November 29, 2022 • 48 mins
Text us your thoughts on the episode or the show!
In this episode, we talk with George Xing (Shing), Co-founder and CEO of Supergrain, a customer engagement platform. Prior to founding Supergrain, George held multiple roles at Lyft, including Director, Head of Decision Science Products and was a data scientist at Indiegogo. Before that he was a Fixed Income Analyst with Morgan Stanley.
Tune in to hear:
- George's career journey and how it led to starting Supergrain.
- Why he describes Supergrain as a warehouse-native solution for marketing / marketing ops.
- Trends that George is seeing in the Martech / Revtech space and how the Supergrain solution aligns with these technology trends.
Episode Brought to You By MO Pros
The #1 Community for Marketing Operations Professionals
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Michael Hartmann (00:00):
Hello everyone.
Welcome to another episode ofOpsCast, brought to you by
marketing ops.com, powered bythe Mo Pros.
I'm your host, Michael Hartmann.
Joined today by one co-host.
Today I've got Mike Rizzo by myside.
So joining us today is GeorgeXing, co-founder and CEO of
Super Grain, a customerengagement platform.
So prior to and founding SuperGrain, George held multiple
(00:21):
roles at Lift, includingdirector, head of decision
science products, and was a datascientist at Indie Gogo.
Before that, he was a fixedincome analyst list with Morgan
Stanley.
Uh, which I find fascinating,the combination of things here.
George, thanks for joining ustoday.
George Xing (00:36):
Thanks, Michael.
Thanks guys.
Uh, super glad to be here andexcited to chat.
Marketing automation andmarketing ops.
Michael Hartmann (00:43):
Awesome.
Yeah, no, I know we talked alittle bit ago, what I was
actually looking back, it was,we're now late November
recording this.
It was late August I think.
So it's been a while since wechatted, so I'm sure there's
been even more movement, whatyou're doing.
But before we get into sort ofthe novel way that Super Green
is approaching, uh, marketingautomation, and I think it goes
beyond that, we'll get that in alittle bit.
(01:05):
Like, can you first maybe.
Share a little more about, youknow, your career and your
journey and how it led tostarting super grade as a
beginning point.
Let's start from there.
George Xing (01:16):
Yeah, as you mentioned, uh, thanks for that
intro by the way.
Uh, started my career actuallyin finance, which is like
totally different on a fixedincome trading floor.
working at a bank in Hong Kong,so literally the other side of
the world.
Um, and I think the only commonthread was that there was
something numeric and somethinganalytical about the work I was
(01:37):
doing.
Other than that, I hated workingin finance.
So after about a year I movedover, uh, you know, came back to
the States, wanted to be in theBay Area.
I'm originally from the Eastcoast.
Super cold winters.
Didn't wanna be in New Jersey,so came out.
Joined a startup calledIndigogo.
And the only role that I wasqualified to do was something,
(01:58):
uh, that at the time was, uh,they call data analytics, which
I wasn't familiar with, butturned out to be something that,
uh, was useful for the business.
And, uh, from there, the rest ofmy career has been in data and
analytics.
So spent a couple years atIndiegogo.
Um, and then, uh, while I was inSan Francisco, I started using.
(02:21):
This ride sharing service calledLyft, um, ended up joining them
as the, the first data hirethere.
And then, you know, uh, it justkind of as the company grew, uh,
I started working on datascience, business intelligence,
data engineering, and built outa number of teams there.
Um, and it was really at Liftwhere I saw a lot of the
(02:42):
interactions between data and alot of the stakeholders, um,
marketing teams, growth.
and a lot of the challenges thatwere involved.
So a lot of my job was justsitting at the intersection of
data decision making, figuringout how to get the right data,
sets the right data points tothe right people in order to
make decisions.
And whether that was, um,marketing decision makers,
(03:06):
whether that was running emailcampaigns were, whether that was
running automation.
So as you can imagine, at Lift,we had a number of very, very
complex, um, automations foreveryth.
Uh, from driver incentives to,uh, driver onboarding to new
users and coupons and thingslike this.
(03:27):
So that's when I really gainedan appreciation for the role of
data, um, and, and marketingand, and marketing technology.
Uh, that was kind of my firstintroduction to that.
Michael Hartmann (03:38):
So I, I'm interested in a coup.
Couple of follow ups, if youdon't mind, indulge me here.
So the.
Of the, the work in finance.
So if for our listeners, you maynot be as familiar with George,
I'm, I'm a big advocate thateverybody should at least learn
the basics of finance.
So I'm, it's interesting thatyou say you were in finance, but
just like it still applied towhat you did, you're doing now,
(04:01):
but not as directly.
Do you, do you still find thatyour, your education, your
knowledge of finance is stillvaluable at this point?
George Xing (04:10):
Yeah, I, I think for sure.
I think if nothing else thatbecame.
Pretty proficient at Excel, uh,when I was working there.
And I feel like that is almostthe foundation for any kind of
analytical number based job isyou, you start with a
spreadsheet and you figure outhow models work and, and I think
(04:32):
a lot of the foundationalprinciples, you know, now that
people talk about SQL and datawarehouses and how you slice and
dice numbers.
Build bi dashboards.
Uh, there's a lot of parallelsbetween that and just looking at
a spreadsheet at the end of theday.
And so I think that gave us,gave me a lot of foundations.
And again, I think it's probablyfor that reason that they hired
me at Indiegogo cuz I wasn'tqualified to do anything else.
Michael Hartmann (04:56):
Yeah.
You were, you're sort of anExcel Jackie, right?
Yeah.
Yeah.
And so it's interesting cuz I,um, although I wasn't the one
doing a lot of the modeling, Iworked at a, a.
A consulting company, second joboutta my, on my career that did
a lot of work with real estatecompanies.
So there was a lot of, um,cashflow modeling that went into
(05:16):
that for different kinds of realestate, you know, portfolios and
stuff like that.
And I, so I was around it enoughthat I learned, that's how I
learned.
And by the way, they all hatedExcel because they were so good
at Lotus 1 23 that, that's howit dates me.
You probably don't even, nevereven heard of it, right?
Cuz you could all do it withkeystrokes.
So Mike's over there laughing atme.
Mike Rizzo (05:36):
I know, like I totally know the tool, but I
never used it.
Um, but like I got my foray intotech and.
I always heard about it andnever really saw it so.
Michael Hartmann (05:47):
So I think at one point in early versions of
Excel, they like to try to getpeople to adopt it.
I think they kept like there wasa way you could turn on like.
How to use some of the lotuslike control keystrokes.
Right.
You could do slash whatever.
Right.
So that way you didn't have totake your hand off the keyboard
and use the mouse.
Right.
It was inefficient.
(06:07):
Anyway, I'm
Mike Rizzo (06:09):
so funny, there's like, there's like a whole email
product around this idea ofkeystroke, uh, managing your
inbox now too.
So it's come full.
Michael Hartmann (06:17):
Yeah.
George Xing (06:17):
my, my biggest, uh, transition when I was moved into
tech was, uh, moving from a,from a Windows machine to Mac
and not knowing my Excelshortcuts and not having the
right, so
Michael Hartmann (06:29):
I, I'm using a Mac right now and I love the Mac
though.
I always told you the one thingI do not like about it is that I
can't use the shortcuts in Excelthat I was so used to.
George Xing (06:38):
Yeah, totally.
Mike Rizzo (06:40):
funny.
Michael Hartmann (06:40):
it's, that's
Mike Rizzo (06:41):
It is a hard, I have to agree, it's a hard transition
to make.
Like my first role in a tech joblike tech SAS company, I was
issued a, you know, standardissue, least MacBook Pro or
whatever, and I was like, I wasa gamer man.
I built my own computers, right?
So I was a PC guy, like forsure, all the way.
I was looking at this thinglike, what do I do with this
(07:04):
Like, like it's a whole newlanguage.
And then as a marketer, I beganto value it greatly.
Um, just some of the stuff,yeah.
The Excel side of things.
Oh,
Michael Hartmann (07:14):
All right, so let me get us back on track.
So George, sorry I went a littlebit sideways there, but, um,
okay.
So that experience with IndieGogo and Lyft, um, said you
exposed you to data and reportsof the data for marketing go,
I'll call it go to market or, orrevenue functions, but so is
what was the, was.
The catalyst for you going, Hey,there's an opportunity to start
(07:37):
Super Grain, is that kinda whereit's, was that the seed that
where it started from?
George Xing (07:43):
Yeah, I, I would say, you know, the, the general
theme that I've been passionateabout, and I think that, you
know, really I cultivated whileI was at LI was just kind of
this desire to, um, help peoplemake better decisions with data.
And that was the, that was maybethe seed that.
Wanted me to start something.
I, I think I wanted to dosomething entrepreneurial for a
(08:05):
long time.
When I left Lift, it was rightbefore the pandemic, and then
shortly during the pandemic, Ispent a lot of time just
thinking about what I was gonnado next.
Took some time off, and thenwith this idea of how do I, um,
enable people withinorganizations to make better
decisions with data and do whatI was doing.
(08:27):
but helping other companies dothat as well.
Um, eventually, I would saythrough a number of iterations,
we ended up with Super Grainand, um, you know, the direction
that we're building in now.
Michael Hartmann (08:38):
Got it.
Okay.
So let's talk a little bit aboutwhat you're doing.
And I think this is, this wasreally fascinating.
