But like, I don't know.
When you go from, from that towhere we're headed now, I'm
like, should we really be thatsurprised?
(28:47):
Like people
Michael Hartmann (28:48):
pretty, pretty,
Mike Rizzo (28:50):
you know, in, um, I don't know.
They're engineering new outcomesnow to try to solve for these
problems that, you know, they'relike, well, I want it to be the
way it used to be, so I'm gonnatry to find.
I'm trying to find my workaround
Michael Hartmann (29:05):
on all that stuff.
Yeah, well, no, I, I'm not, I'mnot surprised either cause I
started my career in marketing.
I went from doing financialsystems and warehouses to
building a consumer database.
And this was a long time ago, 30years ago to call it.
And even then, right there wasa, like, when I saw what was
(29:26):
available to almost thehousehold level, I was just
stunned.
And then it's only gotten.
More robust and more datapoints, more, more activity and
all that.
Mike Rizzo (29:39):
Yeah, I
Sylvain Giuliani (29:40):
mean, I mean, yeah, I take like Google,
Facebook, the more first partydata you feed them, the more
they learn and the more theybetter target.
It's like you, it's kind oflike, uh, uh, an ai, but the
buzz was not made around it.
And for good reason.
You know, they want to say like,oh, you know, we have this AI
that allows you to targeteverybody.
(30:00):
It's like, that sounds sketchy.
Uh, but now the cat, the cat,the cat is out of the bag.
It's like, So this G P T thingis kind of like what we were
doing with ass, like, kind of, Iguess you can say that, you
know,
Mike Rizzo (30:13):
what was doing
Michael Hartmann (30:15):
hy hyper targeting at the individual
level.
Right.
It's just, that's yeah.
It's, and it's and good.
Yeah.
And like really good at it.
Mike Rizzo (30:22):
Yeah, it's very good.
It's, yeah, it's really scary.
Good at it.
Um, I, I have a question for youstill on just sort of your
thoughts on like, Yeah, I, I'vebeen, I've been tussling a bit
with, um, you know, I think,okay, so the underlying message,
or maybe the direct message ifit was said, uh, very plainly
(30:44):
is, look, you, you need to knowwhat to do with the data.
Right?
We just said, we just got donesort of talking about it.
Um, the, my challenge though islike, is that, do you, do you
feel that there's a role today?
Like that actually tries to takethat responsibility on.
(31:07):
No, like you actually said, thebest of breed marketers figure
this out, but like, is that justa new marketer that we're
talking about, like a new futurekind of marketer?
Or is it the data team?
Like are they gonna become quasimarketers?
Like, I don't know who'sresponsible for this.
Sylvain Giuliani (31:23):
I think, I mean, that's a great question.
Like, you know, it's like who'sgoing to eat wood?
Like, you know, there's alwaysthat debate is like, is data
team just gonna become likemarketing team or is marketing
team gonna absorb data team?
Right?
Or like ops teams or whatever,right?
It's like there's always this, Ithink it's just a new skillset.
I think if you just today theproblem is that I see in most
organization is like, It's somarginalized, right?
(31:44):
It's like, Hey, I'm a dimensionperson.
My job is to build campaignmessaging and like the, you
know, promote offers, right?
Where it's an ebook, an event orwebinar, whatever.
And so it's like my, I, I focuson such a spin slice and I, and
then simply, you know, marketingapps is like, well, I'm just
operationalized what thedimension person wants.
Like, you know, they wanna do awebinar, they wanna put a
(32:04):
nurture before and after, so I'mjust gonna do the thing that
they told me to do, kind of likeas a system, system level,
right?
And.
And I think because of this kindof like ultra ization, like
nobody's kind of like thinkingabout the overarching, like what
is the, what is the customerjourney and what the data fits
into this whole thing, right?
Like I think that's like, uh, Ithink there's no role for that.
(32:25):
And I think this is kind oflike, so because of that, It's
the new skills that everybodyshould have, right?
It's like, uh, you know, I wastalking to, I would not name
this company, right?
But like they, by design thisultra data savvy and automation,
they do data and automation,everything, right?
But they by design, never hiremarketing people in marketing,
right?
It's like if you have donemarketing before, you are
(32:47):
corrupt, right?
And so therefore, oh yeah.
Therefore you can't work for usbecause you don't have this
kinda like data automationmindset that we need, right?
Like, and they're like, We can,we can teach you marketing and
messaging and things like that,but we can't teach you being
like data curious and like thiskind of computer science
engineering mindset.
Right.
(33:08):
And anyway, I'm not saying it'sbad or right or wrong, but like
Right.
It's a different approach.
Right.
It's like you need that skillset, right.
Like this, a new, new type ofpeople.
I agree.
Michael Hartmann (33:16):
Yeah.
I, I, I mean, I could argueabout whether or not they sh
they could also teach some ofthe.
Data analysis stuff too as well,but like that, but the point, I
guess
Mike Rizzo (33:25):
there's, yeah, but like there's a difference
between, I, I do think you canteach somebody like the, the
what and the how goes into towhat SI's talking about from
that team.
But like, um, there's, there's adifference between like systems
thinking, And like dataorchestration and, and data
architecture and then liketherefore architecture that like
(33:48):
goes into that and like objectorientation and how those things
all sort of work together.
Like there's a differencebetween that.
Like, oh, I need to make sure wehave access to this data in this
structured way.
Like I.
Because it can get concatenatedand then we're gonna be able to
turn it into a string and the js o response, we're gonna be
able to do da, da, da, da da.
There's a difference betweenthat level of thinking and
(34:09):
someone who is like, who'strying to activate on that data.
And then somebody who's like,okay, like if I showed you this
j s o response, do you know whatyou're looking at?
And they'd be like, I have noidea.
And it's like, could you figureout if we, if we told you that
we could do such and such withthe data, what would you do
next?
Um, there's like, yeah.
Cause there's a, there's a verybig difference between those
(34:32):
types of it's
Sylvain Giuliani (34:32):
systems thinkers.
Yeah.
Someone told me it's like, youknow, it's like knowing
impossible is, is not, is, youknow, it's nothing as you think
tagline.
Right?
It's like there's a lot, likethere's sort of marketing people
are there that just don't knowwhat they could do.
Right?
Like it's just, It's justbecause it's part of the show
them.
It's like, yeah, exactly.
It's like, that's cool, but Idon't know what I can do.
(34:54):
Again, I'm not saying it'severywhere, but like, it's the,
the school of thought, the waywe were, are trained is to just
focus on certain things and likethis kind of like, Way of data
that that is coming is like, ifyou're not forced to use it, you
can kind of coastline and justbe good enough.
Like, you know, and, and it's,it's nothing
Mike Rizzo (35:13):
bad about
Michael Hartmann (35:13):
ok, so, oh, so you're hit, you're hitting on
where I, where I was gonna takea little bit issue.
Like, I, I get what you're bothsaying.
I, I still think there, I'm notgonna say everybody, right, but
there are gonna be people andsomeone, this comes from my own
experience like, When I gotoutta college, I actually just
told the story recently.
Somebody else, uh, we worked atPricewaterhouse.
They sent you off to for threemonths to get trained in their
(35:34):
way of doing at that time,cobalt programming, right?
Mostly right.
But how you did nice consultingwork for them and they had
cohorts of people all the timegoing through this training
center in my cohort were, waswith this one lady who had just
graduated from Duke with areligion major and.
(35:54):
Teaching her how to code waseasy, right?
cause she knew how to learn.
So I, I think there's like a, Ithink there's two things you
need sort of baseline, right?
You need a certain level of, uh,aptitude, right?
And knowledge and ability tolearn and the desire to learn.
Yes.
And then you need, then you needattitude, right?
You need, you need to be like,you have to have the desire, you
(36:15):
have to be coachable.
You be willing to try stuff andfail and learn.
And I think that's, that wouldbe the missing piece to me.
Like if somebody was in thatmore traditional marketing and
wanted to move it to the kind ofrole you're talking about.
That, that's, that's why I waslike resisting that because I
think in a lot of cases weunderestimate how much people
could actually learn, even ifit's outside of their, the
domain they think of as theirexpertise.
Mike Rizzo (36:38):
No, no.
I, I think look, everybody canlearn something.
Right.
Um, I think to that, to thepoint of, uh, what you were
talking about with that teamthat says like, Hey, you've sort
of been corrupted.
Yeah.
I think that's a little bit, youknow, extreme potentially, but
um, I think there is adifference between.
Where like si you hinted at itjust a second ago and
(36:59):
interestingly, it's a comment,it's a conversation I was having
with, uh, Lon Mendoza, whoyou've spoken to I think before
Si potentially.
Um, he's the, the, the founderof T M W, the MarTech Weekly.
And so we were having ourmonthly catch up last night and
um, and one of the things wewere talking about was like,
(37:20):
The, we both recently have sortof been gone out into the fray
on like, Hey, we're just gonnatry to figure out how to make
ends meet.
Right?
Like, I was put into a situationnow where I'm like, all right, I
gotta do marketing ops.com, someconsulting work to keep the
lights on.
And, and he went out and decidedto fully commit to his role.
And, and so we're, we're reallyentrepreneurs at this point, and
so, mm-hmm.
(37:42):
Um, we were talking about thisidea of.
The employee life, right?
The, the employee balance oflike where you're working and
what you're doing.
