August 26, 2025 • 49 mins
Text us your thoughts on the episode or the show!
In this episode of Ops Cast by MarketingOps.com, powered by The MO Pros, hosts Michael Hartmann, Mike Rizzo, and Naomi Liu speak with Dean de la Peña, VP of Identity, Data Strategy, and SaaS at Resonate.
Dean discusses the role of predictive intelligence in marketing and explains how brands can utilize more comprehensive data signals to enhance audience targeting and personalization. He also outlines the importance of identity resolution and data structure in building effective campaigns.
Topics covered include
• How to apply predictive consumer intelligence to marketing workflows
• The value of identity resolution in campaign planning
• Practical approaches to scaling personalization based on real data
This episode is intended for marketing operations professionals looking to improve their use of data in audience engagement.
Episode Brought to You By MO Pros
The #1 Community for Marketing Operations Professionals
Join us at MOps-Apalooza: https://mopsapalooza.com/
Save 10% with code opscast10
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Michael Hartmann (00:01):
This is an episode of OpsCast brought to
you by MarketingOpscom, poweredby all those mo-pros out there.
I'm your host, Michael Hartman,flying solo today, but joining
me today is my guest, Dean de laPena from Resonate, a company
at the forefront of predictiveconsumer intelligence.
Dean is the VP of Identity,Data Strategy and SaaS and
brings a wealth of experience inhelping marketers go beyond
(00:21):
surface-level targeting tocreate meaningful, scalable,
personalized experiences.
So, Dean, welcome to the show.
Dean de la Peña (00:27):
Thank you, happy to be here.
Michael Hartmann (00:29):
Yeah, let's do this.
So we're going to start at thebeginning, really, because I
think the heart of what we'regoing to be talking about is
personalization and what thatreally means in the context of
AI, your world with AI.
But when we think about this,how do you define
personalization from a marketingcontext?
And maybe what's the differencethat you see that, between what
(00:51):
people aspire to do and what weactually see in practice these
days?
Dean de la Peña (00:56):
That's a great question.
I think of personalization as areally fulsome one-on-one
conversation, getting to knowthe person that you're trying to
message to, right?
So you know, I think about it alittle bit as like the holy
grail of marketing.
It's basically like what, if wewere able to have a one-on-one
conversation with everyone thatwe were trying to connect with,
(01:20):
everyone that we're trying toengage, right?
And if you think about having aone-on-one conversation with
every adult in the US and youtook an hour for that discussion
, you'd learn a ton and it wouldtake you about five times the
length of recorded human history, right, right, we don't have
that kind of time.
So, you know, I think, forreally good reason, that's a
(01:41):
daunting prospect and becauseit's daunting, people have a
tendency to fall back on whatthey know, right?
So you know, I think thedifference between what you know
, ultimately we're aspiring tofrom a personalization context,
and what we see in practice, Ithink people tend to fall back
on.
Hey, give me a sense for whatthe demographics are, who you
(02:03):
know, what are some of the ageand gender characteristics of
the person that I'm trying tochat with, some of the
deterministic data, and I'llcall it a day right and I think
what we, you know, that'ssomething that you can get your
hands on easily, it's somethingyou can get your head around,
but it's not really answeringthat question of who are you,
how can I engage with you?
Michael Hartmann (02:21):
And, at the end of the day, to me at least,
personalization is really aboutadapting what we communicate and
how we communicate to thespecific person we're trying to
talk to.
Yeah, with as much context aswe understand.
I was just kind of smiling forthose who are not watching that
Cause I was.
I was thinking, like mostpeople probably think of
personalization.
Victory is, uh, you know,inserting first name into your
(02:42):
email, your message, right?
No, but I think what's reallyinteresting that just struck me
in what you were described andyou said it from the beginning
is it's because I think a lot ofpeople me included, when I
think of personalization,probably think of a single
communication, right, and youtalked about it as in terms of a
conversation, as opposed tothat.
So, which is interesting,because that's like that
(03:05):
actually, I think to me likementally changes how I would
think about it.
Right, there's, if it's a onetime thing, like it feels like,
there's like almost feels evenmore daunting right to use your
word because you have to get itright in that one moment,
whereas if it's a, if you thinkof it as one of many chances to
try to understand that person,you could like maybe give
(03:28):
yourself a little bit of leewayto like don't make assumptions
about what you know and you canchange how you talk to somebody
you know and try to build moreunderstanding.
Dean de la Peña (03:40):
Build more understanding and build a
relationship right.
Yeah, it's like, if I thinkabout my relationship with
brands and I'm a tough cookieit's hard to get me to click
into a digital advertisement.
It happens, and where I findmany of the marketers that are
engaging me are most successfulis when they really are
connecting with me.
They know generally what I careabout and they're building a
(04:02):
message right, and I thinkhistorically we've seen that,
you know, not from a digitalperspective, and and and.
Over time you'll see linear tvcampaigns, right, and, and.
Those campaigns will evolveover time and and.
Once they have the messagefirmly established and they've
got the catchphrase in your headand you know people are excited
about that, they'll build onthat, like I think about the um,
(04:23):
uh, many years ago.
There's, you know, the budlight commercial about that.
They'll build on that, like Ithink about the um, uh, many
years ago.
There's, you know, the BudLight commercial, um, and I'm
not affiliated with Bud Light inany way, right, but just I'm
thinking about this as anexample.
Michael Hartmann (04:32):
So funny Cause I was thinking about a Bud
Light series that I just justpopped up popped up recently in
my one of my social feeds.
Dean de la Peña (04:38):
Yeah, I mean, it's like I remember the.
I remember here we go, you know, and if I think about that
message, the message of action,and you know, enjoying when
yourself and like having a lotof fun, and I remember that once
they really established thatyou know they were starting to
transition that into more oflike hey, we have a dog and the
dog's name is we go, and it'slike here we go, and the dog
would bring you a beer andeverybody's having a lot of fun,
(04:59):
and it's that kind ofconsistent message.
If you can do thatunderstanding who the person is
you're trying to talk to, youknow you're actually building
that much more one-to-onerelationship and that engagement
over time, which is how youdelight your consumers, how you,
you know, really change howthey perceive you and your brand
(05:21):
and engage with you over timeyou and your brand and engage
with you over time.
Michael Hartmann (05:26):
Well, and that's, I mean, to me, like this
taps into human nature, right,that's how I think we build
actual human connection anyway,right, it takes time and it
takes some consistent effort tobuild that, because what you're
trying to build is trust, right,and that's a tough thing to do.
So I'm curious.
My guess is anybody listeningto this, any marketer, any
(05:49):
marketing ops person, is goingto go like, yeah, that sounds
right or at least in the righttrack.
Why do you think we've all kindof collectively missed the mark
or come below where we think weshould be for trying to do
personalization?
Dean de la Peña (06:05):
Great question.
I think there's two parts to myanswer to it.
The first is more of ahistorical view and recognizing
that technology is a reallyimportant part of it and I think
, historically, marketers havebeen limited in their ability to
think about and answer theright questions that create the
(06:27):
underpinning for that reallystrong relationship.
Right, because, like, we'rebasically trying to ask the
question and answer are you, youare a specific consumer,
michael like, are you in themarket for my product?
Do we share similar values?
Can I engage you on thosevalues?
Do you want me I engage you onthose values?
Do you want me to engage you onthose values?
(06:48):
Like, is that something thatyou actually do?
Want this messaging and what'sthe right message for you
specifically?
Right, and that's been reallyhard for a long time.
You know we've traditionallyrelied again on data that I'll
talk a little bit later aboutasking the right question and
we're kind of asking the wrongquestion if we're just looking
at age and income and I thinkwe've fallen back on some of
(07:11):
those characteristics andhistorically you'd find the
regions geographically or you'dfind broad segments of consumers
that generally met the kind ofthe vibe right or the set of
characteristics that you weretrying to engage with.
Like you had a strategy.
You knew who your consumer kindof was.
So let me find a group where alot of those people probably
(07:32):
live, and then I'll go and I'llscattershot, I'll peanut butter
my marketing to that group.
I love it.
Peanut butter, my marketing,peanut peanut butter marketing.
Um, and you know, and the idea,and that the idea and that
works right.
You hope you capture as muchattention and interest as you
can, but that really falls afoulof the old John Wanamaker quote
(07:54):
.
I know 50% of my marketinggoals work.
Trouble is I don't know whichhalf right, right, yeah, and so
it's just an inefficientapproach.
You're not really engaging, andif you don't know the specific
individual, you know you'relucky.
If you are, you're lucky, notintentional, if you are creating
that kind of one-on-onerelationship, and I think you
(08:17):
know that's the technology sideof it.
Now, technology has come a longway.
Analytics has come a long way,right, and the predictive AI
people's ability to createreally intelligent machine
learning models and predictiveAI that's become a reality.
Now, right, no-transcript?
(09:08):
We're going to have a lot moresuccess.
Michael Hartmann (09:11):
So I mean, do you think that's what
differentiates how you approachthis Like?
If so, like could you like?
What do you mean by asking thewrong questions versus the right
questions?
Dean de la Peña (09:19):
Yeah, that's fantastic.
So there's a really good quotethat I think about a lot from a
guy named John Tukey.
He's a statistician since past,but he wrote a book, I think
the Future of Data Analyticsback in the 60s, and so this is
back in the 60s, right?
This is way before you knowmodern.
(09:49):
To the wrong question, whichcan always be made precise,
right?
So instead of saying, look, Iwant to, you know, find the
general age group group, like, Iknow that you know, the people
who are engaging on my productare typically 25 to 35 year old
(10:10):
female on the West coast.
Whatever, that is Right, that'snot really the right question.
The right question is are youspecifically going to engage on
my message and be excited aboutthe opportunity to purchase my
product, have a conversationwith me, to build a relationship
with my brand, right?
So you know, we tackle thathead on using Ray, the
(10:33):
predictive AI that we've builtover a decade and, you know,
more than $100 million at thispoint of investment in that AI,
and we use that AI along withconsented consumer online
behavior and that lets us reallyhave that one-on-one
conversation at scale, right?
So the question you asked, likewhat is the right question In
my mind, is really deeplyunderstanding people's intent,
(10:57):
their motivations, their values.