I know when Mike introduced us,so, uh, and I may get this wrong
a little bit, but you, you just,super grains described as a
warehouse native, warehouse,native solution for marketing
and marketing ops, and maybeeven goes beyond that customer
experience.
So, for our listeners, can youdefine what you mean by
(09:03):
warehouse native solution andmaybe even give some examples
of, of what kind of scenariosthat would mean, or, you know,
use cases or maybe even compare,like today you would use.
Something that's not warehousenative.
And we don't have to pick anyspecific, say, marketing
automation platform, but cuz Ithink they're all fairly
(09:24):
similar.
Um, versus what you, how yousolve the kind of the
challenges.
George Xing (09:30):
Yeah, yeah, for sure.
And so, super Grain, uh, is a,is a warehouse native.
We're also B2B focused marketingautomation platform.
And what we really mean bywarehouse native is that we, uh,
there, there's a couple things.
One, we integrate directly withcloud data warehouses like snow.
And big query.
Um, I, I think the secondprinciple is that, uh, we're
(09:54):
using cloud data warehouses asthe source of truth for customer
data, uh, which is I think oneof the big differences, uh,
versus traditional, uh,platforms which ingest all the
customer data themselves andpresume to be the source of
truth for customer data.
And this includes CDPs and.
Um, you know, traditionalmarketing automation solutions
(10:15):
and a number of other customerengagement tools out there.
Um, but the key, uh, you know,so, so that's the key thing.
We integrate directly with, uh,the customer's cloud data
warehouse.
And to the extent that we can,we run compute and run a lot of
the operations directly on topof their infrastructure.
(10:35):
Um, maybe to talk about why we,we took this approach, I think.
With the big trend that we saw,and I started to see this while
was at Lyft as well, is thatcompanies are, are really
struggling to keep data acrossdifferent platforms in sync.
So they have customer data in anumber of different places.
(10:58):
It might be their crm, it mightbe a cloud data warehouse, it
might be a customer engagementtool, a S A Cs, um, platform.
And at the end of the day, it'sall the same data, but different
teams are making changes andupdates to, uh, you know, the
same data in different systems.
And you have this complex web ofpipelines to make sure that they
(11:19):
stay in sync.
Um, one of the things that westart seeing about how companies
are solving this problem iscentralizing all their customer
data in a cloud data warehouselike Snowflake or big.
And that becomes the source oftruth that powers all their
downstream systems of engagementwith customers.
And when we started to see that,we thought, Hey, naturally,
(11:42):
instead of building yet anotherdata platform that ingests
customer data, why don't we justsit on top of the data platform
that customers are ready, uh,converging around, which is
their own cloud data warehouse.
Mike Rizzo (11:57):
I have so many thoughts on this
Michael Hartmann (11:59):
Uh, like I, I was like, I've got a couple of
follow up questions too, Mike,but you go.
Mike Rizzo (12:03):
Um, you know, George, when you and I connected
for our listeners like.
we don't normally have founderscome on and talk about products.
Right.
Like as, as a sort of a rule forthe show.
That's something, something thatwe just, we don't generally do.
But the reason we wanted to talkto George was specifically on
this show was because thisconcept of warehouse native is
(12:25):
super new.
Like it, I, I would argue you'relike at sort of the cutting edge
of like this new set of adoptionthat's coming down the pike.
At least for me, when the firsttime I heard it, it was
definitely an aha moment, rightwhen I suddenly thought about,
wow, if there's an organizationthat's mature enough to try to
(12:47):
actually centralize theiroperations on a on snowflake or
big query on a data warehousescenario.
I have not worked for anyorganization that has done that
yet, However, for those that aredoing this, um, gosh, it does
make a lot of sense, uh, topotentially have these apps that
just sit right on top of that,um, and allow you to interact
(13:08):
with that data in a way that iseffectively is wholly owned.
Right?
Like, you know, to some degree.
My take on this in Georgia, I'dlove to hear your thoughts.
To some degree, Salesforce ownsyour.
HubSpot owns your data, right?
Like at the end of the day,they're sitting in their
infrastructure, their server'slike, yes, you own it, you own
(13:28):
the rights to it.
You can extract it and move itto somewhere else if you'd like.
Um, but it's certainly not yourown servers or anything like
that.
Um, you don't, you're notbuilding around your own sort of
infrastructure.
You're, you're, you'resubscribing to software as a
service.
Um, and.
What I'm seeing in the marketis, at least with stuff like
(13:49):
this, is this sort of shift backto, well, we actually,
especially with privacy andcompliance laws and all these
things coming together, weactually need to own our data
more holistically on our ownservers that, you know, maybe
have other software tools thathelp enable our interactions
with our data sets, uh, andpotentially the enterprise that
(14:10):
needs to now activate that dataand needs tools like a Super
Grain or something like.
To sort of get back into theflow of, you know, doing a go to
market motion.
Is that sort of like, am I, am Ioff?
Like is this like, is there aton of companies doing this and
I'm just not working at any ofthem and like, am I speaking
around in circles or
George Xing (14:30):
No, no, no.
You're, you're totally right.
And, and I, and I think you, youpointed out some of, a couple of
the key reasons that I thinkthat the warehouse native
approach, um, is compelling to anumber of people that we talk
to, which is, uh, pri you know,privacy and data, data
ownership, right?
So with a warehouse nativeapproach, instead of moving your
(14:53):
data into, um, another thirdparty, Uh, and copying it there
every single time.
You need to use it for emailautomation or, or CRM purposes.
You just kind of, it just sitson your own cloud, your own data
cloud.
And then you can imagine thesame way that you have an iPhone
(15:15):
and you can install applicationsthat run on your iPhone.
You can essentially have dataapplications that run on top of
the data that you already own.
Um, it's like if you have youriPhone, And you have all your
customer data there.
Instead of figuring out a way toplug in a USB stick or something
like that and transfer it toanother device, um, you just
(15:35):
have, uh, all these applicationsthat use the same underlying
data and they can also sharedata, uh, and communicate with
other apps that are alsoinstalled on your phone.
Um, that's maybe kind of theanalogy.
Um, and that's certainly a fewyears, probably down the road
before we get there, but that'scertainly the vision that.
(15:56):
I think a lot of this, uh, canpotentially get to, um, in a few
years once the technology isready, once the ecosystem
matures.
And it makes a lot of sense forthe end users because you don't
have to pay the cost of movingdata back and forth.
You don't have to worry aboutdata syncing issues or
inconsistency between differentapplications and different
(16:17):
platforms.
Um, and at the end of the day,you have a single source of
truth, uh, that, um, becomes alot easier.
Uh, activating and doingpersonalization and
segmentation, which, you know, Ithink a lot of folks probably
listen to this would agree.
One of the hardest parts of, um,you know, using product data and
(16:40):
other data that lives inside of,uh, a cloud data warehouse for
messaging purposes is actuallygetting the data into those
platforms.
So it also solves that problem.
Michael Hartmann (16:51):
Yeah.
So, um, I wanna make sure Iunderstood this right.
So it sounds like, is there, didI hear correctly?
There's a sort of a suppositionfor this.
Uh, warehouse native conceptthat the, the organizations that
are gonna be successful alreadyeither have or are in process of
(17:11):
building will keep it tocustomer a, a customer data
warehouse.
Is that, did I, I'munderstanding that right.
George Xing (17:21):
Yeah, I, I would see, I would say that one of the
things that we observe is thatonce companies get to a certain
stage, certainly the startupsthat we work with, um, when they
get to series A or series B, itbecomes part of the natural
evolution of how they thinkabout their data maturity.
So, um, around that time, theystart to invest in, uh, many
(17:42):
different systems.
They start realizing the painof, uh, having data in a bunch
of different places.
And there's a solution out ofthe box where now they will, um,
invest in a data warehouse,start centralizing all that
information there.
Also start building out datacapabilities to manage the data
that lives inside of thatwarehouse, and then start
(18:03):
thinking about how to leveragethat single source of truth in
downstream places.
And so that evolution happens alittle bit earlier for some
companies, a little bit laterfor other companies, but as a
whole, um, certainly becauseit's now easier than ever before
to set up a data warehouse.
It's happening, um, I would sayearlier than, than, uh, it was
(18:25):
as an industry, uh, even acouple years ago.
Michael Hartmann (18:29):
Interesting.
Yeah.
I mean my, my head went straightto like, I've, the idea of
simply defining terminology thateveryone agreed to is what is a
customer, right?
Seems like it should be an easything to define, but it's
usually like I've only been atone place that's even come close
to doing it.
Um, so do, I mean, do you runinto those challenges where
(18:50):
these companies, um, are stillstruggling with.
The, the taxonomy, so to speak,of what they're doing in terms
of that data.
So they have the, they have thislike, Hey, this need to do, they
realize it to to scale orwhatever, to grow the way they
need to.
They need to be more deliberateand conscious about how they
manage customer data.
(19:11):
Have they gone through, are theytypically still going through
that process of DEF definitionsor are they usually beyond that
at that point?
George Xing (19:20):
Yeah, I, I would say it's a spectrum, you know,
as with most things are, um,typically we see that, uh, the
first use case is that there's aparticular.
Piece of product data or someproduct metric that is really
important to the business thatthey wanna use in some type of
(19:41):
email campaign or for some typeof onboarding flow.