And, um, you know, I like, Ipersonally feel so much better
(38:03):
about like being out and tryingto hustle a little bit for
myself.
That there's like, there,there's this very dramatic
difference between.
Today's payday.
Cool, right?
I'm, I'm getting paid today andI'm gonna go to dinner tonight
and I'm gonna take the kids to aballgame this weekend or
(38:24):
whatever.
And then there's this other sideof it, which is like, I.
Man, I don't know if I'm gonnaget a paycheck this week cause I
didn't.
I didn't do enough work forclients or whatever it is.
And so going back to this ideaof learning and are you going to
take the time to figure out theart of the possible and try to
push yourself forward, yourorganization forward and all of
(38:46):
those things, like, it sort ofcomes down to this like, Are you
lazy or not?
Like are you motivated or
Michael Hartmann (38:55):
not?
I, I don't think, I don't thinkit's really, I think in some
cases it's not laziness ormotivation, cause I know lots of
marketers who work really hard.
I think one of the things thatgets in the way of a lot of
marketers with doing differentstuff than they just like rinse
and repeat kind of stuff isbecause there are put up.
Um, sort of, uh, gates orhurdles, right, with getting
(39:18):
stuff to market, right?
So they, they're, there's a,there's a lot of effort that
goes into getting somethingright as opposed to good enough,
right?
So you said that, or I think itwas about what you said and
quickly seeing if it works,right?
So that, I think that's one ofthe things I've see, especially
at bigger companies where you gothrough, yeah, so many levels of
review and approval, and then ittakes, people are wondering
(39:40):
like, why can't we just updatethat webpage?
So why does it take so many?
So, A week, right?
It's not because you, it's nothard to update, but people are,
they're afraid of like, it'sgonna be off brand, or I'm like,
like I tend to lean towardslike, just go right and do
(40:01):
enough to make sure you're notgonna do something that's really
egregiously wrong.
And even if you do that, likelearn from it.
Don't do it again.
Yeah.
Yeah.
Sylvain Giuliani (40:09):
It, it, I'll, I'll kind of summarize what you
both said.
Cause everything, it's like,it's a combination of both ways.
It's like you need to write theright attitude, right?
But you also have to have theright context to get the right
attitude.
Like, you know, you can bechilling at a company and, you
know, do a good job and youlike, I, I'm chill.
I'm doing worklife balance andgoing on the weekend, uh, fun,
right?
And then to your point, Mike,you're like, I'm an entrepreneur
(40:30):
here.
Like, you know, I, I got, Igotta make it work.
Like, like that.
The context forced you to have adifferent attitude is either
you're going to.
You know, do it or becomecreative and figure it out, or,
you know, you'll just fail andthen you know, you'll go try
something else because you, youcouldn't like have the right
attitude using your word here,Michael.
Right.
I think that's just, that's justa cycle of life and I don't
(40:51):
think there's, like, everybodyshould be, always be at full
speed and things like that.
It's just, oh goodness.
It's like, well, not fullpeople, but like you can just,
yeah.
Using the examples, like you gothere, you learn a ton of
things, you learn how toprogram, that's great.
Like then, you know, you go trysomething new, you learn
something and then, you know,maybe you don't learn something
for like three years.
That's okay too, right?
Like and then, yeah, I think theproblem is like when people stop
(41:12):
learning and people stopadapting, like, you know, so
many.
You so many people, AmazonMarket across the board, like
you said, what was the last timeyou learned something to uplevel
all your skills and doesn't seemlike take a course or something.
It's like you learn at your jobworking with a new peer, new
colleagues, things like that.
And like, actually I've beendoing the same thing for five
years.
You know?
And I feel like that's like,that's the problem.
(41:33):
That's the real problem.
And these like, that's likewhere you get like swept away by
people who come in.
It's like, I've been using G PT, it's my co-pilot.
Not, not saying they replacingpeople.
I think G P T thinks a lot like,you know, uh, I copilot an
enhancer to people, right?
And it's like, you just, like,what, how is this person writing
so much ad copy and the A copyworks better?
(41:53):
Oh, they just brainstorm with GP T all day.
Right?
And, uh, they as are performingbetter than yous just because
they can include it quickerbecause they have this assistant
to help them iterate.
Like just simple things likethat, right?
Like get them ahead and thenit's like one fixed it to the
other.
It's like, oh, why is the, whysales doing better?
It's like he's using intent datafrom blah, blah, blah.
Right?
And you know, like that's the.
(42:13):
That's the mentality, uh,that's, that you don't want to
sink into is like the personwho's like, I haven't learned
something new in five years.
Yeah, yeah.
What was, I mean, the colleagueor,
Mike Rizzo (42:21):
or anything?
Definitely not.
I, I, I, and then the otherpoint that like, so where I
branched off of for a moment waslike, hey, if, if, if someone
understands, you know, the artof the possible when it comes to
sort of data orchestration and,and what you can do with it, um,
(42:41):
So sometimes it's like, youknow, to your point, Hartman,
like, uh, there might be a lotof like roadblocks in the way to
be able to get something done.
And so I, I could, I can arguethat that's actually just a
factor of laziness.
Like, I just have to, I have todeal with all this stuff and I
don't really want to, uh,depending on the person's
(43:04):
response to that situation.
But if you, if you were.
If you were presented with achallenge and someone said, Hey,
I really need to try to figureout how to do this thing, and
you now know how to take datafrom one place and move it to
another, and you're this newkind of data marketer, but then
you recognize openly like that'sa lot of work.
That's, that that actually fallsoutside of the norm of the way
(43:25):
we normally do things at thiscompany.
Yeah.
Um, do I really want to trythat?
Is that really, is the companygoing to like that?
Is that for me?
Is it for them?
So there's all this, like youbecome this entrepreneur.
It's like, I can solve thisproblem.
I promise you I can solve thisproblem, but like, no one here
has done this before and so I'mgonna forge out on this path.
(43:49):
That like feels very different.
And that is where I was like,that's what branched off of.
If you start learning this newskillset and you figure out how
to solve these really uniqueproblems in unique ways that
don't fit into the quote unquotebox of how things are normally
done, you're going to run upagainst hurdles and then you're
going, it's going to take veryspecific.
Individuals who want to try toforge that path
Michael Hartmann (44:11):
forward.
Well, it's, it's permission.
It, it, what you're describingis like swimming upstream
against the, the idea that you,we should all be following best
practices.
Yes.
Right?
Yeah.
Because fallacy, by, by, by bydefinition, that's doing what
everybody else is doing.
And that's like not a recipe forlong-term success probably.
Um, so we're, I didn't, this hasbeen a great conversation.
(44:32):
I want to kinda get back to the,the, not cookie per se, but I, I
think.
I keep coming back to, I've feltfor a long time that like just
within marketing ops, but let'sgo to broader, to marketing in
general.
If there's a gap in, um, skillsand experience of knowing how
to, um, really.
(44:53):
Take advantage of data that'sbeen collected and um, it feels
like that's still gonna be, notonly is that still a problem,
that's gonna be an ongoingproblem if we don't address it,
especially as we continue, like,as we're developing even more
and more data through firstparty cookies and things like
that.
What, like, do you see, likewhat, how would you suggest for
our listeners that they start toaddress it?
(45:15):
Is it skill up leveling?
Is it looking for specificpeople?
Like would it.
I, I, yeah, I have, I don't havethe answer.
I just see it as a problem andI, I don't know how to solve it.
Sylvain Giuliani (45:25):
I think it's like being curious about it.
Like if you know you have thispain, there's a lot way to solve
it.
I think, you know, I, my, theway I've learned data is like, I
was friend with someone at workthat was in the data team, and,
you know, I had like all becauseI had all my questions about
data and he was the person Iused to go to, and so, Bit by
bits I started to learn aboutthe concepts of data.
(45:45):
Right.
And then, you know, I was like,I just don't wanna talk to him
anymore.
Like, I don't want him to be abottleneck.
Right.
And so I started learning.
And so like, cause but I, I haveenough information about the
world data that I could startlearning by myself.
Right.
And you know, after that you,you take a course, you, you do
some training online, like tolearn SQL to start query in a
new practice.
And it's, it's not thatcomplicated, right?
(46:07):
Like I think it's, but it startswith like, Having, being
curiosity and having a need.
Like if, I think if today peopleare like, I'm just gonna go this
weekend and start learningsequel, and you have no purpose
for it, like it's gonna be, soit's, it's gonna be a troll to,
to learn something.
And you don't want learning tobe a chore, right?
Like, I think we, we have otherthings to, to worry about that.
Then that's, and so that's wherewe approach, like I will meet up
(46:29):
someone in your team that ismore data savvy than you work
with them, find interestingproject to work together.
Go learn how to.
Query data.
Don't become a data engineer orsomething, but understand how
data Yeah.
Is model, like, how to query it.
Like what a big core concept.
Right.
And then finally, I'll, I'll dothe G P T plug, but like, you
know, the rate of things aregoing, like you can start like,
(46:51):
I mean, I use it.
Myself today, I can write prettygood sequel, but like sometimes
I'm like, ah God, I just gottaknow how to do this.
And I good.
I know enough that I can ask gpslike, can you write me a
windowing function that quitethis data, this is my data set.
And then it just quite a data, Iknow enough to know that what is
giving me is correct, but likeI.
I rely on my good friend, g p t,to, to help me do the work, you
(47:12):
know?