What makes you tick and whatdifferentiates us is our ability
to line that up with theirbehavior and use AI to scale
that so that we have a reallygood sense for, hey, based on
how you act and again, based onconsented data right, privacy,
protected, careful about thatBased on how you're acting, we
(11:21):
can make an assessment of whatyou care about and we've seen
really strong results inconnecting that.
Michael Hartmann (11:26):
Ultimately, Okay, so I'm a, so I think I'm
following the.
The pushback maybe not pushback, but I think what I'm hearing,
though, is a little bitdifferent than asking the right
question.
I guess it's what a questionare we trying to answer?
As opposed to asking the rightquestion to the consumer, right,
that's what?
But the other part is what Iheard you describing what we're
(11:49):
looking for.
You essentially said we look atbehaviors rather than what
people tell us.
Is that kind of part of it too.
Dean de la Peña (11:58):
A bit of both.
We look at behavior tounderstand, based on what
several and I'll talk a littlebit about, kind of how our
secret sauce works and how we dothis, but we ask people what
they care about.
We talk to the consumer and wealso understand how many of them
behave and can connect thatbehavior which we have for
(12:18):
everyone, to really understandthe intents and motivations of
the individuals that we haven'tspoken to directly, Right, and
so you know to to maybe one wayfor me to analogize and clarify
what I had highlighted about youknow, asking the right question
, understanding the consumer ina different way than I think we
have historically.
(12:38):
Yeah, you know, if I thinkabout um, let's say that I shift
.
I shift from more of amarketing perspective to a sales
perspective and I want to sellyou, Michael, Like I'm a
clothing retailer, a fashionretailer, and I want to talk to
you, Michael, specifically aboutyou know, whether I can get you
(13:00):
excited about my product.
I'm going to have aconversation with you, right and
, as we're having today, right,I get to ask you questions about
what you care about, understandgenerally, how you feel, take a
look at what you're wearingright, which is going to tell me
a lot about.
You know your fashion, what youcare about.
I might, as an example, see youwalk out of the gym and if you
(13:22):
walk out of the gym, even ifyou're wearing street clothes, I
bet you make a pretty solidguess that you were just wearing
athletic gear, that you careabout athletic clothing or
athleisure clothing.
Right, and that's reallypowerful.
The traditional approach wouldbe to look at you from afar down
the street I don't knowanything about you, I don't know
(13:43):
your name and to just get asense of what you look like,
right.
How old do I think you are?
You know, what's your gender,what zip?
Michael Hartmann (13:50):
code are you?
Dean de la Peña (13:51):
in?
What zip code are you in?
Like I mean, I see you down thestreet, right, so I know you're
in the general area, but I'dmuch rather talk to you, learn
about you see, how you react toquestions that I'm asking, and
that's just so much morepowerful in how we can engage
and connect with right.
That connection with theconsumer, then, more of that
(14:13):
distant view of some of thefacts that don't matter as much.
You know, when we think aboutwhat people care about and how
they purchase.
Michael Hartmann (14:20):
I mean, the reason I asked you the question
is because I think so.
Parts one, I think one Istarted my career in marketing
and in database marketing, soeven back then, right, the
volume of data that wasavailable sort of stunned me,
and it's only increased.
Um, but the the bigger thing tome is that I'm I get a little
(14:40):
skeptical about the when we askquestions of consumers to get
their input, because of twothings.
One, sometimes people just flatout lie, right, I mean, I think
that's the case.
We've probably all done it tosome degree, just to get past
the gate.
Or, probably more prevalent iswe answer things the way we
think that we should, or what webelieve we do it, but our
(15:04):
behavior actually does somethingdifferent.
The example I go to is if youask people about how much they
care about their online privacy,virtually everyone's going to
say, absolutely, I don't want mystuff shared, blah, blah, blah,
blah, blah.
But then nobody actually reads.
You know, end user licenseagreements for all these online
apps that are consuming all thisdata and generating all these
insights about you, right?
(15:24):
And so, like, that's why likethe trade?
Because they're they'reconsciously or not they're,
they're assessing the trade-offof like I'm willing to give up
that.
What I care about, right, it'snot that they don't care about
the privacy piece, but on theother hand, right, there's a lot
of benefits and conveniencethat comes with letting that,
letting go of that to somedegree, so that I can have one
(15:47):
click purchase or one clicksubscribe or whatever.
So that's why I was asking thequestion about are we asking
questions of people?
Are we looking at behavior orsome combination of both?
Dean de la Peña (15:59):
No, I think you raise great points Right and I
also don't think you're wrong.
You know what people do asopposed to what they say they do
or are going to do cansometimes be different.
But you know our observationand this is borne out in the
results that we get with ourclients and their ability to
really move the needle on bettermarketing, more efficient
marketing, more effectivemarketing, higher return on your
(16:21):
ad spend right Is that, by andlarge, you know people are
answering questions in goodfaith, like they're trying to do
their best.
Michael Hartmann (16:28):
I actually do think that's true too, yeah, and
I see that in when we've hadthis is more of a B2B context,
but you know gated forms foraccessing content, for example,
and you have optional questions,like.
I'm always stunned at how muchpeople still fill that out, and
they fill it out in a way thatis actually mostly truthful,
yeah mostly truthful, and Ithink you know the other piece
(16:48):
of it too is.
Dean de la Peña (16:49):
that's where I get back to why I like that
Tukey quote so much right,Better to ask the right question
and get the approximate answerthan the wrong question
precisely.
I think there's some noise inthat right and with really good
predictive AI, with goodmodeling and analytical
techniques, you can sort throughthat noise and get a really
(17:11):
useful value driving answer.
That might be a little bitimperfect, right, like we're not
necessarily going to get thatquestion right every single time
, but again, I would so muchrather know if I give you a
nudge and I connect with youwith my messaging, you're going
to buy my product and I knowthat correctly 80% of the time.
(17:31):
I'd way way rather know thatthan know with a hundred percent
certainty that you're male.
Michael Hartmann (17:39):
Right, yeah, no, so I guess that's a good
this is.
That's a good clarification.
I'm glad we kind of wrestledthrough that so you guys do
mostly B2C work right.
Dean de la Peña (17:51):
So we are typically working with marketing
agencies and brands, so alittle bit more B2B2C.
Michael Hartmann (17:57):
Okay, okay, okay, but ultimately you're
helping consumers.
Dean de la Peña (18:03):
We're helping marketers connect with the
consumer.
That's right.
Yeah, we're helping marketersconnect with the consumer,
that's right.
Michael Hartmann (18:06):
Yeah, yeah, okay.
So I just want to make thatclarification because I mean,
it's not that I don't think itapplies to B2B, because I do
think many B2B marketers shouldbe thinking more like B2C
marketers, particularly when itcomes to targeting, messaging,
the whole bit, but that's awhole separate topic of its own.
But I'm just curious.
So, um, okay, I bought intothis idea like asking better
(18:31):
questions, even if the answersmaybe are not as precise.
Um is better at better toinform your targeting and
messaging.
And what?
Are you hitting the rightpeople at the right time?
It's like what are some ofthose inputs and signals that
help you and your clientsidentify the people that you
(18:52):
know will are better fits forwhatever their product or
service they're selling?
Dean de la Peña (18:57):
Yeah, absolutely.
I mean.
So you know, the the mostimportant input again is for a
subset of individuals, right,like we're not, we're not, we're
not asking 250 million people Ijust talked about how that
would take us, you know, justuntenably long to do.
But we do ask these questionsof consumers, right, what?
(19:17):
What are, what are yourmotivations?
You know, are you in the marketto buy a variety of specific
products?
How do you feel about certainissues?
Right, more about who you are.
We get to that level of depthand then you know, we're also,
like, I think, traditionallyright, if you're doing that kind
of survey and you think, from amarket research perspective,
you take that and you build outpanel insights.
Michael Hartmann (19:38):
Yeah.
Dean de la Peña (19:40):
Doesn't get you that far right, like that's
part and parcel of the oldchallenges with personalization,
where, ok, now I've got mypanel insights, that helps me
guide my creative a little bit.
But what I really care about isfinding that person and
connecting with them and whothey are, and that's where that
falls short.
Michael Hartmann (19:56):
It kind of gets you to the ability to do a
persona, which is anamalgamation of something that
tries to get it.
Dean de la Peña (20:03):
Exactly, and so what we're able to do then is
with the AI and with you knowagain, the consented online
behavior, like we're not talking, you know we're talking like
big data at this point right,billions of signals that help us
understand, you know, thebehavior across the consumer
base in the US adult consumerbase, us and then using Ray to
(20:26):
connect those dots.
And again, like the way I kindof think about that from a human
perspective is again being ableto assess a lot about you know
a friend of yours, or if youmeet someone new but you have
the opportunity to sit downacross the dinner table from
them, you'll learn a lot aboutthem, right.
You'll learn about kind of whothey are in a much better way
than just some of the basicdeterministic facts.
(20:47):
And if you think about whatpredictive AI is doing, it's
making those connections betweenbehavior and underlying intent,
motivation, preferences atscale.
It sees so much more than we dowith our hundreds or thousands
of acquaintances and it just isable to develop.
(21:09):
Basically, you can think of itas a lot of practice, right, of
just being able to kind of linkthose things.
Michael Hartmann (21:15):
Yeah, it's interesting.
So you, you know we can't havea conversation feels like these
days without talking about AI.
I think when you and I talkedbefore, you talked about
something to me combines twothings that I haven't seen
together a whole lot, althoughI'm bullish on it.
I guess, as you called it,predictive AI, right.
So I'm familiar with the ideaof predictive analytics, which
(21:37):
is hard on its own, and then itsounds like predictive AI is
combining that with kind of AIto is combining that with kind
of AI to either assist in thator generate different kind of
insights based on some sort ofan analyst.
Talk to me about what thatmeans.
Like, what is predictive AI andhow are you using that?
Dean de la Peña (21:59):
I'm really glad you asked that question because
I think that is.
I think it's getting lost inthe zeitgeist right now.
So I think of predictive AI andgenerative AI as two sides, two
parts of the same whole right.
And generative AI is what'sbuzzy right now.
(22:20):
We think of chat, gpt, claudeGrok right, all of these
different, you know, primarilychat based interfaces where you
can ask a question and get backa very fulsome answer, not
always exactly correct, butcertainly it it.