So to, to make this reallyconcrete, um, let's say they,
you know, a customer wants tosend, um, and kind of like a
upsell message to the admins of.
One of their customers everysingle time there's five, uh,
(20:02):
new users on that account.
Right?
Uh, when, when that accountreaches five total users, well
that calculation, like thenumber of active users on that
account is a calculation that isdone inside of the, the cloud
data warehouse.
So one of the first thingsthat'll see is they'll throw all
(20:23):
the data that they need intotheir cloud data.
W.
In order to calculate thatmetric.
And then the next step isgetting that metric into, you
know, the system of engagementfor doing marketing automation
to support that onboarding flowthat they want to, um, want to
do.
And, you know, I, I think likeover time they will expand and
(20:47):
calculate more metrics.
They will build more of like acohesive taxonomy around all of
this.
But it really just starts withthat first use.
And then spans, expands fromthere.
Mike Rizzo (20:59):
I love, I love that you bring that up.
Just earlier today I was on.
Call with a community member,and we were talking about, um,
how challenging that that exactproblem is like in marketing
operations.
Uh, we are frequently asked, Ithink, at least in the B2B SAS
space, right?
Like, how do you do anonboarding nurture?
(21:21):
How do you do an activation ofsome kind for all these users?
How do you do.
Feature use case usage basedsort of marketing activities.
And the answer is, is like it'sreally hard to do, like even
understanding whether or not afeature is currently in use.
Like a lot of, a lot of folksthat haven't ever had a chance
to talk to a product team or adeveloper, uh, they don't
(21:44):
realize that, like, that, thatisn't inherently built into the
code base to just send thatstuff around.
Right.
You know, the, the, theengineers are not sitting there
going, wow, the marketer's gonnawant to know when a user does
this, so I'm gonna make sure Iwrite a piece of code.
Like, that's not what they'redoing.
They're there to build reallygood product, you know, and, and
one of the reasons why I lovethe idea of pairing up, you
(22:06):
know, marketing apps andmarketers with, with, uh,
product teams, is to try to getto those questions a little
sooner rather than later.
But that's fundamentally, it'sreally hard to do.
And so, you know, I love thatyou're tackling.
With like sort of a layer of,uh, of technology that can like,
at least facilitate that right
George Xing (22:26):
Yeah.
Yeah, for sure.
And, and this is the exact kindof use case that we're trying
to, trying to go for because wesee it so frequently.
Um, one of the other things Ithink that, uh, I just thought
of as, as you were talking aboutkind of.
This interplay between theproduct side and what the
engineer's instrumenting andwhat the marketer's able to to
(22:48):
use in terms of the data is alsojust that I, I think some of the
lines that we see, at least, youknow, from the conversations
that we have between, uh,marketing, marketing ops and
growth teams, which often haveengineers and PMs, those start
to blur because who really ownsan onboarding flow really, you
(23:09):
know, Is it the pm, the growthpm who's kind of, uh, sending in
app notifications?
Um, or is it, you know, themarketing team that's kind of
creating this nurture series,uh, who are really hitting the
same customers with very similarmessaging just through different
(23:30):
channels?
Or is it, you know, the CS team?
Is, uh, kind of reaching out tosomebody who's going through a
free trial and trying to getthem to, uh, you know, debug
their data connection becausethey have trouble adding out
data source or getting reallyvalue outta the product.
And, and so one of the thingsthat we see a lot is, okay, you
(23:52):
have a marketer that's actuallysending out an email and signing
that as like a CS manager, uh,in the byline, or same thing
happens with sales.
Um, and, and a lot of theselines start blurring, even
organizationally.
Uh, you just have kind of oneperson sometimes that smaller
companies that's doing, doingeverything, and maybe kind of,
(24:15):
you know, one, one takeaway thatI have is I wonder if we're
headed for a future where thingsare much more aligned
organizationally around thecustomer journey rather than,
You know, strict handoffsbetween different parts of the
customer funnel because, um,certainly one of the things we
see with as more product databecomes, um, important in the
(24:40):
B2B marketing journey is that,uh, it's much more about
there's, there's no linear pathfrom or hand off between one
function to the other.
Uh, it's all about kind of thecustomer journey, which is very
business.
So I'm curious if you guys seethat in, in, um, in, in some of
(25:02):
the things that Yeah.
Mike Rizzo (25:04):
I definitely do.
And I, I think, I think thatthere's a chance for, you know,
we're seeing more of this like,um, you know, adoption of like,
what do we believe the customerjourney should be?
Um, and then how do we sort ofimplement that, right.
And I think like one of the corechallenges that I've come across
(25:29):
is, The, the realization thatlike there isn't, there's no
like, um, silver bullet thatlike solves all of the, the
problems, right?
Like we can't, there's no oneanswer to any of this thing
Michael Hartmann (25:43):
You wanna repeat that?
Cause I wanna make sureeverybody heard that.
Mike Rizzo (25:47):
right?
George Xing (25:47):
there's no silver
Mike Rizzo (25:48):
There's no silver bullet that, that solves this.
And George, I think for you,like it's a dangerous hole as a
product owner.
You're probably, you're gonnahear from investors and, and
people who are using yourproduct, they're gonna be like,
well, can you build AI to tellus what the user journey should
be like?
You know?
And like the thing is, is for
George Xing (26:09):
don't worry.
We've already gotten that.
Yeah.
Mike Rizzo (26:11):
see, not
Michael Hartmann (26:12):
imagine.
What's the build in the bestpractice?
Mike Rizzo (26:15):
Right.
Right.
I think, I think my As, assomeone who.
Is now a curator of like,programming that is community
led, right?
Like I am, I'm saying like, Hey,what, what does everybody really
want?
And how can we try to bring thatto the, to this community?
Um, I think if you can, to thebest of your ability, try to try
(26:38):
to work alongside your currentmarket, your customers, to
figure out what it is that they,what they believe their journey
should be and how they want toadopt your product.
But I think, don't be afraid to.
Create, like carve your own sortof path, right?
Um, imagine like an art gallery,right?
When someone puts an art gallerytogether, they actually want to
(27:00):
tell a story through thatjourney on your sort of trip
down, this like one particularartist's like timeline or maybe
a genre and a set of categor.
That person who put that gallerytogether did it with intention.
And the turns, the lefts and therights and the decisions that
you get to make are the storythat they want you to
(27:21):
experience.
And so I think it's like on usto also try to figure out like,
what do we believe the journeyshould be?
And like let's hedge our betsand try that for a little while,
and then take people down thatpath.
But to your point earlier,George, like the lines are
blurred and it very much dependson your business.
Like maybe CS people are tryingto do upsells through the PLG
(27:44):
motion.
Maybe sales people are doing it.
Maybe the product people aredoing it.
Maybe the growth team is doingit.
George Xing (27:50):
Mm-hmm.
Mike Rizzo (27:51):
But all of that, no silver bullet one and two.
You know what?
Just come up with the journeyYou want to take them on, like
curate the experience that youwant them to have and focus on
just that.
And don't divert like,
Michael Hartmann (28:05):
you gotta give it a chance.
Yeah, we've been, so I thinkit's interesting that you asked
that question, George, becauseit Yeah.
We've been doing some thinkingabout, um, re.
Rethinking our whole go tomarket, particularly in the
marketing side approach for ourbusiness to think around it in
terms of a, a marketing funnel.
(28:26):
And I have, honestly, I havemixed emotions about, or mixed
feelings about it because Ithink it, it, on the one hand it
does provide, uh, some amount ofstructure so that we're
thinking.
At least having common way ofthinking about how we're going
to market and so we can measureit and all that kinda stuff.
At the same time, I just don'tbelieve that's the way that
people buy and we we're not evenin a tech space.
Right?
I don't just don't think that'sthe way people buy even a B2B
(28:48):
space.
Right?
They, when they're, and I thinkof that for my own stuff, like
when I'm, when I have anopportunity to research
something that I'm just maybejust curious about, I'll spend a
fair amount of time doing thatand then I might not do
anything.
For weeks or months.
And then cuz I'm kind of stillthinking about it, right?
And it's not this very linearthing, it's very chaotic almost.
(29:09):
So, um, I do, I think that'spart of the challenge too, is
that, but I think it does matterwhere, where it does matter is
if you can put this informationin front of anybody who might
interact with that customer orpotential customer, regardless
of where they are in theorganization.
I think that's where you startto have the opportunity to
really build trust with thosepeople, which then builds
(29:31):
loyalty and, and advocacy if youget to that point.
But without that, it's reallyeasy, you know, if your customer
success or customer supportteams don't have visibility into
what's going on, marketing andsales, they can really kill a
deal, right?
Um, so, or a salesperson mightcome in at the wrong time when
there's a pro a product issue,right?
George Xing (29:53):
Yeah, yeah, yeah.
I mean, we hear this all thetime.
It's, it's, you know, People,people say, what's the classic
thing where they say, um, yourproduct reflects your org
structure, right?
Um, yeah.
And, uh, or you ship your orgstructure and, and I think this
is a case where, um, the toolsthat we use reflect kind of the
(30:18):
org structures that have beenkind of like set.
As best practice, right?
There's a CS team, there's asales team, there's a marketing
team.
So of course, all those teamshave their own dedicated tools
to manage customercommunications for their part of
the funnel.
But what happens when you knowthe funnel isn't quite so
(30:39):
linear?