Uh, and I think that's, youcan't just start there, but I
think in the future it will helpyou.
Uh, you might be able to startthere, but I think that's,
again, it's a good acceleratorassistant
Michael Hartmann (47:23):
that that's my, well, and, and, and I
suspect that there's a number ofour listeners, audience or
whatever that um, don't have acolleague who's a data expert,
so maybe also going intocommunities, right.
To find others and just.
Yeah, look for like, it'samazing how much people will
help you learn.
Mike Rizzo (47:41):
Yes.
Sylvain Giuliani (47:42):
I mean, I use communities every days, you
know?
Yeah.
Every time I have Salesforceproblem, you know, God knows we
have, uh, so many all the time.
24 7.
You know what, it's, it's builtlike it's built.
It's built that way on.
That's their business model.
That's where the community,that's where community come,
comes in, you know, it's like,Hey, how's the.
What's the best of the 20million way I could do it in
(48:04):
Salesforce?
Tell me what's gonna go wrong ifI do it.
This one like you, that's whatthe C comes in the shell
Michael Hartmann (48:08):
experience is Eva evaluating the tradeoffs
very quickly, like crowdsourcingit.
Sylvain Giuliani (48:12):
Yeah.
Yeah, exactly.
Mike Rizzo (48:15):
Do it all the time.
Michael Hartmann (48:17):
Yeah, it's, I I, I do too.
Um, well, I, it feels like wecould go on and on here, but I
think we're gonna have to, towrap up here.
So, so thanks for joining us.
This is, uh, so this is one ofthe ways that I learned still
and stay current, is just like,just having these conversations.
I, I think I learn somethingevery time.
But, uh, so thank you so much.
(48:37):
If folks are interested in kindof keeping up with what you're
doing, the communities, etcetera, et cetera, what's the
best way for them to do that?
Um,
Sylvain Giuliani (48:45):
I'm on LinkedIn under my name So easy.
Uh, you know, like every goodmarket, I mean they're replying
to dm, you know, our socialnetwork.
It's the hassle I'm at there.
My email was Sylvain getsensors.com.
I'm sure I'll put it in the shownotes, that link and all that
stuff.
But these are kind of the twobest way.
If you have questions, you wannachat about anything.
Or is it down to, to each otherabout other people's, like tech
(49:06):
data problem.
Uh, and also you, I'm in themarketing apps community, so you
can DM me there
Michael Hartmann (49:11):
too.
Fantastic.
Well, this has been, it's beengreat, Mike.
Thank you.
Uh, yeah, thank you.
Hopefully we'll get Naomi backhere soon.
I know she's had, uh, Quite anadventure recently.
I'll let her Yep.
Share that when she'll be backwhen she's ready.
She'll be, she'll be back, um,to all of us, you listeners and
viewers, if, if we actually getthis out on video.
(49:31):
So, uh, thank you for allowingus to invade your personal
space.
I think I stole that from somemovie or something.
I don't remember.
Um, good job.
Anyway, until next time.
Bye everyone.
Bye everybody.
Mike Rizzo (49:45):
Bye.
Hello everyone.
(00:00):
Welcome to another episode ofOpsCast, brought to you by
Marketingops.com, powered by theMO Pros.
I'm your host, Michael Hartmann.
Joined today by one co-host,Mike Rizzo, who has been just,
he just flew in, so Mike.
Yes, I just zoomed
Mike Rizzo (00:14):
right in.
Thank
you.
Michael Hartmann (00:15):
I'm back.
I was zoomed right in.
Mike Rizzo (00:17):
I just zoomed right in cause we're doing this on
Zoom today.
And uh, yeah, I was at theMOps-Apalooza venue site visit,
getting ready for a big event,so that was
Michael Hartmann (00:27):
exciting.
All right.
MOps-Apalooza.
September, October.
I think I should know this.
I know I should know this.
There's a reason why we haven'thad episodes.
Okay.
November.
Wow.
See November.
I can only keep a few more datesin my head with all the projects
I have going on, so, sorry.
Fair.
It's okay.
(00:47):
Shame on me.
All right.
Well let's, let's get this,let's get this party started
cause we've got our guest today.
Si Giuliani, uh, We are gonna betalking to him a little bit
about, um, gosh, a couple ofdifferent things, but I think it
ultimately, it kind of comesdown to the sort of pending, um,
or, or increasing need to, tohave first party cookies and how
(01:08):
that applies to marketingoperations.
So, Uh, so is currently head ofgrowth at Census, a reverse ETL
provider, which I'm, I don'teven know what that means, so he
might have to tell us what thatis.
But, uh, he is also an activemember of several communities,
primarily in the revenue space.
He's also an investor mentor andhas experience with several
startups, both as a founder anda leader.
So, Sylvain, thanks for joiningus and welcome.
(01:32):
Thank you for having me.
Yeah.
What, what the hell, what thehell is reverse etl, by the way?
I just like, just get, I mean,
Sylvain Giuliani (01:39):
yeah, yeah, yeah.
I mean, reverse CTL is the actof, uh, getting data out of your
cloud data warehouse that youusually, uh, powering like your
BI tools and getting the datainto your operational tools such
as, you know, Marketto,Salesforce, you name it.
Like we have hundred plusintegration nowadays.
Oh, okay.
And, uh, okay.
So, you know, next time someonesays like, Hey, I see this data
(02:01):
and Looker.
I wish I could do somethingabout this churn problem.
Who are the people who churn?
Well, you can use RiverUniversity to identify the
people who are churning.
Get them into market load so youcan send them emails
Michael Hartmann (02:12):
before they got.
Got it.
Okay.
All right.
That makes a little sense.
Yeah.
All right.
That was not, that was notintended to be a pitch.
I just like, I saw that term andI was like, I don't know what
that means.
I kind of know what it means,but, all right.
Um,
Mike Rizzo (02:26):
we need more on the topic.
In fact, we were talking to Siand Si and their team about, um,
helping us learn more about thetopic.
So I think there's probably moreof that coming soon.
Okay.
Because there's not enough of usthat know what reverse ETL is,
that's for sure.
Michael Hartmann (02:40):
So we're excited about it.
You know, this, this whole ideaof like, Well, the, well, we're
now recording.
It's, it's Cinco de Mayo, 2023as we're recording.
And, um, there's been the, the,the latest MarTech landscape
just came out, what, earlierthis week, maybe late last week.
Yeah.
And what's the number?
(03:01):
14,000?
Is that what I saw?
I haven't had a chance to lookat it.
It's over 10,000.
I know that, but yeah.
Yeah, it's 11,000 something.
Now this whole need to sort ofstitch together data, I think
is, is a big one.
So yeah, definitely a topic weprobably need to, we need to dig
deeper into at some point.
So, Just, yeah.
You know, Sylvain, we'll haveto, uh, we'll have to bring you
(03:22):
back on for that.
Maybe.
I love it.
But one of those things isprobably, so first party
cookies, right?
So, um, you and I had aconversation quite a while ago,
but um, and things maybe havechanged a little bit, but, um,
kinda what's like, I love yourtake on kinda where things are
with first party versus thirdparty.
It's been a while since, I thinkwe actually had a guest on, like
(03:44):
over a year ago talking about.
This a little bit, but more,more so from the sort of
deprecation of third party dataor intent data.
But what are you seeing?
What are the trends, impacts?
Is there any equality stuff likejust a general overview?
Sylvain Giuliani (03:59):
Yeah, no, I mean, you know, the cookie
apocalypse happened.
I, I feel like in, in our space,I think, uh, you know, if you
think back a couple of yearsback, like as a marketer, you
were using one of those 10,000tools out there and to get data
in, you know, your hotspot,Marketo.
It was very simple.
You just slap some JavaScriptcode in your app, like on your
website or some code SDK in yourmobile app, right?
(04:21):
And it will do all the trackingfor you.
Facebook will do all theattribution for you, will tell
you like the conversion rate ofall of your campaigns, things
like that.
So didn't require you any kindof like data, uh, skillsets or
data transformation, skillset.
Then create happen and all thatdata basically.
Went away or you are very lowconfidence about it because it's
(04:41):
getting blocked and a lot andyou know, it's just not possible
to track it basically.
Right.
And so what we are seeing, youknow, as census as someone who
like took to a lot of like datateam and marketing team who are
using data is like a huge if tofirst party data, right?
And so what is first party datareally?
I mean, it's kind of like thesame thing, but you are the one
generating it and not a thirdparty vendor.
(05:01):
So it's like, you know, if yougo on census website today, we
are gonna track your click,right?
But it's like, Our own libraryis tracking your click.
So nobody's blocking it at,doesn't block it.
Right?
So, or it's event that might behappening in your, on the server
side.
Right.
So you know, when you create adocument in Figma, like that
event is generated on the serverside, so nothing can block it.
Right.
So we can know, oh, si create adocument in Figma and you shared
(05:24):
it with Michael.
Right?
So that's again, like that's anevent.
That is first party data.
And then you want to again, As amarketer, you wanna consume that
data because you wanna buildlike audiences segmentation.
Maybe you wanna do sometransactional email or push
notification based on like, uh,a trigger that they achieved.
Like, hey, you said like 10, uh,10.
10 documents in Figma, somethinglike that, right?
And so you, that data goesthrough the warehouse, that's
(05:46):
first party data goes throughthe W and then you wanna sync it
again to a tool like Marketo.