It really helps speed uptraditional research, putting
ideas together, language Right,and you know that's that's
(22:44):
obviously come a tremendouslylong way.
But that generative AI part ofthe world is.
I see that as a tremendousworkflow enhancer.
It's an efficiency enhancer.
It helps to pick up tasks thatwe used to have to do manually.
(23:04):
I used to have to write an emailmanually.
I used to have to write anemail manually.
I used to have to, you know,put together my thoughts on X, y
and Z, or do proofreadingmanually, right, and you know
coding used to have to be doneLike you had to get that out of
your fingers on the keyboard,right?
It didn't.
You didn't have quad to help.
You know, do 80% of the heavylifting for you and that's
(23:29):
incredibly valuable.
But what it doesn't do is itdoesn't do what the deep
learning, neural net modelingand the you know like the
extreme gradient boostingnonlinear modeling that allows
you to take these massive datasets and make sense of them and
translate that into predictionsof who these individual
consumers are with a tremendousdegree of accuracy.
That lets us connect with themwhere they are.
And so, you know, increasinglyfrom a software perspective and
(23:52):
not just from our ability toassess kind of who consumers are
, but for our ability to helpour clients get really efficient
, good use out of thatinformation, to connect that
directly to the DSP or socialmedia and the digital targeting
that you know our clients areable to do.
Generative AI has beenenormously helpful at sitting on
(24:14):
this this Ray asset that wehave in predictive modeling to
help our clients really, youknow, make that happen
efficiently, directly,automatically.
Michael Hartmann (24:24):
Yeah, I mean to me when I said I was I've
been bullish about that as apotential thing is because I've
directly, automatically yeah, Imean to me what I said I've been
bullish about that as apotential thing Because when I
was in database marketing at abig telecom company but we
actually had what today would becalled data scientists but they
were doing predictive analytics, modeling for different kinds
(24:44):
of things like churn, likelihoodto buy, etc.
Etc.
But the way I think about it,as I understood the process
right, there would have to besome sort of prediction.
They had, like almost likescientific method.
They have a hypothesis.
You're going to go through abunch of data, they're going to
do some modeling, see if theirhypothesis was true and then
turn that around into apredictive model.
See if their hypothesis wastrue and then turn that around
(25:06):
into a predictive model.
The piece that like, the piecethat I feel like this AI stuff
could do, is not totallyeliminate the need for the
hypothesis part, but likebecause it's like they could
comb through these large, largevolumes of data and look for
pattern, identify patterns thatwe may not otherwise, you know,
(25:28):
be able to do without.
You know, with, you know,manual efforts, limits on
technology and so on, and so tome, like that's, like I'm
excited about that as apossibility, I I still think, I
still stand by, like I stillthink you need humans in the
middle of it, at least for thenear future, because it may.
In fact, we had a guest on youknow one, one guest on recently
(25:55):
who talked about like if you put, if you were to put some of
these AI models in for, say,best customers, it's going to
tell you like people whoreturned stuff right, are are
great to target, and it's like,well, no, they return stuff.
It's not really not great.
So, anyway, so that's that'swhat I'm bullish about that, but
(26:21):
I don't know what else to whatelse to do there.
I think, I think I'm curiousabout that.
So you also mentioned to methat you, you, you have a large
scale consumer study.
You've kind of hinted at that.
Yeah, how, like?
How is that?
How do you tie that togetherwith the predictive AI?
And I think you called it Ray,that is, ray, the predictive AI
(26:43):
modeling engine.
Dean de la Peña (26:50):
Yes, yeah, so the the survey that we run, so
it's the US Consumer Study, it'sours, it's proprietary right
and that's what helps us reallyget deep on understanding of
individual consumers.
There are a lot of, and reallythe power of Ray is our ability
to connect the dots between thisdeep understanding of some
(27:11):
consumers, broad spectrumbehavior and demographics, and
an ability to say, all right,for everyone who we haven't
asked, or for everyone who wehaven't talked to, or for
everyone we don't have data on,sort of that dependent variable,
that outcome that we're tryingto understand, can we actually
predict that outcome foreveryone else?
Right?
So that's the key and you know,the consumer study is
(27:40):
essentially how we, one of theways, in many ways the deepest
way that we can create some ofthat truth, that ground truth,
data at scale right, where weactually say look like, let's
ask you, right, if we don't know.
You know what motivates you ifwe don't know that.
You know you are reallyuncertain about the economy and
how it's unfolding or you aretotally excited about the
direction that you know theeconomy is taking right now and
(28:03):
you feel very confident and surein your spending habits.
You're we don't know that youare about to have significant
family milestones and that'sgoing to change your how you
interact with the world writlarge, including how you
purchase and what you care about.
We ask right, and again, it'sthat ability to be able to that
(28:24):
gives us a lot of control overthe depth of question and the
depth of insight that we cancreate.
And again, of course, right andyou know, looming in all of
this conversation is privacy andthe ethics and regulation
around making sure that you knowwe're doing right by the
consumer, that we're connectingwith them when they want to be
(28:44):
connected with and not otherwise.
That survey, obviously peoplechoose to take that.
They choose to provide us thatbenefit of understanding and
that's an important part of theoverall process as well.
Michael Hartmann (28:57):
Gotcha Makes sense, all right.
So all this is making sense tome, no-transcript.
Dean de la Peña (29:26):
Yeah, absolutely so.
I'll give you one example.
Michael Hartmann (29:28):
You know, obviously sanitized to protect
the innocent here, but um, it'sBud Light, isn't it?
Dean de la Peña (29:30):
Bud Light?
Yeah, exactly.
No, it's not.
Um, I will, I will say I'll,I'll give you the negative.
This is not.
This is not Bud Light.
Um, but uh, you know we haveworked with um one client in
specific where you know they areusing the benefit of our
one-to-one understanding of theconsumer as they are engaging on
(29:52):
customer acquisition right.
They're trying to recruit aspecific type of consumer, like
they know, they know wherethey're strongest, they know
where they're weakest and theyknow who they need to be having
a conversation with, right.
Michael Hartmann (30:03):
And so I just how do?
They know that.
I'm curious Like is it becausethey haven't analyzed their
customer base and know what thatlooks like, both in terms of
like profile, demographic typestuff, as well as other
characteristics, maybebehavior-based things like that?
Dean de la Peña (30:21):
Yeah, I'd say you know in in three broad
spectrum ways, right?
The first is you know theyunderstand the trends in the
market in general.
So just from a very umbrellastrategic perspective, let's say
, as an example, you're in thefood and bev industry and this
client actually is in the foodand beverage industry and you
(30:44):
know that Ozempic, wegovie, someof these, you know these are
increasingly impacting people'srelationship with food, how they
spend, how much they spend.
You know how it changes theirhabits and their behaviors.
Right, understanding that andbuilding that in, there's the
(31:11):
second piece of it, which iskind of one layer deeper, which
is more of a study of how thedemographics around their
industry are changing and whatthey care about.
And the third piece is thelayer deeper than that, which
they're actually using ourunderstanding of the consumer to
develop a much more targeted,much clearer analysis of who is
still in the market and who'snot.
So you know, it's not just ourability to connect the consumer.
(31:33):
The most important part is ourability to connect the consumer
and the brand, but there's alsothe element of helping the brand
actually understand the nuancesof that consumer and who is
best to.
You know, have thatconversation with right, but
then in enabling them to executetheir digital marketing to the
right group of people in a muchmore focused, thoughtful way and
(31:56):
to develop messaging that makessense and is again just more
laser focused.
It's more engaging, right?
Um, they've seen 38% increasein their engagement metrics as a
result, right, relative totheir incumbent approach,
without the benefit of ourunderstanding of the consumer,
and, you know, when you thinkabout that, I mean that's a,
(32:17):
that's a big number.
But it also makes sense becauseit's the difference between
somebody you know, talking atyou, right, you don't care what
they have to say, it's notrelevant to your interests to
somebody who, like, really hitsyou between the eyes with oh my
God, I actually really, I reallydo care about that.
I was just thinking about, youknow, and so I'll give you an
(32:38):
example of my personal life.
Um, I am a pretty avid newly apretty avid cyclist.
I live in the city of Chicagoand I was walking down the
street back to my house one day,two Octobers ago, and there's a
bike store that was closingdown, and as I am walking past
(33:00):
the bike store, as I'm walkingup to the storefront on my way
home, I was thinking to myselfit's a beautiful day.
This walk is taking too long.
I would love a bike.
I really should think aboutbuying one.
And I turn my head to the leftand I see we're going out of
business.
All of our frames 45% off.
(33:21):
You know, get them while theylast.
Now, that's purelyserendipitous, right?
That is luck, right?
They obviously, you know theywere on hard times and they
needed to, you know, shift theirbusiness model and so they were
selling their bikes at adiscount.
But that was a great example ofright message, right time.
And I'm now the proud owner ofa weirdly bright blue Bianchi
(33:44):
bicycle.
And you know what?
It's an awesome bike.
It's also one of the last fewframes that they had in the shop
, right, because it was socompelling to me in that moment.
That was what I was looking for, I'd had the thought and it
just connected the dots.
And that's the power of reallyknowing in near real time what
(34:06):
your consumers are thinking.
Hey, you know if some of theconsumers out there are
concerned about impendingtariffs, right, and you know
they buy a lot of French wine,or they buy products from their
home country or a country thatis well known for Japanese
ceramics.
Whatever it might be.
(34:27):
I think a lot of peopleprobably purchased ahead of
tariffs.
Actually, I saw an article inthe journal recently about Swiss
watches.
Swiss watches are not cheap tostart with, and if you're adding
a tariff to that they're goingto get more expensive, and so
you actually saw in the economicdata an uptick in people
purchasing at that moment, andknowing who is and isn't aware
(34:49):
of those dynamics, who is andisn't in the market for that, it
starts to become pretty easy tosee how you can drive 35, 38,
40, 45% increases in row as anengagement by finding that
individual right when they'reready to make the purchase.
Michael Hartmann (35:03):
Yeah, makes sense.
Two questions are totallyunrelated, so let's go with one
first.
So one, okay.
So the idea here, it soundslike, is trying to replicate
that sort of serendipitous reallife experience to an online
digital experience.
Can you go a little deeper onhow that works?