And when those lines get blurreda little bit, well then those
tools still kind of then, thenthose tools basically start
overlapping and um, and itbecomes kind of a coordination
c.
we see this all the time.
Um, we talk to customers all thetime who are, you know, large at
scale companies that everyone'sheard of, um, and they still
(30:59):
struggle with, uh, this today.
And, you know, even basic thingslike subscriptions and, uh,
making sure that you don'treceive an email from another
system if you've unsubscribed,um, you know, from one.
Um, those things are, are.
(31:20):
Are, um, are hard to, to kind ofmanage, uh, which could be
surprising.
Um, I think at least, you know,as, as someone who doesn't come
from a marketing background, um,that was surprising kind of to
hear.
But it just kind of shows kindof maybe the complexity of some
of these challenges and alsojust the inherent nature and
(31:41):
evolution of some of the thingsthat we're seeing in the market.
Michael Hartmann (31:44):
Yeah, I, I think it's, it's really
interesting and, but I cantestify to the fact that there
are definitely companies wheremultiple systems are not in sync
when it comes to things like optouts.
So definitely think that's thecase.
So it, so what I wanna kind goback to a little bit, you, um,
(32:07):
When you and I first talked, youknow, and you mentioned it
already, right?
I, I immediately was thinkingabout CDPs and I think around
the time we talked, we actuallyhad a guest not long before that
on talking about CDPs.
Cause it was something that I'dbeen hearing about, didn't know
much about.
So how do you, like, what's thedistinction between the, the.
Kind of warehouse nativesolution concept and CDPs.
(32:30):
Is it, are they competingthings?
Are they compat?
Like are they complimentary?
Like what's the, what's thedifference there?
How do you see, see those?
George Xing (32:39):
Yeah, I mean it, I think it really comes down to
the source of truth distinctionthat I was talking about
earlier.
Um, what CDPs do is they ingest.
Customer data from a number ofdifferent source systems.
It could be your eventcollector, it could be third
party tool, it could be yourcrm.
And then they let you do, uh,they do identity resolution,
lets you do segmentation on topof it and then, you know, send
(33:02):
it downstream usually to othertools for ad targeting, emails,
et cetera.
And, you know, I, I.
The big kind of shift in thewarehouse native approach is,
you know, you don't need aseparate data store to do all
this organization andcentralization and identity
stitching.
You actually just want yourcloud data warehouse to be the
(33:25):
place where all that happens.
And then you're accepting the,the, the cloud data warehouse as
truth and, and so concretelywhat that means, You're putting
the onus or, or giving thecontrol of, um, the data
modeling exercise, the identityresolution, how you want to
(33:49):
calculate certain metrics to thecustomer rather than being
opinionated about, okay, this iswhat an active user is, or, you
know, this is exactly how you,you can choose between these two
options for identity resolutionand, and because every business
is the.
That is, is not the same.
Rather, um, that means that, uh,you're just gonna be able to
(34:13):
have a lot more flexibility andend up with data models,
metrics, um, identity, uh, thatis tailored to your business,
uh, versus something that comesout more out of box that you
don't have much control over.
Um, that's, that's kind of likemaybe just from a.
(34:34):
Usability standpoint.
And then the other thing I wouldsay is that kind of goes hand in
hand is that CDPs typically,again from from our
conversations, take a long timeto implement because you're
migrating a lot of data, you'redoing all this crunching, um, in
(34:55):
like a separate system.
If you can do all thatcalculation computation inside a
cloud data warehouse.
The customer is alreadyoperating, then it simplifies
the process to implement fromsometimes months to days.
And you know, obviously that's ahuge, uh, difference when it
comes to business results andgetting up and running.
Michael Hartmann (35:18):
Okay, so let me, I wanna see if I can play
this back cuz I think I,something just hit me.
So I think what I'm hearing, soCDP based on, not what I
understand, what you described,it's sort of pulling in data and
then pushing it back out.
Or some version of it back outto these different systems.
They all have their owndatabases and data structures.
(35:40):
Um, and maybe it's doing somecalculations, but maybe those
applications are doing'em aswell for their specific needs.
Is that, uh, do I have thatabout right?
For cdp?
George Xing (35:52):
Yeah.
Michael Hartmann (35:52):
Okay.
All right.
But, and so I think what I'mhearing differently with this
warehouse native is that ratherthan, um, Sort of ingesting,
pulling data in from all thesedifferent apps or, or, or
solution platforms?
The, the database is thedatabase and those, these, if
(36:13):
you have warehouse nativeplatforms, they're basically
just apps that are accessing,um, and maybe sending some data
back, but it's based on probablytransactional level kind of
stuff as opposed tocomputational kinds of things.
Is that the way, am Iunderstanding that right?
George Xing (36:32):
Um, yeah, I, I would say the, the, the CDP is
requiring kind of, it's tryingto be the source of truth and,
and in the absence of a clouddata warehouse, you basically
have to do everything that acloud data warehouse does and
then more, and whereas, In awarehouse native approach is
you're, you're saying, okay, youhave a data warehouse, so I'm
(36:54):
not gonna do all the backendprocessing, I'm not gonna do all
the computation, the ingestion,and everything I would have to
do as a cdp.
Um, I'll just do kind of the,almost the UI layer or the
orchestration layer, so thesegmentation, the business logic
and the actual activation.
And so it's kind of decouplingthe, the database part from.
(37:19):
The interaction and the UI andthe application side.
A CDP kind of combines the twoand one because in a world where
you didn't have, you know, clouddata warehouses, you had to do
both.
Michael Hartmann (37:30):
Yeah.
Well, and so reason I was tryingto clarify is where my head went
to, cuz as I've already likeshared, right?
I'm old enough to remember backwhen it was like mainframes and
terminals and client server wasa thing like this sounds very
much like that.
I mean, it's.
That's kind of was, it's notquite the same, uh, but
certainly, definitely notmainframe terminal, but the
(37:52):
client server.
That concept of where you'resort of splitting some of the,
the, um, processing effortsright between different sides
of.
The, the, it sounds somethinglike that, which, so it's
interesting to me to see kindathe evolution has been to go to
all these sort of, um, specificapplications with their own
(38:14):
databases and their purposes toone that's more like kind of
going back to old school withnewer technology to support the
ability to distribute the, thedata and all that.
So, To me that's interesting,right?
Like I think I finally, thathelped me click on how to
differentiate this from thingslike cdp.
So tell me if I'm I'll totallyoff here.
George Xing (38:36):
yeah.
No, no.
I think, no, I think that'sexactly right.
I mean, you, you need, um, youneed one database at the, at the
end of the day, right?
And.
Your customer data should bestored in, in one place.
You don't need multiple copiesof it.
If you can have everythingpointed to that single copy and,
um, you know, in a world whereyou have a cloud data warehouse
(39:00):
that is your copy of the data,uh, you don't need a CDP plus
another tool, plus another toolthat all have separate copies of
your data.
Um, and uh, and, and I think,you know, that's why.
You know, you're starting to seea lot of companies move from,
you know, their CDPs over tokind of, uh, adopting cloud data
(39:22):
warehouses,
Mike Rizzo (39:26):
Makes a ton of sense.
I'm just like, I don't know,heart man.
You might have one more questionand, and so we could go with
yours.
Michael Hartmann (39:34):
I'm, I'm, I'm already realizing we, like, we
could have, we, we may need tofollow up on this one.
Mike Rizzo (39:39):
right.
Um, my, my question is, is like,you know, do you feel.
Who's responsible for this,like, transformation around how
to think about the activation ofthis data and like, is it, is it
a collective responsibility?
Like, like, you know, peoplejust need to get in a room and
(40:02):
talk and like, and like episodeslike this can help educate them
on this is why we need to talkabout it.
Or is it it like, I don't, who'sowning
Michael Hartmann (40:12):
Is it?
Is it a business operations?
Is it, yeah.
No, I think that's actually areally, really good question.
So I'm curious to hear whatyou're seeing out there, George.
George Xing (40:21):
Yeah, I like the short answer is where we think
about this question a lot.
Um, cause in our go to market,it obviously kind of ha has a
big impact on.
And where, you know, who we goafter and who we talk to, um,
and how we position ourselves.
I think one thing that, whatI'll say is, and I don't think I
(40:42):
have like a silver bullet answerhere either, but I, you know, I
can share some things that weobserve.
Um, one is that I think, uh, youknow, obviously the adoption of
cloud, cloud data warehousesmakes it easier for companies
earlier.
To get their data into a bettershape.
So what I mean by that is, uh,if you talk to me two years ago,
(41:07):
right, the average series Bcompany would probably not have
their data in a good place.
Like, um, they might not evenhave a data team.
Whereas today, the averageseries a company that we talk.
Has a cloud data warehouse andhas at least one person who's
managing it.
(41:28):
So just kind of the shift ofdata maturity, um, you
Mike Rizzo (41:33):
that's awesome.
George Xing (41:34):
Yeah, yeah.
And, and what that means isthat, you know, um, the people
who really benefit from this arethe downstream stakeholders.
So, uh, marketing teams go tomarket teams that consume.
I think they're just gonna haveaccess to higher quality data
faster in the life cycles.
So you're gonna start to seeearlier stage startups have more
(41:55):
sophisticated life cycle andmarketing programs.
I think that's, that's thedirection that we're headed and
that reflects a need generallyin the market where companies
are collecting more and moreproduct data earlier, earlier in
their life cycles and doing morepersonalization and targeting
and workflows based on thatproduct data.