That's how way Senten tool likesensors comes in handy, right?
There's other ways to do it.
You can just write customscript, get you type the data
team to write code for you.
But most, most importantly, it'slike how do we, what we see is
like a lot of marketing peoplewanting to have.
Self-serve, access, themythical, like, you know, I
(06:06):
don't wanna talk to your datateam to self serve the data.
Like, kinda like what you usedto have with like the third
party data.
Uh, that's what we've seen.
That's kinda like the shift thatis happening.
Uh, you know, we can talk alittle about, also about like
CDPs, you know, the rise anddemise of CDPs.
It's also powered by this kind,like trend I would say.
But, uh, I will stop that to seeif that's, Answer your question.
Michael Hartmann (06:24):
The rise and demise of CDPs.
That's, it was a quick one ifit's demise already, but, um, so
there's two things that aregoing on.
I think two things that strikeme about this.
One is this sort of move awayfrom third party data.
It reliance on that.
Feels, feels like it's like adown, like a down negative
thing, but there's a potentialupside to first party data, but
(06:47):
it also comes with extra.
You need different additionalskill sets maybe, and tools to
capture even more data thanyou're already capturing in, in.
I don't know about you, but mostmarketing ops teams I know are
already swimming in data but notable to do much with it.
So how do we over like, yeah.
Feels like two sort ofchallenges at the same time.
Sylvain Giuliani (07:09):
Yeah, and I think that's like the, I think
there's.
Like next to the cookie agra,uh, I would say of third party
data.
Acas, right?
Like there was a big trend, uh,in the back, which was like the
rise of the warehouse, like, youknow, so the past few years, you
know, storing data, transformingdata has never been that easy.
Like, you know, it's like you UKit's very cheap to deploy.
It's very cheap to implement.
(07:30):
And so on the other side of yourbusiness, an IT team, a data
team, are probably investedtremendous amount of time and
money into building a warehouseand ultimately a bi.
Solution.
Right.
You know, like there's aexecutive team or board meeting
that is being powered by adashboard somewhere, right?
And so all that data is beingclean, transformed because big
decisions are being made on thatdata, right?
(07:50):
And so that's the, that's thebig shift that happened in the
back.
And so now it's like asmarketing team, that's great
because we can just build on topof that work.
Like we don't have to reinventthe wheel.
We don't have to.
That's why I say like the demiseof CDP, right?
Because the CDP is like you haveto redo all that work from
scratch again, as a marketingteam, like nowadays.
Solution plug into the west.
I'm not even talking aboutsensors, but you see, like if
(08:12):
you're a price customer, houseplus customer, you will see
them.
They're talking about like, Hey,we can connect to the warehouse,
we can send data to the west.
Like, you know, they're signed,everybody's trying to make a
warehouse, the first classcitizen, right?
And so as a marketing team, wecan leverage the work that has
been done by IT and data team.
And then just consume data thatwe can trust.
Right?
And I think that's like also athing that is accelerating the
(08:32):
adoption of first party databecause you don't have to start
from scratch essentially.
Right?
So all of those event that wetalked about that is happening
in the background, like numberof document created, who created
those documents, like anythinglike that, all of that is
already exists.
Like it's powering BI report,you can just reuse it as a
marketing team, right?
And I think that's, uh, one ofthe biggest, uh, you know,
accelerator of that trend.
(08:53):
Yeah, I
think,
Mike Rizzo (08:54):
I mean, I think a lot of this layers into there's,
you know, the PLG motion for us,for the SaaS companies out there
where they're trying tounderstand, you know, you were
still, you were hinting at, youknow, oh, you shared, uh, 10,
uh, links from Figma, right?
So like that, this idea thatyou're collecting usage behavior
from your, from your first partysort of interactions.
(09:17):
Um, and I think it's important,like, you know, for, for the
audience out there to understandthat like, The death of the
cookie, so to speak, isn't likearound this concept of, you
know, your HubSpot and munchkincodes and all that stuff, like
that's not going away.
Like, not like you've been doingthat and it's not going
anywhere.
(09:37):
Those are first party cookies.
That is your data.
It is collected, it is not gonnago anywhere.
People can turn off the cookielike consent banner and yes,
that part will stop flowing justlike any other cookie.
Michael Hartmann (09:51):
But that's already happening today.
Mike Rizzo (09:52):
Well like that.
Like that's already happeningtoday.
So in large part, like I wouldlove, and I'm gonna interject
here and this is where Michaelloves me cause I just jump in on
things and despite the outlineI'm sure we were probably
getting here.
Uh, I would love.
For you to say, you know, to, tojust share your thoughts on, you
know, really like, is the B2Bmarketer at the end of the day,
(10:17):
small to medium size, even the,even the enterprise.
Like, are they really impactedby this idea of like the cookie
being retired and like goingaway?
Like it feels like it's more ofthe e-com side of things and
like the consumer, the B2Cconsumer than anything else.
Sylvain Giuliani (10:35):
I think for sure like the small companies
reverse resource or econs, likethings like that.
Like they're getting destroyedbecause again, it was so easy to
just inject the Facebook likeyou know, tracker, right?
Like that's the thing's like,yeah, you don't have to be an A
genius.
You copy paste this thing.
Or even worse, sometimes youhave a direct integration.
You don't even know what'shappening, but you's saying to
track, right?
Yeah.
They're getting impacted.
(10:56):
I think like that.
Because that's going away.
It makes it harder for even likelarger company to do quickly
deployment.
Like, you know.
Now if I, like, I mean, let'stake an example.
We, the three of us, we work ata normal size, like B2B company,
300 people, right?
And we're like, Hey, you knowwhat?
We are gonna start doingadvertising on TikTok, right?
Like, so.
(11:16):
Now we can't just put the, thetracker, right?
Like now it's like, okay, weneed to sync our audience
somehow.
Like, you know, so, because wewanna retarget, so how do we do
that?
Okay, we can do a CSV uploadonce in a while.
Okay, that sucks, but I can comelike do it, but it sucks, right?
And then it's like, but really Ineed to send like conversion
event to, to, to feed themachine, the algorithm
technology.
(11:36):
Hey, this is good audience.
Like keep, keep showing the adsto the right people to train
their algorithm, right?
And then to do that, you need tosend event again.
You can do the Monday morningCSV upload, right?
Like it's okay, but like youdon't have that ease ness of
like just, yeah, just plug thesystem together and you just
work.
And then you're like, oh, whereis my conversion even coming
from?
Right?
Like, okay, some of them arecoming from spo, like you said,
(11:59):
those trackers are not goingaway.
If you're like, Hey, a demo formrequest, like, you know, SPO.
Credit tracking them form, likefor the mission, like nobody's
gonna tell you to take thataway.
But again, if you wanna say,Hey, I'll give you some example
of our customers, right?
It's like, hey, they, they showawareness ads.
So B2B plg company, right?
They show awareness ads or likefeature that you are not using.
So they're like, Hey Mike, weknow that you are notion.
(12:22):
User, right?
Where you're our user audienceand, but you're not haven tried
like our new AI feature, right?
For example.
And so they're gonna retargetyou aggressively everywhere
until you stop, uh, you startusing the ai, right?
And so that's kinda like the,those dynamic audience, this is
where it's get more complicated.
And before it was stillrelatively easy to do with the
third party tracker.
(12:42):
Now it's just impossible withoutlike, Some infrastructure or
help from data and engineering,right?
So that's, that's, I thinkthat's how they get impacted.
It's not like it's impossible,but it's just like an extra hoop
that they're to jump for thatthey didn't
Michael Hartmann (12:54):
have to do it before.
And I, and I don't, I actually,since I've, I actually am the
one here who's works at, hasworked at larger companies
currently working at a largecompany.
Like I don't think there'sreally much differentiating
their, you, you were headingwhere I expected you to go.
So, which was when Mike askedthat question, I suspected the
issue was actually how do youpush that?
First party data to thoseadvertising platforms now to do
(13:18):
the targeting in different ways.
And it sounds like that'sreally, that's gonna be the new
challenge because we're, they'regonna be dependent on general
stuff that they will probablystill try to package and keep to
the, you know, but it's gonna berelatively, if I remember right,
right, it's gonna be chunks of afew thousand people as opposed
to, you know, hundreds or dozensor, or, you know, whatever is
(13:40):
the right appropriate amount foryour, your business.
Mm-hmm.
Sylvain Giuliani (13:43):
Yeah.
And, and to me there's like thiskind of like duality, right?
It's like as marketer, like, youknow, you go to any marketer in
the world and you say, Hey,we're gonna build like agencies.
Like, nobody's gonna be like, ohmy God, this is a novel concept,
right?
It's more like, how do we build.
Better audiences.
Like, you know, I already have ahundred, but like I need to go
even further.
I need to press more.
(14:03):
And so it's like, I want, I knowI wanna do it.
What's stopping me is likeaccess to that data.
Right.
And so you took some of some ofthe things away from me by
having like this kind of thirdparty coming away or like, Like
that, kinda like third partyside of the house.
And so this is also why you seea lot more first party data
being used, right?
Because like, oh, I want to, Iwanna do a, no, I wanna do a, a
(14:24):
targeting that is like, youknow, location based, usage
base, like pre, like a likecalled purchase base, like yada
yada.
And so it's like you, youstopped talking to people.