Dean de la Peña (35:31):
Sure, I mean, it boils down to being able and
again given the consentassociated with it being able to
identify what makes a consumertick when they are in the
process of the you know digitalbid, right?
So it's like when you're seeingthe banner ads on you know your
(35:51):
, your edition of the wallstreet journal or wherever it is
that you're online or you'regetting advertisements on TikTok
or social media, making thatconnection at that moment and
realizing that that's what youcare about.
And so it's about making thisunderstanding of who the
consumer is, what they careabout in that moment, how they
(36:13):
work, and connecting the dots,kind of at that specific moment
where they are ready toessentially consume that or
related content.
Um, and doing that through theyou know the DSP, ssp and the
digital marketing process.
Michael Hartmann (36:29):
Gotcha.
I mean, are you and is ittailoring messaging to to some
degree, or is there sort of likewe've got a, maybe not three,
messages, but we've got 3000messages, right, and it's a
little more focused?
Dean de la Peña (36:41):
Yeah, I think it's a.
It's a question of tailoringmessages and it's tailoring
timing.
Michael Hartmann (36:46):
Yeah.
Dean de la Peña (36:46):
Okay, and so you know one of the things that
we had talked about I think whenwe talked last Michael was
around hey, segmentation versusone-on-one marketing, and I kind
of see it as tool for the jobright.
Segmentation is still reallyvaluable and I think building
your segments off of a reallydeep understanding of individual
consumers and building that upinto groupings that help you,
you know, kind of get a sensefor the gestalt um helps you
(37:09):
create the creative.
It helps you strategize, ithelps you maximize your reach.
It breaks out the, it breaksout the broader spectrum
strategic consumer group youcare about into chunks that are
manageable to the human brain.
Right To your point.
Humans are, and will continueto be part of this process.
Really important for us tounderstand how we each you know,
(37:31):
how other humans view the world, what's going to land to build
that creative process.
Like I personally believe, evenas a technologist and someone
who's been in AI for 15 years, Ireally believe that humans stay
a critical part of that and youknow our ability to help them
kind of make those connectionsfaster is really the kind of
(37:52):
core of the game.
Michael Hartmann (37:53):
Yeah, all right.
So my second question is reallyso this is consumer focused.
I'm trying to think of how thiswould potentially be analogous
in a B2B world, and so one ofthe things in fact, we just
talked about this in a recentpodcast something like the
anatomy of a deal, right.
So looking at, yeah, here's adeal we won last quarter.
(38:15):
Right, here are all the touchpoints across sales, marketing,
whatever right, includingdigital ones, and that can be
like, I think, your point.
That's why I was asking thequestion about how did that
client you talked aboutunderstand what a good customer
looked like?
Right, I think that's theanalogous there, and then kind
(38:37):
of going from there.
Typically, b2b buying doesn'thappen by an individual right.
Very often it's a group right.
So could could something likewhat you're describing also be
adapted to look at it from thatstandpoint, or is it?
Is it really dependent on thisum, like ongoing survey etc.
(38:58):
That you're doing where you'regathering additional input from
people, or do you do you thinkthat individual consumer stuff
could apply in the b2b world aswell?
yes, to the mic drop there, yeah, yeah exactly so, um, I I'm
going to.
Dean de la Peña (39:18):
I believe the answer to you know the sort of
overarching question of is thisvaluable in different contexts
B2C, b2b the answer to thatquestion is yes.
I'm going to reframe it just alittle bit by abstracting notion
of hey, is it like the consumersurvey that helps us create
(39:39):
some of this data at scale, oris it, you know, these other
pieces?
It's more about the principlesat play and the technologies of
being able to use and it takes alot, it's hard to develop these
modeling technologies, theseneural net, you know, nonlinear
models, the analyticaltechniques required to take data
(40:01):
set A right, these predictablecharacteristics of are we likely
to win the deal with XYZbusiness, or who is expected to
be a really good business thatwe want to be working with and
we want to, you know, prospectand try to attract from a
marketing perspective and thedata points that make that so.
(40:23):
And so you know, I guess, thequestions that I would ask from
a B2B perspective and you knowI've lived that world in the
past more from a salesperspective than a marketing
perspective but how do we talkto other businesses and how do
we, you know, kind of connectthose dots and show the value
that we can provide.
What makes a good client?
You know it's like it boilsdown to.
(40:44):
You know the extent to whichtheir strategy aligns with the
product that you're providing.
You know, do they?
Are you a nice to have versus aneed to have for them?
And you know what, in general,has been their perspective.
Like you know, are they acompany that is renowned for
finding the best technology anddeploying it?
Are they a company that is muchmore careful with their
(41:06):
investments and very tight knit?
Are they a company that likesto build things themselves?
Right?
These are all the key datapoints that I would use in my
own head as a human to decidewhether that's a good
opportunity.
And those are the kind of datapoints that you'd want to be
able to collect, broadlyspeaking, for some of these
businesses to help you, as amachine learning expert, build
(41:29):
the AI to make thosedeterminations.
And so, from that perspectiveand from our perspective, right,
where we are generally mosthelpful is in this really
complex, like we focused on thisreally complex problem of
people, and so you know, to yourpoint, right, the survey is a
part of that there's nointrinsic reason why you can't
(41:50):
start to collect that kind ofinformation for businesses and
apply the same general approachthat makes sense, yeah, and
that's kind of what I assume forbusinesses and apply the same
general approach.
Michael Hartmann (41:58):
Does that make sense?
Okay, yeah, and that's kind ofwhat I assumed.
I just was like just because Iknow a large portion of our
audience is primarily B2B and Ithink it's good for them to hear
about some of these things thatgo on in the B2C world, because
the scale is significantlyhigher, right, just so that it's
(42:19):
interesting to me, all right.
So maybe one last thing beforewe wrap up.
I mean, you've touched on thisa little bit.
You brought up the point aboutprivacy and compliance and that
kind of stuff.
You know clearly, right, thiskind of volume of data, there's
potential risks and concernsabout privacy.
Like, how do you handle that?
How do you balance that right,between, like, the goals to help
(42:41):
your clients achieve what theyneed to, while also being
respectful of the people whoseinformation you've you've got
responsibility for?
Dean de la Peña (42:50):
I love that question and in part because I'm
a bit of a privacy nut, right.
So you know I'm the guy thatrejects all cookies all the time
, uses the Apple, the AppleiPhone, you know, private relay,
proxy, right.
Like you know, I operate, youknow very much in that, and so I
(43:12):
come from the perspective thatI really want others to respect
my own choices.
From that perspective andthat's really what it starts
with is being respectful,ethical and privacy first, right
.
So that's step one.
You toe that line and in fact,actually we even go further than
that because, as we think aboutthe kind of data sizes that
(43:33):
you're talking about and beingcareful with and good stewards
of that data, we rely ontechnology that includes vector
embeddings as an example, whichbasically takes a data set that,
even if it's massive, actually,especially if it's massive
would otherwise be humanreadable, and turns that into
something that is condensed,actually easier for a machine to
(43:55):
use, easier for a model tounderstand and completely
unintelligible to the human.
It's gobbledygook, right.
So it really starts with, youknow, focusing on consented data
, right, using data thatconsumers allow for the use to
connect them, because, you know,to your point, right, consumers
actually want the value.
Where I do give up my privacyrights is Google Maps is a great
(44:22):
example.
I turn on the GPS and I letthem know everywhere that I go
and track that in history,because it's extraordinarily
useful and as we are helpingbrands connect with consumers
that want that connection, thatwant that message.
We want to make sure that we'redoing that.
But that starts with makingsure that the data that we use
is consented Our survey, ofcourse, is double consented and,
(44:45):
where there are sensitivitieson a state-by-state basis, right
where, hey, we're not allowedto use certain data for certain
purposes for any reason, makingsure that we always have the
appropriate constraints andguardrails there to protect
ourselves and our clients.
Michael Hartmann (44:59):
That's great.
No, I mean, it sounds to melike, really like it's a guiding
principle, right?
Dean de la Peña (45:03):
That would be the term for you as an
organization, and so that'sgreat you know, michael,
actually, on that note, beforewe close out on that, that topic
, right, I think the elephant inthe room is that makes it, that
makes the um makes it harder,right, because people want
privacy, um, it's harder toactually make that connection,
it's harder to sort through thedata and that's a.
(45:25):
That's a real differentiatorfor us is our ability to still
drive the kind of results thatyou know we're seeing with our
clients and to get the, theclarity not always right, right
like 80 confidence that we'regetting the right answer,
answering the right question,but that our ability to do that
well and accurately andprecisely, I think, really sets
(45:45):
us apart in a market where thatproblem, that problem is
increasingly difficult to manage.
Michael Hartmann (45:50):
Yeah, that makes sense.
All right, so we do need toprobably wrap up here, but just
before we move on, is thereanything that we like, like you
wanted?
Like anyone listening orwatching this walks away from
this like one key point that wemay not have covered, that you
want to make sure they heard, orwe covered everything?
Dean de la Peña (46:07):
I think we've covered about everything.
I mean, um, I think, just thethe advent of of, you know, data
science and how it's connectingto the human experience and
helping us, you know, kind ofhave that more fulsome
conversation, that fulsome humanconversation, right, whereas
really propping up how we aspeople talk to each other and
(46:29):
engage in the right ways at theright times.
That's what I'm about and Ithink that's really it's really
cool that we have theopportunity, as resonate, to
support that, that focus andthat I have the opportunity to
talk to you about it today.
Michael Hartmann (46:44):
So fantastic, all right.
Well, so let's move.
If people want to learn morecause I'm sure we didn't
actually cover everything, butlike, if they want to learn more
or connect with you or hearmore about what you're doing,
what's the best way for them todo that?
Dean de la Peña (46:58):
You connect with you or hear more about what
you're doing?
What's the best way for them todo that?
You can always find me onLinkedIn.
Shoot me an email.
Michael Hartmann (47:01):
Happy to chat.
All right, perfect, that soundsgreat.
Well, dean, again thank you.
This has been a funconversation.
It's kind of taken me back tomy roots and database marketing
and consumer stuff, so it'salways fun to go and think about
that, thanks.
Thanks also to our long timeand first time guests and
listeners, or listeners andsupporters.
(47:21):
We always appreciate that.
If you have suggestions fortopics or guests or want to be a
guest, you can always reach outto mike, naomi or me and we
would be happy to get started onthat until next time.