Um, I think on the other, talkto marketers.
(42:17):
Um, you know, we also kind ofsee a similar type of
convergence where, um, you know,I think there's this stereotype,
uh, at least, um, and I thinkit's kind of wrongly placed
that, you know, marketers arenot technical or data savvy.
(42:37):
I think it's actually theopposite.
Um, I think marketers areactually very data savvy and we
start.
Marketers are actually quitetechnical, very proficient, able
to speak about, um, technicalconcepts and understand the way
that their data is structuredand interface very directly with
their data team and say, no,this is, I need the data in this
(42:58):
format.
Um, I need you to get me thisdata, point a certain way, talk
to their product teams, talk toengineers, and be able to ask
for what they need in a veryspecific way.
That wasn't the case.
Three, five years ago when I wasworking, you know, at Lift.
And, and I think what that meansis there's just a much tighter
(43:21):
collaboration between thoseteams, and I think it means that
they're able to get what theyneed faster.
I think there's just betterdialogue and better
collaboration.
Mike Rizzo (43:32):
Mm-hmm.
George Xing (43:32):
Um, so those are two things that I see from my
kind of like selfish standpoint.
I hope that this trend acceler.
Because it'll make our job a loteasier in terms of educating the
respective people, um, gettingdata people to appreciate the
go-to-market challenges and the,and the business use cases, but
(43:53):
also getting kind of the, thego-to-market teams to appreciate
the technology considerations.
Um, that's a lot of where we sitand trying to kind of pull those
two teams together and get themto talk to each other.
Michael Hartmann (44:07):
Yeah,
Mike Rizzo (44:08):
I really like that, um, George, and I appreciate
your answer.
Um, hopefully people tune intothis episode and we can
accelerate that, that learningas well, uh, with you.
So thank
Michael Hartmann (44:18):
Well, I, I think what was really, really
caught my attention when wefirst started talking about this
was just, I think in the back ofmy mind and probably said it out
loud and public too a few times.
Like it feels like there's been,even with all the volume of new.
Marketing, revenue technologycompanies out there.
(44:39):
They seem to be really niche.
There's not really beensomething that seemed really
sort of a, a a, I hate to usethis paradigm shift, but like a
significant change in direction,um, like this.
So that, um, it's, it's reallyinteresting and I, I'm gonna be
really curious to see how itplays out.
(44:59):
So, George, thanks for joiningus.
Anything, anything last minute.
And truly, like I have a wholeset of new questions that
weren't even like, we didn't getto everything we had planned on
even here.
So, um, anything like last bitof nuggets you wanna share with
our listeners before we, we, wesign off here.
George Xing (45:20):
Um, I, you know, I think that the, the one thing
that I, I'll say that I learnedis, uh, or, or, you know,
learned from working with ourcustomers is that, um, Almost
universally, you tend to kind ofunderestimate the benefits of,
um, being able to use betterdata to run the same campaigns.
(45:43):
It's like, you know, we talk topeople that say, okay, yeah,
like we have data that we're notusing product data that we're
not using, but we haven't reallyprioritized improving this
onboarding flow because, youknow, we got like 50 other
things that we're working on.
And I, and I say, okay, wellyou.
We can run a really likelightweight, simple poc.
(46:05):
It'll take like a couple hoursto get started and then, you
know, at that point, if you cantell, if you see no results or
uh, then no harm, no foul.
Right?
But why not give it a try?
And inevitably, you know, Ithink people under estimate, and
it's not because we have amagical tool or anything, right?
(46:25):
It's just because people alreadyhave data that they're not
leveraging and.
The impact of personalizing thesame emails with just a little
bit more, or, um, being able todo a little bit of more routing
logic to send a specific segmentof users different messages or
messages that they wouldn't havereceived before, um, will drive
(46:49):
conversion at like a verycritical step of their
activation funnel.
I think especially for foundersand, and people that haven't run
a lot of these programs before,your intuition is that, hey, it
might not matter that much inthe beginning, but you know, uh,
it turns out that in generalpeople underestimate the impact
(47:10):
that it could have.
So, um, I would encourage peopleto give it a shot.
Michael Hartmann (47:15):
no, I think, I mean that's what's something
I've been preaching for a whileis to focus not on like these
big, easy to measure conversionfor things, but looking at.
Micro conversions through acustomer's journey and, and
really focusing on improvingthose incrementally and they
have a multiplier effect, right?
Can really make a difference.
So, fascinating stuff, George.
Thanks for, thanks for joiningus today.
(47:36):
Um, if folks want to keep upwith you, connect with you or
learn more about Super Grain,what's the best way for them to
do that?
George Xing (47:44):
You can go to super grand.com, which is our website.
Uh, also feel free to shoot me anote at George Super Grand dot.
Always, uh, excited to haveconversations like this, chat,
marketing, uh, chat data.
So, um, don't hesitate to reachout.
Thank you.
Michael Hartmann (48:00):
Yeah.
Great.
George has been a pleasure.
I my mind is really now, Mike,thank you as always
Mike Rizzo (48:07):
Thank you, George.
Michael Hartmann (48:08):
we'll get, Naomi on next time as well,
hopefully, and to all you outthere listening and, uh,
continuing to support us, wethank you and continue to give
us your feedback and support andideas and suggestions for topics
and or.
Guess so.
Until next time, we'll talk toyou later.
Thanks everyone.
Bye.
Mike Rizzo (48:26):
Bye.
Hello everyone.
Welcome to another episode ofOpsCast, brought to you by
marketing ops.com, powered bythe Mo Pros.
I'm your host, Michael Hartmann.
Joined today by one co-host.
Today I've got Mike Rizzo by myside.
So joining us today is GeorgeXing, co-founder and CEO of
Super Grain, a customerengagement platform.
So prior to and founding SuperGrain, George held multiple
(00:21):
roles at Lift, includingdirector, head of decision
science products, and was a datascientist at Indie Gogo.
Before that, he was a fixedincome analyst list with Morgan
Stanley.
Uh, which I find fascinating,the combination of things here.
George, thanks for joining ustoday.
George Xing (00:36):
Thanks, Michael.
Thanks guys.
Uh, super glad to be here andexcited to chat.
Marketing automation andmarketing ops.
Michael Hartmann (00:43):
Awesome.
Yeah, no, I know we talked alittle bit ago, what I was
actually looking back, it was,we're now late November
recording this.
It was late August I think.
So it's been a while since wechatted, so I'm sure there's
been even more movement, whatyou're doing.
But before we get into sort ofthe novel way that Super Green
is approaching, uh, marketingautomation, and I think it goes
beyond that, we'll get that in alittle bit.
(01:05):
Like, can you first maybe.
Share a little more about, youknow, your career and your
journey and how it led tostarting super grade as a
beginning point.
Let's start from there.
George Xing (01:16):
Yeah, as you mentioned, uh, thanks for that
intro by the way.
Uh, started my career actuallyin finance, which is like
totally different on a fixedincome trading floor.
working at a bank in Hong Kong,so literally the other side of
the world.
Um, and I think the only commonthread was that there was
something numeric and somethinganalytical about the work I was
(01:37):
doing.
Other than that, I hated workingin finance.
So after about a year I movedover, uh, you know, came back to
the States, wanted to be in theBay Area.
I'm originally from the Eastcoast.
Super cold winters.
Didn't wanna be in New Jersey,so came out.
Joined a startup calledIndigogo.
And the only role that I wasqualified to do was something,
(01:58):
uh, that at the time was, uh,they call data analytics, which
I wasn't familiar with, butturned out to be something that,
uh, was useful for the business.
And, uh, from there, the rest ofmy career has been in data and
analytics.
So spent a couple years atIndiegogo.
Um, and then, uh, while I was inSan Francisco, I started using.
(02:21):
This ride sharing service calledLyft, um, ended up joining them
as the, the first data hirethere.
And then, you know, uh, it justkind of as the company grew, uh,
I started working on datascience, business intelligence,
data engineering, and built outa number of teams there.
Um, and it was really at Liftwhere I saw a lot of the
(02:42):
interactions between data and alot of the stakeholders, um,
marketing teams, growth.
and a lot of the challenges thatwere involved.
So a lot of my job was justsitting at the intersection of
data decision making, figuringout how to get the right data,
sets the right data points tothe right people in order to
make decisions.
And whether that was, um,marketing decision makers,
(03:06):
whether that was running emailcampaigns were, whether that was
running automation.
So as you can imagine, at Lift,we had a number of very, very
complex, um, automations foreveryth.
Uh, from driver incentives to,uh, driver onboarding to new
users and coupons and thingslike this.
(03:27):
So that's when I really gainedan appreciation for the role of
data, um, and, and marketingand, and marketing technology.
Uh, that was kind of my firstintroduction to that.
Michael Hartmann (03:38):
So I, I'm interested in a coup.
Couple of follow ups, if youdon't mind, indulge me here.
So the.
Of the, the work in finance.
So if for our listeners, you maynot be as familiar with George,
I'm, I'm a big advocate thateverybody should at least learn
the basics of finance.
So I'm, it's interesting thatyou say you were in finance, but
just like it still applied towhat you did, you're doing now,
(04:01):
but not as directly.
Do you, do you still find thatyour, your education, your
knowledge of finance is stillvaluable at this point?