I mean, I, I was talking tolike, uh, a large gaming
company, right?
And.
For them.
It's like we have audiences inthe order of magnitude of like
hundred thousands of audiences,right?
And it's just like everything islike programmatically done and
(14:45):
it's all powered by first partydata, right?
Like, and that's the, and youknow, and I asked them, yeah, I
mean that sounds like a lot ofwork.
Is it even worth it?
Like to go that big?
Right.
Uh, like in terms of like, forthem it's worth it because they
have large user base and it'snot about.
Reaching to people, they canreach anyone, but it's like
reaching to people to rightplace with the right message.
Right.
And that's the only thing, like,that's the power of first party
(15:06):
data is like, they know thatseal plays on, uh, Xbox and like
School of Duty, right.
And Mike, you more like aPlayStation person or a PC
person.
And so, and you like in like,you know, AARD game or something
like that of them, and they'llshow you the right ad in the
right place at the right timeinstead of like spending the ad
dollars showing ads toeverybody.
It was like, I keep seeing it.
A for wall of white, I'm a callperson.
(15:27):
Right?
Like that.
So that's the
Michael Hartmann (15:31):
how, how, how like, so my understanding, cause
I haven't had the opportunity toreally work with them a whole
lot.
But these intent data platforms,right?
Are they assumed they're pretty.
Relying on third party cookies.
Is, is that true or is it like,are they gonna be like, affected
by this as well?
So, so in other words then our,the marketers who are relying on
(15:51):
intent data for, you know,whether, whether they're using
for lead scoring or whateverthey're using it for, to.
Prioritize targets or whateveryou think they're gonna be
affected.
I think it depends
Mike Rizzo (16:01):
on where your intent data is coming from.
Total.
Like it's a total like reverseIP lookup play.
And for people from working fromhome, that's a challenge.
Um, but people going back to theoffice, that's helping.
And then I think it's also like,like G2 for example, you know,
the review platform.
They've got a crap ton of intentdata that they can signal out
(16:21):
back to people, but that's theirfirst party intent data, right?
So like there's value in thatdata.
Sylvain Giuliani (16:27):
Um, yeah, there's not too much issue
there, like intern data.
I mean, we can go into the pageslike, is it even worth investing
in this?
But like, uh, maybe it's a bigfan of a day.
Mike Rizzo (16:38):
Forget the same subject, throw like ro, like,
you know.
Yeah.
Uh, on those folks, but maybe Icould get some intent people to
come on and tell me about why itworks.
Sylvain Giuliani (16:49):
Yeah, yeah.
I think it works well for a verythin slice of the market.
Yeah.
And it works extremely well forthem, but everybody else, it's
kind of like us.
A money pit that doesn't bringyou anything.
Uh, that, that's my, that's myhot take on it.
Uh, but I think from anidentification point of view,
there's no issue cause you justneed an identity key, right?
(17:09):
Like if you use G2 as anexample, Here's, like G is gonna
be like, Hey, this companycalled uh, accent.com is looking
to buy like a warehouse, right?
Like, and so the only thing thatyou need is like this identity,
like the account, website ordomain, sorry, campaign.
Like that's the thing.
And it'll be the same forpeople.
It'd be like, Hey, this silk isdoing something right.
(17:32):
And so as long as you can matchthose two identity, you know,
that's, that's okay.
That's okay.
And that doesn't really requirethird party data.
I mean, you could like, youknow, using like, Cookies.
Like CrossFit cookie, but like,that's going away.
But like I think nobody reallyrelies on that anymore.
I mean, as far as as
Mike Rizzo (17:49):
I am.
Yeah.
Oh, I think what's interestingtoo, and I know we've got some
other things we want to diveinto here, but around this, like
the innovation that's gonna takeplace on, on the platform side,
right?
Like not, not the marketingautomation, technology side and
the 11,000 vendors there, butmore on the ad tech side.
Right.
So like you see there's likeLinkedIn lead forms, right?
(18:09):
Like yeah, they're going to usethat as your conversion signal
now.
And they may even, like, there'sa very, I could see a business
case where platforms likeLinkedIn and Facebook and all
those other providers are gonnaoptimize for you to have, uh,
potentially lower cost perconversion to go use those ad
platforms that hook up to theirown lead conversion metrics.
(18:30):
Versus trying to send you offplatform to go track a
conversion and do the whole backand forth thing, they're gonna
be like, no, you know what,just.
Just get your lead form here,right?
Yeah.
And then they're probably gonnabe like, you know what, just
upload your list.
In fact, take your data, uploadyour list over here, and run
your email campaign out of ourplatform.
And it's like, so like I, I feellike there's gonna be this big
(18:52):
shift in MarTech around theoutbound marketing capabilities
on the platform side.
Sylvain Giuliani (18:58):
I mean, like you're a hundred percent right?
I mean, if you look at LinkedIn,Add performance, right?
Like, I mean, you run exactlythe same ads.
Sames one is a leading lead genform, right?
So form on LinkedIn, you neverleave the website.
The other one is the same thing,but you click, you have to go to
an earning page.
Like the leading legend, likeobliterates, like the conversion
rate of the, of site conversion.
(19:19):
Right.
And, and you know, like LinkedIntell you straight up, like you
should do it like the, the ui,like they make it on purpose
that the non formm converseless.
The buttons are smaller, theclick zones are smaller.
Right.
Like it's, it's made through foryou to direct you to.
No, just keep people on.
It's fine.
Yeah.
It's gonna be totally, you know,like of course.
And
Mike Rizzo (19:39):
then forward tracking is there, right?
Like Yeah.
It's already happening.
And, and all the conversiontracking is happening on their
platform, which means it's likethere's this like, uh, false
sense of security that like, oh,cool, like I've got all the
things that I need.
But to your point earlier, whenyou want to start to layer in
the deeper segments of like,well, I really need to target
somebody that hasn't used X, Y,Z feature.
(20:02):
And as in a particular regionmm-hmm.
There's no way to be able to dothat connection anymore, which
is why, you know, marketers needto be paying attention.
Right.
Michael Hartmann (20:10):
So, okay.
So how, how will, how will the,and this, this may be, I know,
so, so you and I were talkingbefore we started about like
Aiche P t and how that, I thinkthere may be a tie in here, but
like one of the things we allwant to get to with our
marketing teams and marketingops teams is to.
Better understand like what'sthe preferred channel that
(20:31):
people want to be communicatedthrough, right?
Uh, or that performs best interms of whatever it is we're
trying to optimize.
How, how, how is this?
I guess two parts of it.
How is the, the cookie situationgonna affect that, do you think?
And then second, is there, like,you're describing all this sort
of intelligence, but it's builtinto the, to the, uh, the
(20:51):
platforms like LinkedIn orFacebook or whatever, or the,
you know, us having data andthen trying to deploy it in the
right way back out to them.
Like how, like is there a playfor AI in this somehow?
I don't, like, I don't know whatthat looks
Sylvain Giuliani (21:02):
like.
I think there's, I think youdecomposed like the problems in
like free, in free bucket,right?
It's like one is like, How do Icollect as much data as
possible, which is first partydata.
Like, so even like, you know,identifying prefer channel, like
it's, you need to know whichchannel works well for each
individual, right?
So again, if you think in termof first party data, it's like,
well, I know SI is a human.
(21:23):
Okay.
He does his email address.
Is this Id.
On different platforms.
So I've, I've sold identityresolution across as many
channel as possible.
And now I correct.
Data was like, oh, he consumeemail, he opens email, he
replies to email, he click onemail, right?
So a lot of first party datastream coming from that, and
it's like, well, I know email'sthe cheapest channel I have, so
I'm gonna use that first partydata to actually exclude sale
(21:44):
from all of the profe, like overchannel.
So I'm not gonna show you ads onLinkedIn because, I can get to
him for free Via emails.
Right.
Or via app messaging.
Right.
Or or whatever platform youhave.
Right.
Uh, then if you think aboutlike, Uh, third party platform,
ad platform or even like socialplatform, right?
Like lot of committee, like lotof things happen in Slack
committee now.
(22:04):
Like, you know, I think likethere's something called
marketing ops mobs, PLAcommittee, something like that
out there, right?
Uh, and it's like how do youtrack the, the engagement where
it's like you have to do.
Deep link, like UTM trackers,things like that.
So again, like you can collect,oh, this link is unique to seal
because he has like his ownlike, like, uh, anonymized ID
and he clicked on TikTok, right?
For example, on the dm on TikTokor WhatsApp, message on TikTok,
(22:26):
right?
So it's like I can get that databack and be like, oh, looks like
seals a lot on WhatsApp orwhatever platform you took
about, right?
And so, so I think that's one,like it's collecting that data.
Second point is like, how do wemake sense of that data, right?
Like, so I think this is wherehistorically the best
Michael Hartmann (22:40):
companies that, that, that is the big
question in my mind.
Sylvain Giuliani (22:43):
Yeah, historically it was gotta go
hire a bunch of data scientistpeople, essentially.
Right?
Or like, you know, during thekoolaid of one of the many,
mark, mark tech, advi, uh, youknow, like, uh, marketing
message.
It's like, no, just give us yourdata.
We put in the black box, insightcomes out.
You don't know how it works, buttrust us.
That's how it works.
(23:03):
Like, you know, that'sbasically, but again, it costs
you money.
I think right now what'sinteresting is like, you know,
GPT is trying to be very good.