Bye, everybody, you.
This is an episode of OpsCast brought to
you by MarketingOpscom, poweredby all those mo-pros out there.
I'm your host, Michael Hartman,flying solo today, but joining
me today is my guest, Dean de laPena from Resonate, a company
at the forefront of predictiveconsumer intelligence.
Dean is the VP of Identity,Data Strategy and SaaS and
brings a wealth of experience inhelping marketers go beyond
(00:21):
surface-level targeting tocreate meaningful, scalable,
personalized experiences.
So, Dean, welcome to the show.
Dean de la Peña (00:27):
Thank you, happy to be here.
Michael Hartmann (00:29):
Yeah, let's do this.
So we're going to start at thebeginning, really, because I
think the heart of what we'regoing to be talking about is
personalization and what thatreally means in the context of
AI, your world with AI.
But when we think about this,how do you define
personalization from a marketingcontext?
And maybe what's the differencethat you see that, between what
(00:51):
people aspire to do and what weactually see in practice these
days?
Dean de la Peña (00:56):
That's a great question.
I think of personalization as areally fulsome one-on-one
conversation, getting to knowthe person that you're trying to
message to, right?
So you know, I think about it alittle bit as like the holy
grail of marketing.
It's basically like what, if wewere able to have a one-on-one
conversation with everyone thatwe were trying to connect with,
(01:20):
everyone that we're trying toengage, right?
And if you think about having aone-on-one conversation with
every adult in the US and youtook an hour for that discussion
, you'd learn a ton and it wouldtake you about five times the
length of recorded human history, right, right, we don't have
that kind of time.
So, you know, I think, forreally good reason, that's a
(01:41):
daunting prospect and becauseit's daunting, people have a
tendency to fall back on whatthey know, right?
So you know, I think thedifference between what you know
, ultimately we're aspiring tofrom a personalization context,
and what we see in practice, Ithink people tend to fall back
on.
Hey, give me a sense for whatthe demographics are, who you
(02:03):
know, what are some of the ageand gender characteristics of
the person that I'm trying tochat with, some of the
deterministic data, and I'llcall it a day right and I think
what we, you know, that'ssomething that you can get your
hands on easily, it's somethingyou can get your head around,
but it's not really answeringthat question of who are you,
how can I engage with you?
Michael Hartmann (02:21):
And, at the end of the day, to me at least,
personalization is really aboutadapting what we communicate and
how we communicate to thespecific person we're trying to
talk to.
Yeah, with as much context aswe understand.
I was just kind of smiling forthose who are not watching that
Cause I was.
I was thinking, like mostpeople probably think of
personalization.
Victory is, uh, you know,inserting first name into your
(02:42):
email, your message, right?
No, but I think what's reallyinteresting that just struck me
in what you were described andyou said it from the beginning
is it's because I think a lot ofpeople me included, when I
think of personalization,probably think of a single
communication, right, and youtalked about it as in terms of a
conversation, as opposed tothat.
So, which is interesting,because that's like that
(03:05):
actually, I think to me likementally changes how I would
think about it.
Right, there's, if it's a onetime thing, like it feels like,
there's like almost feels evenmore daunting right to use your
word because you have to get itright in that one moment,
whereas if it's a, if you thinkof it as one of many chances to
try to understand that person,you could like maybe give
(03:28):
yourself a little bit of leewayto like don't make assumptions
about what you know and you canchange how you talk to somebody
you know and try to build moreunderstanding.
Dean de la Peña (03:40):
Build more understanding and build a
relationship right.
Yeah, it's like, if I thinkabout my relationship with
brands and I'm a tough cookieit's hard to get me to click
into a digital advertisement.
It happens, and where I findmany of the marketers that are
engaging me are most successfulis when they really are
connecting with me.
They know generally what I careabout and they're building a
(04:02):
message right, and I thinkhistorically we've seen that,
you know, not from a digitalperspective, and and and.
Over time you'll see linear tvcampaigns, right, and, and.
Those campaigns will evolveover time and and.
Once they have the messagefirmly established and they've
got the catchphrase in your headand you know people are excited
about that, they'll build onthat, like I think about the um,
(04:23):
uh, many years ago.
There's, you know, the budlight commercial about that.
They'll build on that, like Ithink about the um, uh, many
years ago.
There's, you know, the BudLight commercial, um, and I'm
not affiliated with Bud Light inany way, right, but just I'm
thinking about this as anexample.
Michael Hartmann (04:32):
So funny Cause I was thinking about a Bud
Light series that I just justpopped up popped up recently in
my one of my social feeds.
Dean de la Peña (04:38):
Yeah, I mean, it's like I remember the.
I remember here we go, you know, and if I think about that
message, the message of action,and you know, enjoying when
yourself and like having a lotof fun, and I remember that once
they really established thatyou know they were starting to
transition that into more oflike hey, we have a dog and the
dog's name is we go, and it'slike here we go, and the dog
would bring you a beer andeverybody's having a lot of fun,
(04:59):
and it's that kind ofconsistent message.
If you can do thatunderstanding who the person is
you're trying to talk to, youknow you're actually building
that much more one-to-onerelationship and that engagement
over time, which is how youdelight your consumers, how you,
you know, really change howthey perceive you and your brand
(05:21):
and engage with you over timeyou and your brand and engage
with you over time.
Michael Hartmann (05:26):
Well, and that's, I mean, to me, like this
taps into human nature, right,that's how I think we build
actual human connection anyway,right, it takes time and it
takes some consistent effort tobuild that, because what you're
trying to build is trust, right,and that's a tough thing to do.
So I'm curious.
My guess is anybody listeningto this, any marketer, any
(05:49):
marketing ops person, is goingto go like, yeah, that sounds
right or at least in the righttrack.
Why do you think we've all kindof collectively missed the mark
or come below where we think weshould be for trying to do
personalization?
Dean de la Peña (06:05):
Great question.
I think there's two parts to myanswer to it.
The first is more of ahistorical view and recognizing
that technology is a reallyimportant part of it and I think
, historically, marketers havebeen limited in their ability to
think about and answer theright questions that create the
(06:27):
underpinning for that reallystrong relationship.
Right, because, like, we'rebasically trying to ask the
question and answer are you, youare a specific consumer,
michael like, are you in themarket for my product?
Do we share similar values?
Can I engage you on thosevalues?
Do you want me I engage you onthose values?
Do you want me to engage you onthose values?
(06:48):
Like, is that something thatyou actually do?
Want this messaging and what'sthe right message for you
specifically?
Right, and that's been reallyhard for a long time.
You know we've traditionallyrelied again on data that I'll
talk a little bit later aboutasking the right question and
we're kind of asking the wrongquestion if we're just looking
at age and income and I thinkwe've fallen back on some of
(07:11):
those characteristics andhistorically you'd find the
regions geographically or you'dfind broad segments of consumers
that generally met the kind ofthe vibe right or the set of
characteristics that you weretrying to engage with.
Like you had a strategy.
You knew who your consumer kindof was.
So let me find a group where alot of those people probably
(07:32):
live, and then I'll go and I'llscattershot, I'll peanut butter
my marketing to that group.
I love it.
Peanut butter, my marketing,peanut peanut butter marketing.
Um, and you know, and the idea,and that the idea and that
works right.
You hope you capture as muchattention and interest as you
can, but that really falls afoulof the old John Wanamaker quote
(07:54):
.
I know 50% of my marketinggoals work.
Trouble is I don't know whichhalf right, right, yeah, and so
it's just an inefficientapproach.
You're not really engaging, andif you don't know the specific
individual, you know you'relucky.
If you are, you're lucky, notintentional, if you are creating
that kind of one-on-onerelationship, and I think you
(08:17):
know that's the technology sideof it.
Now, technology has come a longway.
Analytics has come a long way,right, and the predictive AI
people's ability to createreally intelligent machine
learning models and predictiveAI that's become a reality.
Now, right, no-transcript?
(09:08):
We're going to have a lot moresuccess.
Michael Hartmann (09:11):
So I mean, do you think that's what
differentiates how you approachthis Like?
If so, like could you like?
What do you mean by asking thewrong questions versus the right
questions?
Dean de la Peña (09:19):
Yeah, that's fantastic.
So there's a really good quotethat I think about a lot from a
guy named John Tukey.
He's a statistician since past,but he wrote a book, I think
the Future of Data Analyticsback in the 60s, and so this is
back in the 60s, right?
This is way before you knowmodern.
(09:49):
To the wrong question, whichcan always be made precise,
right?
So instead of saying, look, Iwant to, you know, find the
general age group group, like, Iknow that you know, the people
who are engaging on my productare typically 25 to 35 year old
(10:10):
female on the West coast.
Whatever, that is Right, that'snot really the right question.
The right question is are youspecifically going to engage on
my message and be excited aboutthe opportunity to purchase my
product, have a conversationwith me, to build a relationship
with my brand, right?
So you know, we tackle thathead on using Ray, the
(10:33):
predictive AI that we've builtover a decade and, you know,
more than $100 million at thispoint of investment in that AI,
and we use that AI along withconsented consumer online
behavior and that lets us reallyhave that one-on-one
conversation at scale, right?
So the question you asked, likewhat is the right question In
my mind, is really deeplyunderstanding people's intent,
(10:57):
their motivations, their values.
What makes you tick and whatdifferentiates us is our ability
to line that up with theirbehavior and use AI to scale
that so that we have a reallygood sense for, hey, based on
how you act and again, based onconsented data right, privacy,
protected, careful about thatBased on how you're acting, we
(11:21):
can make an assessment of whatyou care about and we've seen
really strong results inconnecting that.
Michael Hartmann (11:26):
Ultimately, Okay, so I'm a, so I think I'm
following the.
The pushback maybe not pushback, but I think what I'm hearing,
though, is a little bitdifferent than asking the right
question.
I guess it's what a questionare we trying to answer?
As opposed to asking the rightquestion to the consumer, right,
that's what?
But the other part is what Iheard you describing what we're
(11:49):
looking for.
You essentially said we look atbehaviors rather than what
people tell us.
Is that kind of part of it too.
Dean de la Peña (11:58):
A bit of both.
We look at behavior tounderstand, based on what
several and I'll talk a littlebit about, kind of how our
secret sauce works and how we dothis, but we ask people what
they care about.