George Xing (04:10):
Yeah, I, I think for sure.
I think if nothing else thatbecame.
Pretty proficient at Excel, uh,when I was working there.
And I feel like that is almostthe foundation for any kind of
analytical number based job isyou, you start with a
spreadsheet and you figure outhow models work and, and I think
(04:32):
a lot of the foundationalprinciples, you know, now that
people talk about SQL and datawarehouses and how you slice and
dice numbers.
Build bi dashboards.
Uh, there's a lot of parallelsbetween that and just looking at
a spreadsheet at the end of theday.
And so I think that gave us,gave me a lot of foundations.
And again, I think it's probablyfor that reason that they hired
me at Indiegogo cuz I wasn'tqualified to do anything else.
Michael Hartmann (04:56):
Yeah.
You were, you're sort of anExcel Jackie, right?
Yeah.
Yeah.
And so it's interesting cuz I,um, although I wasn't the one
doing a lot of the modeling, Iworked at a, a.
A consulting company, second joboutta my, on my career that did
a lot of work with real estatecompanies.
So there was a lot of, um,cashflow modeling that went into
(05:16):
that for different kinds of realestate, you know, portfolios and
stuff like that.
And I, so I was around it enoughthat I learned, that's how I
learned.
And by the way, they all hatedExcel because they were so good
at Lotus 1 23 that, that's howit dates me.
You probably don't even, nevereven heard of it, right?
Cuz you could all do it withkeystrokes.
So Mike's over there laughing atme.
Mike Rizzo (05:36):
I know, like I totally know the tool, but I
never used it.
Um, but like I got my foray intotech and.
I always heard about it andnever really saw it so.
Michael Hartmann (05:47):
So I think at one point in early versions of
Excel, they like to try to getpeople to adopt it.
I think they kept like there wasa way you could turn on like.
How to use some of the lotuslike control keystrokes.
Right.
You could do slash whatever.
Right.
So that way you didn't have totake your hand off the keyboard
and use the mouse.
Right.
It was inefficient.
(06:07):
Anyway, I'm
Mike Rizzo (06:09):
so funny, there's like, there's like a whole email
product around this idea ofkeystroke, uh, managing your
inbox now too.
So it's come full.
Michael Hartmann (06:17):
Yeah.
George Xing (06:17):
my, my biggest, uh, transition when I was moved into
tech was, uh, moving from a,from a Windows machine to Mac
and not knowing my Excelshortcuts and not having the
right, so
Michael Hartmann (06:29):
I, I'm using a Mac right now and I love the Mac
though.
I always told you the one thingI do not like about it is that I
can't use the shortcuts in Excelthat I was so used to.
George Xing (06:38):
Yeah, totally.
Mike Rizzo (06:40):
funny.
Michael Hartmann (06:40):
it's, that's
Mike Rizzo (06:41):
It is a hard, I have to agree, it's a hard transition
to make.
Like my first role in a tech joblike tech SAS company, I was
issued a, you know, standardissue, least MacBook Pro or
whatever, and I was like, I wasa gamer man.
I built my own computers, right?
So I was a PC guy, like forsure, all the way.
I was looking at this thinglike, what do I do with this
(07:04):
Like, like it's a whole newlanguage.
And then as a marketer, I beganto value it greatly.
Um, just some of the stuff,yeah.
The Excel side of things.
Oh,
Michael Hartmann (07:14):
All right, so let me get us back on track.
So George, sorry I went a littlebit sideways there, but, um,
okay.
So that experience with IndieGogo and Lyft, um, said you
exposed you to data and reportsof the data for marketing go,
I'll call it go to market or, orrevenue functions, but so is
what was the, was.
The catalyst for you going, Hey,there's an opportunity to start
(07:37):
Super Grain, is that kinda whereit's, was that the seed that
where it started from?
George Xing (07:43):
Yeah, I, I would say, you know, the, the general
theme that I've been passionateabout, and I think that, you
know, really I cultivated whileI was at LI was just kind of
this desire to, um, help peoplemake better decisions with data.
And that was the, that was maybethe seed that.
Wanted me to start something.
I, I think I wanted to dosomething entrepreneurial for a
(08:05):
long time.
When I left Lift, it was rightbefore the pandemic, and then
shortly during the pandemic, Ispent a lot of time just
thinking about what I was gonnado next.
Took some time off, and thenwith this idea of how do I, um,
enable people withinorganizations to make better
decisions with data and do whatI was doing.
(08:27):
but helping other companies dothat as well.
Um, eventually, I would saythrough a number of iterations,
we ended up with Super Grainand, um, you know, the direction
that we're building in now.
Michael Hartmann (08:38):
Got it.
Okay.
So let's talk a little bit aboutwhat you're doing.
And I think this is, this wasreally fascinating.
I know when Mike introduced us,so, uh, and I may get this wrong
a little bit, but you, you just,super grains described as a
warehouse native, warehouse,native solution for marketing
and marketing ops, and maybeeven goes beyond that customer
experience.
So, for our listeners, can youdefine what you mean by
(09:03):
warehouse native solution andmaybe even give some examples
of, of what kind of scenariosthat would mean, or, you know,
use cases or maybe even compare,like today you would use.
Something that's not warehousenative.
And we don't have to pick anyspecific, say, marketing
automation platform, but cuz Ithink they're all fairly
(09:24):
similar.
Um, versus what you, how yousolve the kind of the
challenges.
George Xing (09:30):
Yeah, yeah, for sure.
And so, super Grain, uh, is a,is a warehouse native.
We're also B2B focused marketingautomation platform.
And what we really mean bywarehouse native is that we, uh,
there, there's a couple things.
One, we integrate directly withcloud data warehouses like snow.
And big query.
Um, I, I think the secondprinciple is that, uh, we're
(09:54):
using cloud data warehouses asthe source of truth for customer
data, uh, which is I think oneof the big differences, uh,
versus traditional, uh,platforms which ingest all the
customer data themselves andpresume to be the source of
truth for customer data.
And this includes CDPs and.
Um, you know, traditionalmarketing automation solutions
(10:15):
and a number of other customerengagement tools out there.
Um, but the key, uh, you know,so, so that's the key thing.
We integrate directly with, uh,the customer's cloud data
warehouse.
And to the extent that we can,we run compute and run a lot of
the operations directly on topof their infrastructure.
(10:35):
Um, maybe to talk about why we,we took this approach, I think.
With the big trend that we saw,and I started to see this while
was at Lyft as well, is thatcompanies are, are really
struggling to keep data acrossdifferent platforms in sync.
So they have customer data in anumber of different places.
(10:58):
It might be their crm, it mightbe a cloud data warehouse, it
might be a customer engagementtool, a S A Cs, um, platform.
And at the end of the day, it'sall the same data, but different
teams are making changes andupdates to, uh, you know, the
same data in different systems.
And you have this complex web ofpipelines to make sure that they
(11:19):
stay in sync.
Um, one of the things that westart seeing about how companies
are solving this problem iscentralizing all their customer
data in a cloud data warehouselike Snowflake or big.
And that becomes the source oftruth that powers all their
downstream systems of engagementwith customers.
And when we started to see that,we thought, Hey, naturally,
(11:42):
instead of building yet anotherdata platform that ingests
customer data, why don't we justsit on top of the data platform
that customers are ready, uh,converging around, which is
their own cloud data warehouse.
Mike Rizzo (11:57):
I have so many thoughts on this
Michael Hartmann (11:59):
Uh, like I, I was like, I've got a couple of
follow up questions too, Mike,but you go.
Mike Rizzo (12:03):
Um, you know, George, when you and I connected
for our listeners like.
we don't normally have founderscome on and talk about products.
Right.
Like as, as a sort of a rule forthe show.
That's something, something thatwe just, we don't generally do.
But the reason we wanted to talkto George was specifically on
this show was because thisconcept of warehouse native is
(12:25):
super new.
Like it, I, I would argue you'relike at sort of the cutting edge
of like this new set of adoptionthat's coming down the pike.
At least for me, when the firsttime I heard it, it was
definitely an aha moment, rightwhen I suddenly thought about,
wow, if there's an organizationthat's mature enough to try to
(12:47):
actually centralize theiroperations on a on snowflake or
big query on a data warehousescenario.
I have not worked for anyorganization that has done that
yet, However, for those that aredoing this, um, gosh, it does
make a lot of sense, uh, topotentially have these apps that
just sit right on top of that,um, and allow you to interact
(13:08):
with that data in a way that iseffectively is wholly owned.
Right?
Like, you know, to some degree.
My take on this in Georgia, I'dlove to hear your thoughts.
To some degree, Salesforce ownsyour.
HubSpot owns your data, right?
Like at the end of the day,they're sitting in their
infrastructure, their server'slike, yes, you own it, you own
(13:28):
the rights to it.
You can extract it and move itto somewhere else if you'd like.
Um, but it's certainly not yourown servers or anything like
that.
Um, you don't, you're notbuilding around your own sort of
infrastructure.
You're, you're, you'resubscribing to software as a
service.
Um, and.
What I'm seeing in the marketis, at least with stuff like
(13:49):
this, is this sort of shift backto, well, we actually,
especially with privacy andcompliance laws and all these
things coming together, weactually need to own our data
more holistically on our ownservers that, you know, maybe
have other software tools thathelp enable our interactions
with our data sets, uh, andpotentially the enterprise that
(14:10):
needs to now activate that dataand needs tools like a Super
Grain or something like.