I mean, as a technology,something that is very good at
like pattern matching.
Uh, identify things like I'mrunning like myself right now.
Just experiment like using G P Tto do identity resolution, for
example.
And that works extremely well.
Like, you know, like it can,like, it can detect that these
(23:25):
two humans like are almost thesame as like, Hey, we should do
something with it.
We should match them with a highlevel of confidence.
Right?
And so I think again, likethere's, there's gonna be a way
to repackage this kind of like GP T technology to kind of like
make it like, As, as a featureof a product.
I'm not saying like you're gonnago and charge PT and be like,
Hey, what's the differencebetween these two mics?
Kind of thing.
But like, it's gonna be veryeasy to build this feature, uh,
(23:46):
to make marketing life easier,right?
And then the third, the thirdbucket, right, is like, cool, I
got my data, I got my insights.
What do I do with it, right?
Because, uh, like how do Ideploy, how do I activate my
data in marketing speak, right?
Uh, and I think that's the,that's the next thing.
And so, again, like these freebuckets, like you need to be
exceptional at these threethings to be best of written
(24:09):
marketers nowadays, like, youknow, you like if you can be
good at one.
And that gets you ahead of thepack.
But like the best in classmarketing team I've, I've met
are, you know, they are worldclass at these things.
Yeah.
All of the data.
Super accessible, highconfidence.
They generate tons of this sideof it and they know what to do
with the data and they deploy itin like super tactical
advertising campaign, emailnurture campaign, like, you
(24:30):
know, empowering your sales teamwith insight to be like, Hey,
this account is, you know, thisaccount is at risk, this account
is ready to up sell.
Things like that.
Like that's, that's what thebest marketing team team do,
team do nowadays.
Sorry.
Um, but that's my take.
But that's how AI could fit intoit.
Michael Hartmann (24:45):
Um, wow.
So I, I bet there's a lot ofpeople out there listening like
me, where, where we know we havegaps or challenges with our, our
data.
Either it's not complete, notnormalized, right.
On key things that we typicallywould use.
Could, could something like thathelp with data cleansing kind of
(25:06):
stuff as well?
Sylvain Giuliani (25:08):
Yeah.
Well I think that like, this iswhere it depends what type of
company you have.
Like, you know, I thinkhistorically humans are still
like, you know, like if you tookyour data engineer or data
analyst, like they're gonna tellyou 60% of their job is like
data transformation is a niceway to say it.
Right?
Cleaning, right?
Yeah.
Yep.
And, uh, I think that's notgonna go away anytime soon.
(25:28):
I think, you know, like AI,things like, that's gonna make
that life a lot easier.
Like, I mean, I mean, go to J GB T right now, past like 24
numbers with completelydifferent format and you say,
put all of that into likeintentional formats, you know,
ease, o standard, whatever theZO standard.
It just does it right?
Like, you know, same thing pastlike the company name of like
(25:49):
one with ink, one with Ltd, onewith.com and say, just put them
into normal company name, humanreadable.
It just does it right?
Like, so I think that's going tohelp you tremendously, like
cleaning your data.
Uh, I don't think it's, it isstill ready to like, transform
your data and make sense of it.
Michael Hartmann (26:06):
It's not gonna, it's not gonna do it in
an automated way.
You still have to prompt it and,Give it the clue.
No, you can,
Sylvain Giuliani (26:10):
you No, you can, you can prompt it.
Like, I mean, you can, you canjust be like, this is my role,
you know, company name.
And then, you know, the otherone is like, you know, use G B T
to clean it by making like humanridable and it just kind like,
just keep doing it over andover.
Getting your data set, likethat's like, like these tools
already does it.
It's, and more like, this isjust gonna be a new normal, like
(26:32):
for most of the tools.
Sure.
Because it's so cheap andapproachable to
Michael Hartmann (26:34):
build things in this Yeah.
And the other, like the otherthing that, so I bet.
The, the, the first thing thatcame to mind when you talked
about like matching twoidentities from different places
that look like they're a matchwas, um, one of the challenges I
think everybody runs into isyou've got a really good client
at your, at your company.
They move to a similar role ora, you know, to another company.
Like how do you follow them soyou can go pursue them as a new,
(26:57):
new cli a client at thatcompany?
Uh, that may actually bepossible, if I'm understanding
what you're describing.
Sylvain Giuliani (27:05):
Yeah, I mean it's, it's possible today,
right?
Like, I mean, there's a bunch oftools out there, gem, you know,
APU does it like, I mean, youknow, it is at the end of the
day, it's like you are gonna belate when I think there's two
things, like knowing that personjoin, right?
So it's like, You know, it goesinto product building.
But like, you know, LinkedIn isa good place you can spare, plug
in, you know, you can, uh, Iwait for an email to bounce
(27:26):
like, you know, I send an emailto Mike user, right?
And like, oh, bounce, I guessit's not there so I can at least
understand it's not at thiscompany anymore.
Right?
Uh, so I think that's gonna be,that's a different set of
product.
Like I think this is not whereAI gonna be able to help you,
like magically, cause it needs adata set.
I mean, if you wanna checksomething, it's called people
GBT io.
It's got announced yesterday andit's like, A chat G P T
(27:50):
interface that you can create alot of people and can be like,
Hey, show me all the marketing pof people.
That looks like Mike Riso.
And I also have like thousandsof followers on Twitter.
And it's like, or like changejob.
Recently I just give you overlist.
Right?
It's like, it's like, oh, humanreadable or like NLP to, uh,
query a database of people andprospects.
Mike Rizzo (28:12):
Wow.
Crazy.
Michael Hartmann (28:14):
That's amazing.
Like I'm, I am actually sort ofspeechless right now.
I, I, I can't, I can't decide ifI'm, I'm amazed or I'm, or if
I'm frightened.
So
Mike Rizzo (28:26):
I don't know.
I mean, having come from theworld of, like, I started my
career in ad tech, um, and likeonce I figured out how all this
stuff worked, and of coursewe're like, we're literally in
the midst of a conversationabout how that whole world is
changing right now.
But like, I don't know.
When you go from, from that towhere we're headed now, I'm
like, should we really be thatsurprised?
(28:47):
Like people
Michael Hartmann (28:48):
pretty, pretty,
Mike Rizzo (28:50):
you know, in, um, I don't know.
They're engineering new outcomesnow to try to solve for these
problems that, you know, they'relike, well, I want it to be the
way it used to be, so I'm gonnatry to find.
I'm trying to find my workaround
Michael Hartmann (29:05):
on all that stuff.
Yeah, well, no, I, I'm not, I'mnot surprised either cause I
started my career in marketing.
I went from doing financialsystems and warehouses to
building a consumer database.
And this was a long time ago, 30years ago to call it.
And even then, right there wasa, like, when I saw what was
(29:26):
available to almost thehousehold level, I was just
stunned.
And then it's only gotten.
More robust and more datapoints, more, more activity and
all that.
Mike Rizzo (29:39):
Yeah, I
Sylvain Giuliani (29:40):
mean, I mean, yeah, I take like Google,
Facebook, the more first partydata you feed them, the more
they learn and the more theybetter target.
It's like you, it's kind oflike, uh, uh, an ai, but the
buzz was not made around it.
And for good reason.
You know, they want to say like,oh, you know, we have this AI
that allows you to targeteverybody.
(30:00):
It's like, that sounds sketchy.
Uh, but now the cat, the cat,the cat is out of the bag.
It's like, So this G P T thingis kind of like what we were
doing with ass, like, kind of, Iguess you can say that, you
know,
Mike Rizzo (30:13):
what was doing
Michael Hartmann (30:15):
hy hyper targeting at the individual
level.
Right.
It's just, that's yeah.
It's, and it's and good.
Yeah.
And like really good at it.
Mike Rizzo (30:22):
Yeah, it's very good.
It's, yeah, it's really scary.
Good at it.
Um, I, I have a question for youstill on just sort of your
thoughts on like, Yeah, I, I'vebeen, I've been tussling a bit
with, um, you know, I think,okay, so the underlying message,
or maybe the direct message ifit was said, uh, very plainly
(30:44):
is, look, you, you need to knowwhat to do with the data.
Right?
We just said, we just got donesort of talking about it.
Um, the, my challenge though islike, is that, do you, do you
feel that there's a role today?
Like that actually tries to takethat responsibility on.
(31:07):
No, like you actually said, thebest of breed marketers figure
this out, but like, is that justa new marketer that we're
talking about, like a new futurekind of marketer?
Or is it the data team?
Like are they gonna become quasimarketers?
Like, I don't know who'sresponsible for this.
Sylvain Giuliani (31:23):
I think, I mean, that's a great question.
Like, you know, it's like who'sgoing to eat wood?
Like, you know, there's alwaysthat debate is like, is data
team just gonna become likemarketing team or is marketing
team gonna absorb data team?
Right?
Or like ops teams or whatever,right?
It's like there's always this, Ithink it's just a new skillset.
I think if you just today theproblem is that I see in most
organization is like, It's somarginalized, right?
(31:44):
It's like, Hey, I'm a dimensionperson.
My job is to build campaignmessaging and like the, you
know, promote offers, right?
Where it's an ebook, an event orwebinar, whatever.
And so it's like my, I, I focuson such a spin slice and I, and
then simply, you know, marketingapps is like, well, I'm just
operationalized what thedimension person wants.