We talk to the consumer and wealso understand how many of them
behave and can connect thatbehavior which we have for
(12:18):
everyone, to really understandthe intents and motivations of
the individuals that we haven'tspoken to directly, Right, and
so you know to to maybe one wayfor me to analogize and clarify
what I had highlighted about youknow, asking the right question
, understanding the consumer ina different way than I think we
have historically.
(12:38):
Yeah, you know, if I thinkabout um, let's say that I shift
.
I shift from more of amarketing perspective to a sales
perspective and I want to sellyou, Michael, Like I'm a
clothing retailer, a fashionretailer, and I want to talk to
you, Michael, specifically aboutyou know, whether I can get you
(13:00):
excited about my product.
I'm going to have aconversation with you, right and
, as we're having today, right,I get to ask you questions about
what you care about, understandgenerally, how you feel, take a
look at what you're wearingright, which is going to tell me
a lot about.
You know your fashion, what youcare about.
I might, as an example, see youwalk out of the gym and if you
(13:22):
walk out of the gym, even ifyou're wearing street clothes, I
bet you make a pretty solidguess that you were just wearing
athletic gear, that you careabout athletic clothing or
athleisure clothing.
Right, and that's reallypowerful.
The traditional approach wouldbe to look at you from afar down
the street I don't knowanything about you, I don't know
(13:43):
your name and to just get asense of what you look like,
right.
How old do I think you are?
You know, what's your gender,what zip?
Michael Hartmann (13:50):
code are you?
Dean de la Peña (13:51):
in?
What zip code are you in?
Like I mean, I see you down thestreet, right, so I know you're
in the general area, but I'dmuch rather talk to you, learn
about you see, how you react toquestions that I'm asking, and
that's just so much morepowerful in how we can engage
and connect with right.
That connection with theconsumer, then, more of that
(14:13):
distant view of some of thefacts that don't matter as much.
You know, when we think aboutwhat people care about and how
they purchase.
Michael Hartmann (14:20):
I mean, the reason I asked you the question
is because I think so.
Parts one, I think one Istarted my career in marketing
and in database marketing, soeven back then, right, the
volume of data that wasavailable sort of stunned me,
and it's only increased.
Um, but the the bigger thing tome is that I'm I get a little
(14:40):
skeptical about the when we askquestions of consumers to get
their input, because of twothings.
One, sometimes people just flatout lie, right, I mean, I think
that's the case.
We've probably all done it tosome degree, just to get past
the gate.
Or, probably more prevalent iswe answer things the way we
think that we should, or what webelieve we do it, but our
(15:04):
behavior actually does somethingdifferent.
The example I go to is if youask people about how much they
care about their online privacy,virtually everyone's going to
say, absolutely, I don't want mystuff shared, blah, blah, blah,
blah, blah.
But then nobody actually reads.
You know, end user licenseagreements for all these online
apps that are consuming all thisdata and generating all these
insights about you, right?
(15:24):
And so, like, that's why likethe trade?
Because they're they'reconsciously or not they're,
they're assessing the trade-offof like I'm willing to give up
that.
What I care about, right, it'snot that they don't care about
the privacy piece, but on theother hand, right, there's a lot
of benefits and conveniencethat comes with letting that,
letting go of that to somedegree, so that I can have one
(15:47):
click purchase or one clicksubscribe or whatever.
So that's why I was asking thequestion about are we asking
questions of people?
Are we looking at behavior orsome combination of both?
Dean de la Peña (15:59):
No, I think you raise great points Right and I
also don't think you're wrong.
You know what people do asopposed to what they say they do
or are going to do cansometimes be different.
But you know our observationand this is borne out in the
results that we get with ourclients and their ability to
really move the needle on bettermarketing, more efficient
marketing, more effectivemarketing, higher return on your
(16:21):
ad spend right Is that, by andlarge, you know people are
answering questions in goodfaith, like they're trying to do
their best.
Michael Hartmann (16:28):
I actually do think that's true too, yeah, and
I see that in when we've hadthis is more of a B2B context,
but you know gated forms foraccessing content, for example,
and you have optional questions,like.
I'm always stunned at how muchpeople still fill that out, and
they fill it out in a way thatis actually mostly truthful,
yeah mostly truthful, and Ithink you know the other piece
(16:48):
of it too is.
Dean de la Peña (16:49):
that's where I get back to why I like that
Tukey quote so much right,Better to ask the right question
and get the approximate answerthan the wrong question
precisely.
I think there's some noise inthat right and with really good
predictive AI, with goodmodeling and analytical
techniques, you can sort throughthat noise and get a really
(17:11):
useful value driving answer.
That might be a little bitimperfect, right, like we're not
necessarily going to get thatquestion right every single time
, but again, I would so muchrather know if I give you a
nudge and I connect with youwith my messaging, you're going
to buy my product and I knowthat correctly 80% of the time.
(17:31):
I'd way way rather know thatthan know with a hundred percent
certainty that you're male.
Michael Hartmann (17:39):
Right, yeah, no, so I guess that's a good
this is.
That's a good clarification.
I'm glad we kind of wrestledthrough that so you guys do
mostly B2C work right.
Dean de la Peña (17:51):
So we are typically working with marketing
agencies and brands, so alittle bit more B2B2C.
Michael Hartmann (17:57):
Okay, okay, okay, but ultimately you're
helping consumers.
Dean de la Peña (18:03):
We're helping marketers connect with the
consumer.
That's right.
Yeah, we're helping marketersconnect with the consumer,
that's right.
Michael Hartmann (18:06):
Yeah, yeah, okay.
So I just want to make thatclarification because I mean,
it's not that I don't think itapplies to B2B, because I do
think many B2B marketers shouldbe thinking more like B2C
marketers, particularly when itcomes to targeting, messaging,
the whole bit, but that's awhole separate topic of its own.
But I'm just curious.
So, um, okay, I bought intothis idea like asking better
(18:31):
questions, even if the answersmaybe are not as precise.
Um is better at better toinform your targeting and
messaging.
And what?
Are you hitting the rightpeople at the right time?
It's like what are some ofthose inputs and signals that
help you and your clientsidentify the people that you
(18:52):
know will are better fits forwhatever their product or
service they're selling?
Dean de la Peña (18:57):
Yeah, absolutely.
I mean.
So you know, the the mostimportant input again is for a
subset of individuals, right,like we're not, we're not, we're
not asking 250 million people Ijust talked about how that
would take us, you know, justuntenably long to do.
But we do ask these questionsof consumers, right, what?
(19:17):
What are, what are yourmotivations?
You know, are you in the marketto buy a variety of specific
products?
How do you feel about certainissues?
Right, more about who you are.
We get to that level of depthand then you know, we're also,
like, I think, traditionallyright, if you're doing that kind
of survey and you think, from amarket research perspective,
you take that and you build outpanel insights.
Michael Hartmann (19:38):
Yeah.
Dean de la Peña (19:40):
Doesn't get you that far right, like that's
part and parcel of the oldchallenges with personalization,
where, ok, now I've got mypanel insights, that helps me
guide my creative a little bit.
But what I really care about isfinding that person and
connecting with them and whothey are, and that's where that
falls short.
Michael Hartmann (19:56):
It kind of gets you to the ability to do a
persona, which is anamalgamation of something that
tries to get it.
Dean de la Peña (20:03):
Exactly, and so what we're able to do then is
with the AI and with you knowagain, the consented online
behavior, like we're not talking, you know we're talking like
big data at this point right,billions of signals that help us
understand, you know, thebehavior across the consumer
base in the US adult consumerbase, us and then using Ray to
(20:26):
connect those dots.
And again, like the way I kindof think about that from a human
perspective is again being ableto assess a lot about you know
a friend of yours, or if youmeet someone new but you have
the opportunity to sit downacross the dinner table from
them, you'll learn a lot aboutthem, right.
You'll learn about kind of whothey are in a much better way
than just some of the basicdeterministic facts.
(20:47):
And if you think about whatpredictive AI is doing, it's
making those connections betweenbehavior and underlying intent,
motivation, preferences atscale.
It sees so much more than we dowith our hundreds or thousands
of acquaintances and it just isable to develop.
(21:09):
Basically, you can think of itas a lot of practice, right, of
just being able to kind of linkthose things.
Michael Hartmann (21:15):
Yeah, it's interesting.
So you, you know we can't havea conversation feels like these
days without talking about AI.
I think when you and I talkedbefore, you talked about
something to me combines twothings that I haven't seen
together a whole lot, althoughI'm bullish on it.
I guess, as you called it,predictive AI, right.
So I'm familiar with the ideaof predictive analytics, which
(21:37):
is hard on its own, and then itsounds like predictive AI is
combining that with kind of AIto is combining that with kind
of AI to either assist in thator generate different kind of
insights based on some sort ofan analyst.
Talk to me about what thatmeans.
Like, what is predictive AI andhow are you using that?
Dean de la Peña (21:59):
I'm really glad you asked that question because
I think that is.
I think it's getting lost inthe zeitgeist right now.
So I think of predictive AI andgenerative AI as two sides, two
parts of the same whole right.
And generative AI is what'sbuzzy right now.
(22:20):
We think of chat, gpt, claudeGrok right, all of these
different, you know, primarilychat based interfaces where you
can ask a question and get backa very fulsome answer, not
always exactly correct, butcertainly it it.
It really helps speed uptraditional research, putting
ideas together, language Right,and you know that's that's
(22:44):
obviously come a tremendouslylong way.
But that generative AI part ofthe world is.
I see that as a tremendousworkflow enhancer.
It's an efficiency enhancer.
It helps to pick up tasks thatwe used to have to do manually.
(23:04):
I used to have to write an emailmanually.
I used to have to write anemail manually.
I used to have to, you know,put together my thoughts on X, y
and Z, or do proofreadingmanually, right, and you know
coding used to have to be doneLike you had to get that out of
your fingers on the keyboard,right?
It didn't.
You didn't have quad to help.
You know, do 80% of the heavylifting for you and that's
(23:29):
incredibly valuable.
But what it doesn't do is itdoesn't do what the deep
learning, neural net modelingand the you know like the
extreme gradient boostingnonlinear modeling that allows
you to take these massive datasets and make sense of them and
translate that into predictionsof who these individual
consumers are with a tremendousdegree of accuracy.