To sort of get back into theflow of, you know, doing a go to
market motion.
Is that sort of like, am I, am Ioff?
Like is this like, is there aton of companies doing this and
I'm just not working at any ofthem and like, am I speaking
around in circles or
George Xing (14:30):
No, no, no.
You're, you're totally right.
And, and I, and I think you, youpointed out some of, a couple of
the key reasons that I thinkthat the warehouse native
approach, um, is compelling to anumber of people that we talk
to, which is, uh, pri you know,privacy and data, data
ownership, right?
So with a warehouse nativeapproach, instead of moving your
(14:53):
data into, um, another thirdparty, Uh, and copying it there
every single time.
You need to use it for emailautomation or, or CRM purposes.
You just kind of, it just sitson your own cloud, your own data
cloud.
And then you can imagine thesame way that you have an iPhone
(15:15):
and you can install applicationsthat run on your iPhone.
You can essentially have dataapplications that run on top of
the data that you already own.
Um, it's like if you have youriPhone, And you have all your
customer data there.
Instead of figuring out a way toplug in a USB stick or something
like that and transfer it toanother device, um, you just
(15:35):
have, uh, all these applicationsthat use the same underlying
data and they can also sharedata, uh, and communicate with
other apps that are alsoinstalled on your phone.
Um, that's maybe kind of theanalogy.
Um, and that's certainly a fewyears, probably down the road
before we get there, but that'scertainly the vision that.
(15:56):
I think a lot of this, uh, canpotentially get to, um, in a few
years once the technology isready, once the ecosystem
matures.
And it makes a lot of sense forthe end users because you don't
have to pay the cost of movingdata back and forth.
You don't have to worry aboutdata syncing issues or
inconsistency between differentapplications and different
(16:17):
platforms.
Um, and at the end of the day,you have a single source of
truth, uh, that, um, becomes alot easier.
Uh, activating and doingpersonalization and
segmentation, which, you know, Ithink a lot of folks probably
listen to this would agree.
One of the hardest parts of, um,you know, using product data and
(16:40):
other data that lives inside of,uh, a cloud data warehouse for
messaging purposes is actuallygetting the data into those
platforms.
So it also solves that problem.
Michael Hartmann (16:51):
Yeah.
So, um, I wanna make sure Iunderstood this right.
So it sounds like, is there, didI hear correctly?
There's a sort of a suppositionfor this.
Uh, warehouse native conceptthat the, the organizations that
are gonna be successful alreadyeither have or are in process of
(17:11):
building will keep it tocustomer a, a customer data
warehouse.
Is that, did I, I'munderstanding that right.
George Xing (17:21):
Yeah, I, I would see, I would say that one of the
things that we observe is thatonce companies get to a certain
stage, certainly the startupsthat we work with, um, when they
get to series A or series B, itbecomes part of the natural
evolution of how they thinkabout their data maturity.
So, um, around that time, theystart to invest in, uh, many
(17:42):
different systems.
They start realizing the painof, uh, having data in a bunch
of different places.
And there's a solution out ofthe box where now they will, um,
invest in a data warehouse,start centralizing all that
information there.
Also start building out datacapabilities to manage the data
that lives inside of thatwarehouse, and then start
(18:03):
thinking about how to leveragethat single source of truth in
downstream places.
And so that evolution happens alittle bit earlier for some
companies, a little bit laterfor other companies, but as a
whole, um, certainly becauseit's now easier than ever before
to set up a data warehouse.
It's happening, um, I would sayearlier than, than, uh, it was
(18:25):
as an industry, uh, even acouple years ago.
Michael Hartmann (18:29):
Interesting.
Yeah.
I mean my, my head went straightto like, I've, the idea of
simply defining terminology thateveryone agreed to is what is a
customer, right?
Seems like it should be an easything to define, but it's
usually like I've only been atone place that's even come close
to doing it.
Um, so do, I mean, do you runinto those challenges where
(18:50):
these companies, um, are stillstruggling with.
The, the taxonomy, so to speak,of what they're doing in terms
of that data.
So they have the, they have thislike, Hey, this need to do, they
realize it to to scale orwhatever, to grow the way they
need to.
They need to be more deliberateand conscious about how they
manage customer data.
(19:11):
Have they gone through, are theytypically still going through
that process of DEF definitionsor are they usually beyond that
at that point?
George Xing (19:20):
Yeah, I, I would say it's a spectrum, you know,
as with most things are, um,typically we see that, uh, the
first use case is that there's aparticular.
Piece of product data or someproduct metric that is really
important to the business thatthey wanna use in some type of
(19:41):
email campaign or for some typeof onboarding flow.
So to, to make this reallyconcrete, um, let's say they,
you know, a customer wants tosend, um, and kind of like a
upsell message to the admins of.
One of their customers everysingle time there's five, uh,
(20:02):
new users on that account.
Right?
Uh, when, when that accountreaches five total users, well
that calculation, like thenumber of active users on that
account is a calculation that isdone inside of the, the cloud
data warehouse.
So one of the first thingsthat'll see is they'll throw all
(20:23):
the data that they need intotheir cloud data.
W.
In order to calculate thatmetric.
And then the next step isgetting that metric into, you
know, the system of engagementfor doing marketing automation
to support that onboarding flowthat they want to, um, want to
do.
And, you know, I, I think likeover time they will expand and
(20:47):
calculate more metrics.
They will build more of like acohesive taxonomy around all of
this.
But it really just starts withthat first use.
And then spans, expands fromthere.
Mike Rizzo (20:59):
I love, I love that you bring that up.
Just earlier today I was on.
Call with a community member,and we were talking about, um,
how challenging that that exactproblem is like in marketing
operations.
Uh, we are frequently asked, Ithink, at least in the B2B SAS
space, right?
Like, how do you do anonboarding nurture?
(21:21):
How do you do an activation ofsome kind for all these users?
How do you do.
Feature use case usage basedsort of marketing activities.
And the answer is, is like it'sreally hard to do, like even
understanding whether or not afeature is currently in use.
Like a lot of, a lot of folksthat haven't ever had a chance
to talk to a product team or adeveloper, uh, they don't
(21:44):
realize that, like, that, thatisn't inherently built into the
code base to just send thatstuff around.
Right.
You know, the, the, theengineers are not sitting there
going, wow, the marketer's gonnawant to know when a user does
this, so I'm gonna make sure Iwrite a piece of code.
Like, that's not what they'redoing.
They're there to build reallygood product, you know, and, and
one of the reasons why I lovethe idea of pairing up, you
(22:06):
know, marketing apps andmarketers with, with, uh,
product teams, is to try to getto those questions a little
sooner rather than later.
But that's fundamentally, it'sreally hard to do.
And so, you know, I love thatyou're tackling.
With like sort of a layer of,uh, of technology that can like,
at least facilitate that right
George Xing (22:26):
Yeah.
Yeah, for sure.
And, and this is the exact kindof use case that we're trying
to, trying to go for because wesee it so frequently.
Um, one of the other things Ithink that, uh, I just thought
of as, as you were talking aboutkind of.
This interplay between theproduct side and what the
engineer's instrumenting andwhat the marketer's able to to
(22:48):
use in terms of the data is alsojust that I, I think some of the
lines that we see, at least, youknow, from the conversations
that we have between, uh,marketing, marketing ops and
growth teams, which often haveengineers and PMs, those start
to blur because who really ownsan onboarding flow really, you
(23:09):
know, Is it the pm, the growthpm who's kind of, uh, sending in
app notifications?
Um, or is it, you know, themarketing team that's kind of
creating this nurture series,uh, who are really hitting the
same customers with very similarmessaging just through different
(23:30):
channels?
Or is it, you know, the CS team?
Is, uh, kind of reaching out tosomebody who's going through a
free trial and trying to getthem to, uh, you know, debug
their data connection becausethey have trouble adding out
data source or getting reallyvalue outta the product.
And, and so one of the thingsthat we see a lot is, okay, you
(23:52):
have a marketer that's actuallysending out an email and signing
that as like a CS manager, uh,in the byline, or same thing
happens with sales.
Um, and, and a lot of theselines start blurring, even
organizationally.
Uh, you just have kind of oneperson sometimes that smaller
companies that's doing, doingeverything, and maybe kind of,
(24:15):
you know, one, one takeaway thatI have is I wonder if we're
headed for a future where thingsare much more aligned
organizationally around thecustomer journey rather than,
You know, strict handoffsbetween different parts of the
customer funnel because, um,certainly one of the things we
see with as more product databecomes, um, important in the
(24:40):
B2B marketing journey is that,uh, it's much more about
there's, there's no linear pathfrom or hand off between one
function to the other.
Uh, it's all about kind of thecustomer journey, which is very
business.
So I'm curious if you guys seethat in, in, um, in, in some of
(25:02):
the things that Yeah.
Mike Rizzo (25:04):
I definitely do.
And I, I think, I think thatthere's a chance for, you know,
we're seeing more of this like,um, you know, adoption of like,
what do we believe the customerjourney should be?
Um, and then how do we sort ofimplement that, right.
And I think like one of the corechallenges that I've come across
(25:29):
is, The, the realization thatlike there isn't, there's no
like, um, silver bullet thatlike solves all of the, the
problems, right?
Like we can't, there's no oneanswer to any of this thing
Michael Hartmann (25:43):
You wanna repeat that?