Like, you know, they wanna do awebinar, they wanna put a
(32:04):
nurture before and after, so I'mjust gonna do the thing that
they told me to do, kind of likeas a system, system level,
right?
And.
And I think because of this kindof like ultra ization, like
nobody's kind of like thinkingabout the overarching, like what
is the, what is the customerjourney and what the data fits
into this whole thing, right?
Like I think that's like, uh, Ithink there's no role for that.
(32:25):
And I think this is kind oflike, so because of that, It's
the new skills that everybodyshould have, right?
It's like, uh, you know, I wastalking to, I would not name
this company, right?
But like they, by design thisultra data savvy and automation,
they do data and automation,everything, right?
But they by design, never hiremarketing people in marketing,
right?
It's like if you have donemarketing before, you are
(32:47):
corrupt, right?
And so therefore, oh yeah.
Therefore you can't work for usbecause you don't have this
kinda like data automationmindset that we need, right?
Like, and they're like, We can,we can teach you marketing and
messaging and things like that,but we can't teach you being
like data curious and like thiskind of computer science
engineering mindset.
Right.
(33:08):
And anyway, I'm not saying it'sbad or right or wrong, but like
Right.
It's a different approach.
Right.
It's like you need that skillset, right.
Like this, a new, new type ofpeople.
I agree.
Michael Hartmann (33:16):
Yeah.
I, I, I mean, I could argueabout whether or not they sh
they could also teach some ofthe.
Data analysis stuff too as well,but like that, but the point, I
guess
Mike Rizzo (33:25):
there's, yeah, but like there's a difference
between, I, I do think you canteach somebody like the, the
what and the how goes into towhat SI's talking about from
that team.
But like, um, there's, there's adifference between like systems
thinking, And like dataorchestration and, and data
architecture and then liketherefore architecture that like
(33:48):
goes into that and like objectorientation and how those things
all sort of work together.
Like there's a differencebetween that.
Like, oh, I need to make sure wehave access to this data in this
structured way.
Like I.
Because it can get concatenatedand then we're gonna be able to
turn it into a string and the js o response, we're gonna be
able to do da, da, da, da da.
There's a difference betweenthat level of thinking and
(34:09):
someone who is like, who'strying to activate on that data.
And then somebody who's like,okay, like if I showed you this
j s o response, do you know whatyou're looking at?
And they'd be like, I have noidea.
And it's like, could you figureout if we, if we told you that
we could do such and such withthe data, what would you do
next?
Um, there's like, yeah.
Cause there's a, there's a verybig difference between those
(34:32):
types of it's
Sylvain Giuliani (34:32):
systems thinkers.
Yeah.
Someone told me it's like, youknow, it's like knowing
impossible is, is not, is, youknow, it's nothing as you think
tagline.
Right?
It's like there's a lot, likethere's sort of marketing people
are there that just don't knowwhat they could do.
Right?
Like it's just, It's justbecause it's part of the show
them.
It's like, yeah, exactly.
It's like, that's cool, but Idon't know what I can do.
(34:54):
Again, I'm not saying it'severywhere, but like, it's the,
the school of thought, the waywe were, are trained is to just
focus on certain things and likethis kind of like, Way of data
that that is coming is like, ifyou're not forced to use it, you
can kind of coastline and justbe good enough.
Like, you know, and, and it's,it's nothing
Mike Rizzo (35:13):
bad about
Michael Hartmann (35:13):
ok, so, oh, so you're hit, you're hitting on
where I, where I was gonna takea little bit issue.
Like, I, I get what you're bothsaying.
I, I still think there, I'm notgonna say everybody, right, but
there are gonna be people andsomeone, this comes from my own
experience like, When I gotoutta college, I actually just
told the story recently.
Somebody else, uh, we worked atPricewaterhouse.
They sent you off to for threemonths to get trained in their
(35:34):
way of doing at that time,cobalt programming, right?
Mostly right.
But how you did nice consultingwork for them and they had
cohorts of people all the timegoing through this training
center in my cohort were, waswith this one lady who had just
graduated from Duke with areligion major and.
(35:54):
Teaching her how to code waseasy, right?
cause she knew how to learn.
So I, I think there's like a, Ithink there's two things you
need sort of baseline, right?
You need a certain level of, uh,aptitude, right?
And knowledge and ability tolearn and the desire to learn.
Yes.
And then you need, then you needattitude, right?
You need, you need to be like,you have to have the desire, you
(36:15):
have to be coachable.
You be willing to try stuff andfail and learn.
And I think that's, that wouldbe the missing piece to me.
Like if somebody was in thatmore traditional marketing and
wanted to move it to the kind ofrole you're talking about.
That, that's, that's why I waslike resisting that because I
think in a lot of cases weunderestimate how much people
could actually learn, even ifit's outside of their, the
domain they think of as theirexpertise.
Mike Rizzo (36:38):
No, no.
I, I think look, everybody canlearn something.
Right.
Um, I think to that, to thepoint of, uh, what you were
talking about with that teamthat says like, Hey, you've sort
of been corrupted.
Yeah.
I think that's a little bit, youknow, extreme potentially, but
um, I think there is adifference between.
Where like si you hinted at itjust a second ago and
(36:59):
interestingly, it's a comment,it's a conversation I was having
with, uh, Lon Mendoza, whoyou've spoken to I think before
Si potentially.
Um, he's the, the, the founderof T M W, the MarTech Weekly.
And so we were having ourmonthly catch up last night and
um, and one of the things wewere talking about was like,
(37:20):
The, we both recently have sortof been gone out into the fray
on like, Hey, we're just gonnatry to figure out how to make
ends meet.
Right?
Like, I was put into a situationnow where I'm like, all right, I
gotta do marketing ops.com, someconsulting work to keep the
lights on.
And, and he went out and decidedto fully commit to his role.
And, and so we're, we're reallyentrepreneurs at this point, and
so, mm-hmm.
(37:42):
Um, we were talking about thisidea of.
The employee life, right?
The, the employee balance oflike where you're working and
what you're doing.
And, um, you know, I like, Ipersonally feel so much better
(38:03):
about like being out and tryingto hustle a little bit for
myself.
That there's like, there,there's this very dramatic
difference between.
Today's payday.
Cool, right?
I'm, I'm getting paid today andI'm gonna go to dinner tonight
and I'm gonna take the kids to aballgame this weekend or
(38:24):
whatever.
And then there's this other sideof it, which is like, I.
Man, I don't know if I'm gonnaget a paycheck this week cause I
didn't.
I didn't do enough work forclients or whatever it is.
And so going back to this ideaof learning and are you going to
take the time to figure out theart of the possible and try to
push yourself forward, yourorganization forward and all of
(38:46):
those things, like, it sort ofcomes down to this like, Are you
lazy or not?
Like are you motivated or
Michael Hartmann (38:55):
not?
I, I don't think, I don't thinkit's really, I think in some
cases it's not laziness ormotivation, cause I know lots of
marketers who work really hard.
I think one of the things thatgets in the way of a lot of
marketers with doing differentstuff than they just like rinse
and repeat kind of stuff isbecause there are put up.
Um, sort of, uh, gates orhurdles, right, with getting
(39:18):
stuff to market, right?
So they, they're, there's a,there's a lot of effort that
goes into getting somethingright as opposed to good enough,
right?
So you said that, or I think itwas about what you said and
quickly seeing if it works,right?
So that, I think that's one ofthe things I've see, especially
at bigger companies where you gothrough, yeah, so many levels of
review and approval, and then ittakes, people are wondering
(39:40):
like, why can't we just updatethat webpage?
So why does it take so many?
So, A week, right?
It's not because you, it's nothard to update, but people are,
they're afraid of like, it'sgonna be off brand, or I'm like,
like I tend to lean towardslike, just go right and do
(40:01):
enough to make sure you're notgonna do something that's really
egregiously wrong.
And even if you do that, likelearn from it.
Don't do it again.
Yeah.
Yeah.
Sylvain Giuliani (40:09):
It, it, I'll, I'll kind of summarize what you
both said.
Cause everything, it's like,it's a combination of both ways.
It's like you need to write theright attitude, right?
But you also have to have theright context to get the right
attitude.
Like, you know, you can bechilling at a company and, you
know, do a good job and youlike, I, I'm chill.
I'm doing worklife balance andgoing on the weekend, uh, fun,
right?
And then to your point, Mike,you're like, I'm an entrepreneur
(40:30):
here.
Like, you know, I, I got, Igotta make it work.
Like, like that.
The context forced you to have adifferent attitude is either
you're going to.
You know, do it or becomecreative and figure it out, or,
you know, you'll just fail andthen you know, you'll go try
something else because you, youcouldn't like have the right
attitude using your word here,Michael.
Right.
I think that's just, that's justa cycle of life and I don't
(40:51):
think there's, like, everybodyshould be, always be at full
speed and things like that.
It's just, oh goodness.
It's like, well, not fullpeople, but like you can just,
yeah.
Using the examples, like you gothere, you learn a ton of
things, you learn how toprogram, that's great.
Like then, you know, you go trysomething new, you learn
something and then, you know,maybe you don't learn something
for like three years.
That's okay too, right?
Like and then, yeah, I think theproblem is like when people stop
(41:12):
learning and people stopadapting, like, you know, so
many.
You so many people, AmazonMarket across the board, like
you said, what was the last timeyou learned something to uplevel
all your skills and doesn't seemlike take a course or something.