That lets us connect with themwhere they are.
And so, you know, increasinglyfrom a software perspective and
(23:52):
not just from our ability toassess kind of who consumers are
, but for our ability to helpour clients get really efficient
, good use out of thatinformation, to connect that
directly to the DSP or socialmedia and the digital targeting
that you know our clients areable to do.
Generative AI has beenenormously helpful at sitting on
(24:14):
this this Ray asset that wehave in predictive modeling to
help our clients really, youknow, make that happen
efficiently, directly,automatically.
Michael Hartmann (24:24):
Yeah, I mean to me when I said I was I've
been bullish about that as apotential thing is because I've
directly, automatically yeah, Imean to me what I said I've been
bullish about that as apotential thing Because when I
was in database marketing at abig telecom company but we
actually had what today would becalled data scientists but they
were doing predictive analytics, modeling for different kinds
(24:44):
of things like churn, likelihoodto buy, etc.
Etc.
But the way I think about it,as I understood the process
right, there would have to besome sort of prediction.
They had, like almost likescientific method.
They have a hypothesis.
You're going to go through abunch of data, they're going to
do some modeling, see if theirhypothesis was true and then
turn that around into apredictive model.
See if their hypothesis wastrue and then turn that around
(25:06):
into a predictive model.
The piece that like, the piecethat I feel like this AI stuff
could do, is not totallyeliminate the need for the
hypothesis part, but likebecause it's like they could
comb through these large, largevolumes of data and look for
pattern, identify patterns thatwe may not otherwise, you know,
(25:28):
be able to do without.
You know, with, you know,manual efforts, limits on
technology and so on, and so tome, like that's, like I'm
excited about that as apossibility, I I still think, I
still stand by, like I stillthink you need humans in the
middle of it, at least for thenear future, because it may.
In fact, we had a guest on youknow one, one guest on recently
(25:55):
who talked about like if you put, if you were to put some of
these AI models in for, say,best customers, it's going to
tell you like people whoreturned stuff right, are are
great to target, and it's like,well, no, they return stuff.
It's not really not great.
So, anyway, so that's that'swhat I'm bullish about that, but
(26:21):
I don't know what else to whatelse to do there.
I think, I think I'm curiousabout that.
So you also mentioned to methat you, you, you have a large
scale consumer study.
You've kind of hinted at that.
Yeah, how, like?
How is that?
How do you tie that togetherwith the predictive AI?
And I think you called it Ray,that is, ray, the predictive AI
(26:43):
modeling engine.
Dean de la Peña (26:50):
Yes, yeah, so the the survey that we run, so
it's the US Consumer Study, it'sours, it's proprietary right
and that's what helps us reallyget deep on understanding of
individual consumers.
There are a lot of, and reallythe power of Ray is our ability
to connect the dots between thisdeep understanding of some
(27:11):
consumers, broad spectrumbehavior and demographics, and
an ability to say, all right,for everyone who we haven't
asked, or for everyone who wehaven't talked to, or for
everyone we don't have data on,sort of that dependent variable,
that outcome that we're tryingto understand, can we actually
predict that outcome foreveryone else?
Right?
So that's the key and you know,the consumer study is
(27:40):
essentially how we, one of theways, in many ways the deepest
way that we can create some ofthat truth, that ground truth,
data at scale right, where weactually say look like, let's
ask you, right, if we don't know.
You know what motivates you ifwe don't know that.
You know you are reallyuncertain about the economy and
how it's unfolding or you aretotally excited about the
direction that you know theeconomy is taking right now and
(28:03):
you feel very confident and surein your spending habits.
You're we don't know that youare about to have significant
family milestones and that'sgoing to change your how you
interact with the world writlarge, including how you
purchase and what you care about.
We ask right, and again, it'sthat ability to be able to that
(28:24):
gives us a lot of control overthe depth of question and the
depth of insight that we cancreate.
And again, of course, right andyou know, looming in all of
this conversation is privacy andthe ethics and regulation
around making sure that you knowwe're doing right by the
consumer, that we're connectingwith them when they want to be
(28:44):
connected with and not otherwise.
That survey, obviously peoplechoose to take that.
They choose to provide us thatbenefit of understanding and
that's an important part of theoverall process as well.
Michael Hartmann (28:57):
Gotcha Makes sense, all right.
So all this is making sense tome, no-transcript.
Dean de la Peña (29:26):
Yeah, absolutely so.
I'll give you one example.
Michael Hartmann (29:28):
You know, obviously sanitized to protect
the innocent here, but um, it'sBud Light, isn't it?
Dean de la Peña (29:30):
Bud Light?
Yeah, exactly.
No, it's not.
Um, I will, I will say I'll,I'll give you the negative.
This is not.
This is not Bud Light.
Um, but uh, you know we haveworked with um one client in
specific where you know they areusing the benefit of our
one-to-one understanding of theconsumer as they are engaging on
(29:52):
customer acquisition right.
They're trying to recruit aspecific type of consumer, like
they know, they know wherethey're strongest, they know
where they're weakest and theyknow who they need to be having
a conversation with, right.
Michael Hartmann (30:03):
And so I just how do?
They know that.
I'm curious Like is it becausethey haven't analyzed their
customer base and know what thatlooks like, both in terms of
like profile, demographic typestuff, as well as other
characteristics, maybebehavior-based things like that?
Dean de la Peña (30:21):
Yeah, I'd say you know in in three broad
spectrum ways, right?
The first is you know theyunderstand the trends in the
market in general.
So just from a very umbrellastrategic perspective, let's say
, as an example, you're in thefood and bev industry and this
client actually is in the foodand beverage industry and you
(30:44):
know that Ozempic, wegovie, someof these, you know these are
increasingly impacting people'srelationship with food, how they
spend, how much they spend.
You know how it changes theirhabits and their behaviors.
Right, understanding that andbuilding that in, there's the
(31:11):
second piece of it, which iskind of one layer deeper, which
is more of a study of how thedemographics around their
industry are changing and whatthey care about.
And the third piece is thelayer deeper than that, which
they're actually using ourunderstanding of the consumer to
develop a much more targeted,much clearer analysis of who is
still in the market and who'snot.
So you know, it's not just ourability to connect the consumer.
(31:33):
The most important part is ourability to connect the consumer
and the brand, but there's alsothe element of helping the brand
actually understand the nuancesof that consumer and who is
best to.
You know, have thatconversation with right, but
then in enabling them to executetheir digital marketing to the
right group of people in a muchmore focused, thoughtful way and
(31:56):
to develop messaging that makessense and is again just more
laser focused.
It's more engaging, right?
Um, they've seen 38% increasein their engagement metrics as a
result, right, relative totheir incumbent approach,
without the benefit of ourunderstanding of the consumer,
and, you know, when you thinkabout that, I mean that's a,
(32:17):
that's a big number.
But it also makes sense becauseit's the difference between
somebody you know, talking atyou, right, you don't care what
they have to say, it's notrelevant to your interests to
somebody who, like, really hitsyou between the eyes with oh my
God, I actually really, I reallydo care about that.
I was just thinking about, youknow, and so I'll give you an
(32:38):
example of my personal life.
Um, I am a pretty avid newly apretty avid cyclist.
I live in the city of Chicagoand I was walking down the
street back to my house one day,two Octobers ago, and there's a
bike store that was closingdown, and as I am walking past
(33:00):
the bike store, as I'm walkingup to the storefront on my way
home, I was thinking to myselfit's a beautiful day.
This walk is taking too long.
I would love a bike.
I really should think aboutbuying one.
And I turn my head to the leftand I see we're going out of
business.
All of our frames 45% off.
(33:21):
You know, get them while theylast.
Now, that's purelyserendipitous, right?
That is luck, right?
They obviously, you know theywere on hard times and they
needed to, you know, shift theirbusiness model and so they were
selling their bikes at adiscount.
But that was a great example ofright message, right time.
And I'm now the proud owner ofa weirdly bright blue Bianchi
(33:44):
bicycle.
And you know what?
It's an awesome bike.
It's also one of the last fewframes that they had in the shop
, right, because it was socompelling to me in that moment.
That was what I was looking for, I'd had the thought and it
just connected the dots.
And that's the power of reallyknowing in near real time what
(34:06):
your consumers are thinking.
Hey, you know if some of theconsumers out there are
concerned about impendingtariffs, right, and you know
they buy a lot of French wine,or they buy products from their
home country or a country thatis well known for Japanese
ceramics.
Whatever it might be.
(34:27):
I think a lot of peopleprobably purchased ahead of
tariffs.
Actually, I saw an article inthe journal recently about Swiss
watches.
Swiss watches are not cheap tostart with, and if you're adding
a tariff to that they're goingto get more expensive, and so
you actually saw in the economicdata an uptick in people
purchasing at that moment, andknowing who is and isn't aware
(34:49):
of those dynamics, who is andisn't in the market for that, it
starts to become pretty easy tosee how you can drive 35, 38,
40, 45% increases in row as anengagement by finding that
individual right when they'reready to make the purchase.
Michael Hartmann (35:03):
Yeah, makes sense.
Two questions are totallyunrelated, so let's go with one
first.
So one, okay.
So the idea here, it soundslike, is trying to replicate
that sort of serendipitous reallife experience to an online
digital experience.
Can you go a little deeper onhow that works?
Dean de la Peña (35:31):
Sure, I mean, it boils down to being able and
again given the consentassociated with it being able to
identify what makes a consumertick when they are in the
process of the you know digitalbid, right?
So it's like when you're seeingthe banner ads on you know your
(35:51):
, your edition of the wallstreet journal or wherever it is
that you're online or you'regetting advertisements on TikTok
or social media, making thatconnection at that moment and
realizing that that's what youcare about.
And so it's about making thisunderstanding of who the
consumer is, what they careabout in that moment, how they
(36:13):
work, and connecting the dots,kind of at that specific moment
where they are ready toessentially consume that or
related content.
Um, and doing that through theyou know the DSP, ssp and the
digital marketing process.
Michael Hartmann (36:29):
Gotcha.
I mean, are you and is ittailoring messaging to to some
degree, or is there sort of likewe've got a, maybe not three,
messages, but we've got 3000messages, right, and it's a
little more focused?
Dean de la Peña (36:41):
Yeah, I think it's a.