It's like you learn at your jobworking with a new peer, new
colleagues, things like that.
And like, actually I've beendoing the same thing for five
years.
You know?
And I feel like that's like,that's the problem.
(41:33):
That's the real problem.
And these like, that's likewhere you get like swept away by
people who come in.
It's like, I've been using G PT, it's my co-pilot.
Not, not saying they replacingpeople.
I think G P T thinks a lot like,you know, uh, I copilot an
enhancer to people, right?
And it's like, you just, like,what, how is this person writing
so much ad copy and the A copyworks better?
(41:53):
Oh, they just brainstorm with GP T all day.
Right?
And, uh, they as are performingbetter than yous just because
they can include it quickerbecause they have this assistant
to help them iterate.
Like just simple things likethat, right?
Like get them ahead and thenit's like one fixed it to the
other.
It's like, oh, why is the, whysales doing better?
It's like he's using intent datafrom blah, blah, blah.
Right?
And you know, like that's the.
(42:13):
That's the mentality, uh,that's, that you don't want to
sink into is like the personwho's like, I haven't learned
something new in five years.
Yeah, yeah.
What was, I mean, the colleagueor,
Mike Rizzo (42:21):
or anything?
Definitely not.
I, I, I, and then the otherpoint that like, so where I
branched off of for a moment waslike, hey, if, if, if someone
understands, you know, the artof the possible when it comes to
sort of data orchestration and,and what you can do with it, um,
(42:41):
So sometimes it's like, youknow, to your point, Hartman,
like, uh, there might be a lotof like roadblocks in the way to
be able to get something done.
And so I, I could, I can arguethat that's actually just a
factor of laziness.
Like, I just have to, I have todeal with all this stuff and I
don't really want to, uh,depending on the person's
(43:04):
response to that situation.
But if you, if you were.
If you were presented with achallenge and someone said, Hey,
I really need to try to figureout how to do this thing, and
you now know how to take datafrom one place and move it to
another, and you're this newkind of data marketer, but then
you recognize openly like that'sa lot of work.
That's, that that actually fallsoutside of the norm of the way
(43:25):
we normally do things at thiscompany.
Yeah.
Um, do I really want to trythat?
Is that really, is the companygoing to like that?
Is that for me?
Is it for them?
So there's all this, like youbecome this entrepreneur.
It's like, I can solve thisproblem.
I promise you I can solve thisproblem, but like, no one here
has done this before and so I'mgonna forge out on this path.
(43:49):
That like feels very different.
And that is where I was like,that's what branched off of.
If you start learning this newskillset and you figure out how
to solve these really uniqueproblems in unique ways that
don't fit into the quote unquotebox of how things are normally
done, you're going to run upagainst hurdles and then you're
going, it's going to take veryspecific.
Individuals who want to try toforge that path
Michael Hartmann (44:11):
forward.
Well, it's, it's permission.
It, it, what you're describingis like swimming upstream
against the, the idea that you,we should all be following best
practices.
Yes.
Right?
Yeah.
Because fallacy, by, by, by bydefinition, that's doing what
everybody else is doing.
And that's like not a recipe forlong-term success probably.
Um, so we're, I didn't, this hasbeen a great conversation.
(44:32):
I want to kinda get back to the,the, not cookie per se, but I, I
think.
I keep coming back to, I've feltfor a long time that like just
within marketing ops, but let'sgo to broader, to marketing in
general.
If there's a gap in, um, skillsand experience of knowing how
to, um, really.
(44:53):
Take advantage of data that'sbeen collected and um, it feels
like that's still gonna be, notonly is that still a problem,
that's gonna be an ongoingproblem if we don't address it,
especially as we continue, like,as we're developing even more
and more data through firstparty cookies and things like
that.
What, like, do you see, likewhat, how would you suggest for
our listeners that they start toaddress it?
(45:15):
Is it skill up leveling?
Is it looking for specificpeople?
Like would it.
I, I, yeah, I have, I don't havethe answer.
I just see it as a problem andI, I don't know how to solve it.
Sylvain Giuliani (45:25):
I think it's like being curious about it.
Like if you know you have thispain, there's a lot way to solve
it.
I think, you know, I, my, theway I've learned data is like, I
was friend with someone at workthat was in the data team, and,
you know, I had like all becauseI had all my questions about
data and he was the person Iused to go to, and so, Bit by
bits I started to learn aboutthe concepts of data.
(45:45):
Right.
And then, you know, I was like,I just don't wanna talk to him
anymore.
Like, I don't want him to be abottleneck.
Right.
And so I started learning.
And so like, cause but I, I haveenough information about the
world data that I could startlearning by myself.
Right.
And you know, after that you,you take a course, you, you do
some training online, like tolearn SQL to start query in a
new practice.
And it's, it's not thatcomplicated, right?
(46:07):
Like I think it's, but it startswith like, Having, being
curiosity and having a need.
Like if, I think if today peopleare like, I'm just gonna go this
weekend and start learningsequel, and you have no purpose
for it, like it's gonna be, soit's, it's gonna be a troll to,
to learn something.
And you don't want learning tobe a chore, right?
Like, I think we, we have otherthings to, to worry about that.
Then that's, and so that's wherewe approach, like I will meet up
(46:29):
someone in your team that ismore data savvy than you work
with them, find interestingproject to work together.
Go learn how to.
Query data.
Don't become a data engineer orsomething, but understand how
data Yeah.
Is model, like, how to query it.
Like what a big core concept.
Right.
And then finally, I'll, I'll dothe G P T plug, but like, you
know, the rate of things aregoing, like you can start like,
(46:51):
I mean, I use it.
Myself today, I can write prettygood sequel, but like sometimes
I'm like, ah God, I just gottaknow how to do this.
And I good.
I know enough that I can ask gpslike, can you write me a
windowing function that quitethis data, this is my data set.
And then it just quite a data, Iknow enough to know that what is
giving me is correct, but likeI.
I rely on my good friend, g p t,to, to help me do the work, you
(47:12):
know?
Uh, and I think that's, youcan't just start there, but I
think in the future it will helpyou.
Uh, you might be able to startthere, but I think that's,
again, it's a good acceleratorassistant
Michael Hartmann (47:23):
that that's my, well, and, and, and I
suspect that there's a number ofour listeners, audience or
whatever that um, don't have acolleague who's a data expert,
so maybe also going intocommunities, right.
To find others and just.
Yeah, look for like, it'samazing how much people will
help you learn.
Mike Rizzo (47:41):
Yes.
Sylvain Giuliani (47:42):
I mean, I use communities every days, you
know?
Yeah.
Every time I have Salesforceproblem, you know, God knows we
have, uh, so many all the time.
24 7.
You know what, it's, it's builtlike it's built.
It's built that way on.
That's their business model.
That's where the community,that's where community come,
comes in, you know, it's like,Hey, how's the.
What's the best of the 20million way I could do it in
(48:04):
Salesforce?
Tell me what's gonna go wrong ifI do it.
This one like you, that's whatthe C comes in the shell
Michael Hartmann (48:08):
experience is Eva evaluating the tradeoffs
very quickly, like crowdsourcingit.
Sylvain Giuliani (48:12):
Yeah.
Yeah, exactly.
Mike Rizzo (48:15):
Do it all the time.
Michael Hartmann (48:17):
Yeah, it's, I I, I do too.
Um, well, I, it feels like wecould go on and on here, but I
think we're gonna have to, towrap up here.
So, so thanks for joining us.
This is, uh, so this is one ofthe ways that I learned still
and stay current, is just like,just having these conversations.
I, I think I learn somethingevery time.
But, uh, so thank you so much.
(48:37):
If folks are interested in kindof keeping up with what you're
doing, the communities, etcetera, et cetera, what's the
best way for them to do that?
Um,
Sylvain Giuliani (48:45):
I'm on LinkedIn under my name So easy.
Uh, you know, like every goodmarket, I mean they're replying
to dm, you know, our socialnetwork.
It's the hassle I'm at there.
My email was Sylvain getsensors.com.
I'm sure I'll put it in the shownotes, that link and all that
stuff.
But these are kind of the twobest way.
If you have questions, you wannachat about anything.
Or is it down to, to each otherabout other people's, like tech
(49:06):
data problem.
Uh, and also you, I'm in themarketing apps community, so you
can DM me there
Michael Hartmann (49:11):
too.
Fantastic.
Well, this has been, it's beengreat, Mike.
Thank you.
Uh, yeah, thank you.
Hopefully we'll get Naomi backhere soon.
I know she's had, uh, Quite anadventure recently.
I'll let her Yep.
Share that when she'll be backwhen she's ready.
She'll be, she'll be back, um,to all of us, you listeners and
viewers, if, if we actually getthis out on video.
(49:31):
So, uh, thank you for allowingus to invade your personal
space.
I think I stole that from somemovie or something.
I don't remember.
Um, good job.
Anyway, until next time.
Bye everyone.
Bye everybody.
Mike Rizzo (49:45):
Bye.
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Therapy Gecko
An unlicensed lizard psychologist travels the universe talking to strangers about absolutely nothing. TO CALL THE GECKO: follow me on https://www.twitch.tv/lyleforever to get a notification for when I am taking calls. I am usually live Mondays, Wednesdays, and Fridays but lately a lot of other times too. I am a gecko.