It's a question of tailoringmessages and it's tailoring
timing.
Michael Hartmann (36:46):
Yeah.
Dean de la Peña (36:46):
Okay, and so you know one of the things that
we had talked about I think whenwe talked last Michael was
around hey, segmentation versusone-on-one marketing, and I kind
of see it as tool for the jobright.
Segmentation is still reallyvaluable and I think building
your segments off of a reallydeep understanding of individual
consumers and building that upinto groupings that help you,
you know, kind of get a sensefor the gestalt um helps you
(37:09):
create the creative.
It helps you strategize, ithelps you maximize your reach.
It breaks out the, it breaksout the broader spectrum
strategic consumer group youcare about into chunks that are
manageable to the human brain.
Right To your point.
Humans are, and will continueto be part of this process.
Really important for us tounderstand how we each you know,
(37:31):
how other humans view the world, what's going to land to build
that creative process.
Like I personally believe, evenas a technologist and someone
who's been in AI for 15 years, Ireally believe that humans stay
a critical part of that and youknow our ability to help them
kind of make those connectionsfaster is really the kind of
(37:52):
core of the game.
Michael Hartmann (37:53):
Yeah, all right.
So my second question is reallyso this is consumer focused.
I'm trying to think of how thiswould potentially be analogous
in a B2B world, and so one ofthe things in fact, we just
talked about this in a recentpodcast something like the
anatomy of a deal, right.
So looking at, yeah, here's adeal we won last quarter.
(38:15):
Right, here are all the touchpoints across sales, marketing,
whatever right, includingdigital ones, and that can be
like, I think, your point.
That's why I was asking thequestion about how did that
client you talked aboutunderstand what a good customer
looked like?
Right, I think that's theanalogous there, and then kind
(38:37):
of going from there.
Typically, b2b buying doesn'thappen by an individual right.
Very often it's a group right.
So could could something likewhat you're describing also be
adapted to look at it from thatstandpoint, or is it?
Is it really dependent on thisum, like ongoing survey etc.
(38:58):
That you're doing where you'regathering additional input from
people, or do you do you thinkthat individual consumer stuff
could apply in the b2b world aswell?
yes, to the mic drop there, yeah, yeah exactly so, um, I I'm
going to.
Dean de la Peña (39:18):
I believe the answer to you know the sort of
overarching question of is thisvaluable in different contexts
B2C, b2b the answer to thatquestion is yes.
I'm going to reframe it just alittle bit by abstracting notion
of hey, is it like the consumersurvey that helps us create
(39:39):
some of this data at scale, oris it, you know, these other
pieces?
It's more about the principlesat play and the technologies of
being able to use and it takes alot, it's hard to develop these
modeling technologies, theseneural net, you know, nonlinear
models, the analyticaltechniques required to take data
(40:01):
set A right, these predictablecharacteristics of are we likely
to win the deal with XYZbusiness, or who is expected to
be a really good business thatwe want to be working with and
we want to, you know, prospectand try to attract from a
marketing perspective and thedata points that make that so.
(40:23):
And so you know, I guess, thequestions that I would ask from
a B2B perspective and you knowI've lived that world in the
past more from a salesperspective than a marketing
perspective but how do we talkto other businesses and how do
we, you know, kind of connectthose dots and show the value
that we can provide.
What makes a good client?
You know it's like it boilsdown to.
(40:44):
You know the extent to whichtheir strategy aligns with the
product that you're providing.
You know, do they?
Are you a nice to have versus aneed to have for them?
And you know what, in general,has been their perspective.
Like you know, are they acompany that is renowned for
finding the best technology anddeploying it?
Are they a company that is muchmore careful with their
(41:06):
investments and very tight knit?
Are they a company that likesto build things themselves?
Right?
These are all the key datapoints that I would use in my
own head as a human to decidewhether that's a good
opportunity.
And those are the kind of datapoints that you'd want to be
able to collect, broadlyspeaking, for some of these
businesses to help you, as amachine learning expert, build
(41:29):
the AI to make thosedeterminations.
And so, from that perspectiveand from our perspective, right,
where we are generally mosthelpful is in this really
complex, like we focused on thisreally complex problem of
people, and so you know, to yourpoint, right, the survey is a
part of that there's nointrinsic reason why you can't
(41:50):
start to collect that kind ofinformation for businesses and
apply the same general approachthat makes sense, yeah, and
that's kind of what I assume forbusinesses and apply the same
general approach.
Michael Hartmann (41:58):
Does that make sense?
Okay, yeah, and that's kind ofwhat I assumed.
I just was like just because Iknow a large portion of our
audience is primarily B2B and Ithink it's good for them to hear
about some of these things thatgo on in the B2C world, because
the scale is significantlyhigher, right, just so that it's
(42:19):
interesting to me, all right.
So maybe one last thing beforewe wrap up.
I mean, you've touched on thisa little bit.
You brought up the point aboutprivacy and compliance and that
kind of stuff.
You know clearly, right, thiskind of volume of data, there's
potential risks and concernsabout privacy.
Like, how do you handle that?
How do you balance that right,between, like, the goals to help
(42:41):
your clients achieve what theyneed to, while also being
respectful of the people whoseinformation you've you've got
responsibility for?
Dean de la Peña (42:50):
I love that question and in part because I'm
a bit of a privacy nut, right.
So you know I'm the guy thatrejects all cookies all the time
, uses the Apple, the AppleiPhone, you know, private relay,
proxy, right.
Like you know, I operate, youknow very much in that, and so I
(43:12):
come from the perspective thatI really want others to respect
my own choices.
From that perspective andthat's really what it starts
with is being respectful,ethical and privacy first, right
.
So that's step one.
You toe that line and in fact,actually we even go further than
that because, as we think aboutthe kind of data sizes that
(43:33):
you're talking about and beingcareful with and good stewards
of that data, we rely ontechnology that includes vector
embeddings as an example, whichbasically takes a data set that,
even if it's massive, actually,especially if it's massive
would otherwise be humanreadable, and turns that into
something that is condensed,actually easier for a machine to
(43:55):
use, easier for a model tounderstand and completely
unintelligible to the human.
It's gobbledygook, right.
So it really starts with, youknow, focusing on consented data
, right, using data thatconsumers allow for the use to
connect them, because, you know,to your point, right, consumers
actually want the value.
Where I do give up my privacyrights is Google Maps is a great
(44:22):
example.
I turn on the GPS and I letthem know everywhere that I go
and track that in history,because it's extraordinarily
useful and as we are helpingbrands connect with consumers
that want that connection, thatwant that message.
We want to make sure that we'redoing that.
But that starts with makingsure that the data that we use
is consented Our survey, ofcourse, is double consented and,
(44:45):
where there are sensitivitieson a state-by-state basis, right
where, hey, we're not allowedto use certain data for certain
purposes for any reason, makingsure that we always have the
appropriate constraints andguardrails there to protect
ourselves and our clients.
Michael Hartmann (44:59):
That's great.
No, I mean, it sounds to melike, really like it's a guiding
principle, right?
Dean de la Peña (45:03):
That would be the term for you as an
organization, and so that'sgreat you know, michael,
actually, on that note, beforewe close out on that, that topic
, right, I think the elephant inthe room is that makes it, that
makes the um makes it harder,right, because people want
privacy, um, it's harder toactually make that connection,
it's harder to sort through thedata and that's a.
(45:25):
That's a real differentiatorfor us is our ability to still
drive the kind of results thatyou know we're seeing with our
clients and to get the, theclarity not always right, right
like 80 confidence that we'regetting the right answer,
answering the right question,but that our ability to do that
well and accurately andprecisely, I think, really sets
(45:45):
us apart in a market where thatproblem, that problem is
increasingly difficult to manage.
Michael Hartmann (45:50):
Yeah, that makes sense.
All right, so we do need toprobably wrap up here, but just
before we move on, is thereanything that we like, like you
wanted?
Like anyone listening orwatching this walks away from
this like one key point that wemay not have covered, that you
want to make sure they heard, orwe covered everything?
Dean de la Peña (46:07):
I think we've covered about everything.
I mean, um, I think, just thethe advent of of, you know, data
science and how it's connectingto the human experience and
helping us, you know, kind ofhave that more fulsome
conversation, that fulsome humanconversation, right, whereas
really propping up how we aspeople talk to each other and
(46:29):
engage in the right ways at theright times.
That's what I'm about and Ithink that's really it's really
cool that we have theopportunity, as resonate, to
support that, that focus andthat I have the opportunity to
talk to you about it today.
Michael Hartmann (46:44):
So fantastic, all right.
Well, so let's move.
If people want to learn morecause I'm sure we didn't
actually cover everything, butlike, if they want to learn more
or connect with you or hearmore about what you're doing,
what's the best way for them todo that?
Dean de la Peña (46:58):
You connect with you or hear more about what
you're doing?
What's the best way for them todo that?
You can always find me onLinkedIn.
Shoot me an email.
Michael Hartmann (47:01):
Happy to chat.
All right, perfect, that soundsgreat.
Well, dean, again thank you.
This has been a funconversation.
It's kind of taken me back tomy roots and database marketing
and consumer stuff, so it'salways fun to go and think about
that, thanks.
Thanks also to our long timeand first time guests and
listeners, or listeners andsupporters.
(47:21):
We always appreciate that.
If you have suggestions fortopics or guests or want to be a
guest, you can always reach outto mike, naomi or me and we
would be happy to get started onthat until next time.
Bye, everybody, you.
Popular Podcasts
On Purpose with Jay Shetty
I’m Jay Shetty host of On Purpose the worlds #1 Mental Health podcast and I’m so grateful you found us. I started this podcast 5 years ago to invite you into conversations and workshops that are designed to help make you happier, healthier and more healed. I believe that when you (yes you) feel seen, heard and understood you’re able to deal with relationship struggles, work challenges and life’s ups and downs with more ease and grace. I interview experts, celebrities, thought leaders and athletes so that we can grow our mindset, build better habits and uncover a side of them we’ve never seen before. New episodes every Monday and Friday. Your support means the world to me and I don’t take it for granted — click the follow button and leave a review to help us spread the love with On Purpose. I can’t wait for you to listen to your first or 500th episode!
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
The Joe Rogan Experience
The official podcast of comedian Joe Rogan.