Cause I wanna make sureeverybody heard that.
Mike Rizzo (25:47):
right?
George Xing (25:47):
there's no silver
Mike Rizzo (25:48):
There's no silver bullet that, that solves this.
And George, I think for you,like it's a dangerous hole as a
product owner.
You're probably, you're gonnahear from investors and, and
people who are using yourproduct, they're gonna be like,
well, can you build AI to tellus what the user journey should
be like?
You know?
And like the thing is, is for
George Xing (26:09):
don't worry.
We've already gotten that.
Yeah.
Mike Rizzo (26:11):
see, not
Michael Hartmann (26:12):
imagine.
What's the build in the bestpractice?
Mike Rizzo (26:15):
Right.
Right.
I think, I think my As, assomeone who.
Is now a curator of like,programming that is community
led, right?
Like I am, I'm saying like, Hey,what, what does everybody really
want?
And how can we try to bring thatto the, to this community?
Um, I think if you can, to thebest of your ability, try to try
(26:38):
to work alongside your currentmarket, your customers, to
figure out what it is that they,what they believe their journey
should be and how they want toadopt your product.
But I think, don't be afraid to.
Create, like carve your own sortof path, right?
Um, imagine like an art gallery,right?
When someone puts an art gallerytogether, they actually want to
(27:00):
tell a story through thatjourney on your sort of trip
down, this like one particularartist's like timeline or maybe
a genre and a set of categor.
That person who put that gallerytogether did it with intention.
And the turns, the lefts and therights and the decisions that
you get to make are the storythat they want you to
(27:21):
experience.
And so I think it's like on usto also try to figure out like,
what do we believe the journeyshould be?
And like let's hedge our betsand try that for a little while,
and then take people down thatpath.
But to your point earlier,George, like the lines are
blurred and it very much dependson your business.
Like maybe CS people are tryingto do upsells through the PLG
(27:44):
motion.
Maybe sales people are doing it.
Maybe the product people aredoing it.
Maybe the growth team is doingit.
George Xing (27:50):
Mm-hmm.
Mike Rizzo (27:51):
But all of that, no silver bullet one and two.
You know what?
Just come up with the journeyYou want to take them on, like
curate the experience that youwant them to have and focus on
just that.
And don't divert like,
Michael Hartmann (28:05):
you gotta give it a chance.
Yeah, we've been, so I thinkit's interesting that you asked
that question, George, becauseit Yeah.
We've been doing some thinkingabout, um, re.
Rethinking our whole go tomarket, particularly in the
marketing side approach for ourbusiness to think around it in
terms of a, a marketing funnel.
(28:26):
And I have, honestly, I havemixed emotions about, or mixed
feelings about it because Ithink it, it, on the one hand it
does provide, uh, some amount ofstructure so that we're
thinking.
At least having common way ofthinking about how we're going
to market and so we can measureit and all that kinda stuff.
At the same time, I just don'tbelieve that's the way that
people buy and we we're not evenin a tech space.
Right?
I don't just don't think that'sthe way people buy even a B2B
(28:48):
space.
Right?
They, when they're, and I thinkof that for my own stuff, like
when I'm, when I have anopportunity to research
something that I'm just maybejust curious about, I'll spend a
fair amount of time doing thatand then I might not do
anything.
For weeks or months.
And then cuz I'm kind of stillthinking about it, right?
And it's not this very linearthing, it's very chaotic almost.
(29:09):
So, um, I do, I think that'spart of the challenge too, is
that, but I think it does matterwhere, where it does matter is
if you can put this informationin front of anybody who might
interact with that customer orpotential customer, regardless
of where they are in theorganization.
I think that's where you startto have the opportunity to
really build trust with thosepeople, which then builds
(29:31):
loyalty and, and advocacy if youget to that point.
But without that, it's reallyeasy, you know, if your customer
success or customer supportteams don't have visibility into
what's going on, marketing andsales, they can really kill a
deal, right?
Um, so, or a salesperson mightcome in at the wrong time when
there's a pro a product issue,right?
George Xing (29:53):
Yeah, yeah, yeah.
I mean, we hear this all thetime.
It's, it's, you know, People,people say, what's the classic
thing where they say, um, yourproduct reflects your org
structure, right?
Um, yeah.
And, uh, or you ship your orgstructure and, and I think this
is a case where, um, the toolsthat we use reflect kind of the
(30:18):
org structures that have beenkind of like set.
As best practice, right?
There's a CS team, there's asales team, there's a marketing
team.
So of course, all those teamshave their own dedicated tools
to manage customercommunications for their part of
the funnel.
But what happens when you knowthe funnel isn't quite so
(30:39):
linear?
And when those lines get blurreda little bit, well then those
tools still kind of then, thenthose tools basically start
overlapping and um, and itbecomes kind of a coordination
c.
we see this all the time.
Um, we talk to customers all thetime who are, you know, large at
scale companies that everyone'sheard of, um, and they still
(30:59):
struggle with, uh, this today.
And, you know, even basic thingslike subscriptions and, uh,
making sure that you don'treceive an email from another
system if you've unsubscribed,um, you know, from one.
Um, those things are, are.
(31:20):
Are, um, are hard to, to kind ofmanage, uh, which could be
surprising.
Um, I think at least, you know,as, as someone who doesn't come
from a marketing background, um,that was surprising kind of to
hear.
But it just kind of shows kindof maybe the complexity of some
of these challenges and alsojust the inherent nature and
(31:41):
evolution of some of the thingsthat we're seeing in the market.
Michael Hartmann (31:44):
Yeah, I, I think it's, it's really
interesting and, but I cantestify to the fact that there
are definitely companies wheremultiple systems are not in sync
when it comes to things like optouts.
So definitely think that's thecase.
So it, so what I wanna kind goback to a little bit, you, um,
(32:07):
When you and I first talked, youknow, and you mentioned it
already, right?
I, I immediately was thinkingabout CDPs and I think around
the time we talked, we actuallyhad a guest not long before that
on talking about CDPs.
Cause it was something that I'dbeen hearing about, didn't know
much about.
So how do you, like, what's thedistinction between the, the.
Kind of warehouse nativesolution concept and CDPs.
(32:30):
Is it, are they competingthings?
Are they compat?
Like are they complimentary?
Like what's the, what's thedifference there?
How do you see, see those?
George Xing (32:39):
Yeah, I mean it, I think it really comes down to
the source of truth distinctionthat I was talking about
earlier.
Um, what CDPs do is they ingest.
Customer data from a number ofdifferent source systems.
It could be your eventcollector, it could be third
party tool, it could be yourcrm.
And then they let you do, uh,they do identity resolution,
lets you do segmentation on topof it and then, you know, send
(33:02):
it downstream usually to othertools for ad targeting, emails,
et cetera.
And, you know, I, I.
The big kind of shift in thewarehouse native approach is,
you know, you don't need aseparate data store to do all
this organization andcentralization and identity
stitching.
You actually just want yourcloud data warehouse to be the
(33:25):
place where all that happens.
And then you're accepting the,the, the cloud data warehouse as
truth and, and so concretelywhat that means, You're putting
the onus or, or giving thecontrol of, um, the data
modeling exercise, the identityresolution, how you want to
(33:49):
calculate certain metrics to thecustomer rather than being
opinionated about, okay, this iswhat an active user is, or, you
know, this is exactly how you,you can choose between these two
options for identity resolutionand, and because every business
is the.
That is, is not the same.
Rather, um, that means that, uh,you're just gonna be able to
(34:13):
have a lot more flexibility andend up with data models,
metrics, um, identity, uh, thatis tailored to your business,
uh, versus something that comesout more out of box that you
don't have much control over.
Um, that's, that's kind of likemaybe just from a.
(34:34):
Usability standpoint.
And then the other thing I wouldsay is that kind of goes hand in
hand is that CDPs typically,again from from our
conversations, take a long timeto implement because you're
migrating a lot of data, you'redoing all this crunching, um, in
(34:55):
like a separate system.
If you can do all thatcalculation computation inside a
cloud data warehouse.
The customer is alreadyoperating, then it simplifies
the process to implement fromsometimes months to days.
And you know, obviously that's ahuge, uh, difference when it
comes to business results andgetting up and running.
Michael Hartmann (35:18):
Okay, so let me, I wanna see if I can play
this back cuz I think I,something just hit me.
So I think what I'm hearing, soCDP based on, not what I
understand, what you described,it's sort of pulling in data and
then pushing it back out.
Or some version of it back outto these different systems.
They all have their owndatabases and data structures.
(35:40):
Um, and maybe it's doing somecalculations, but maybe those
applications are doing'em aswell for their specific needs.
Is that, uh, do I have thatabout right?
For cdp?
George Xing (35:52):
Yeah.
Michael Hartmann (35:52):
Okay.
All right.
But, and so I think what I'mhearing differently with this
warehouse native is that ratherthan, um, Sort of ingesting,
pulling data in from all thesedifferent apps or, or, or
solution platforms?
The, the database is thedatabase and those, these, if
(36:13):
you have warehouse nativeplatforms, they're basically
just apps that are accessing,um, and maybe sending some data
back, but it's based on probablytransactional level kind of
stuff as opposed tocomputational kinds of things.
Is that the way, am Iunderstanding that right?
George Xing (36:32):
Um, yeah, I, I would say the, the, the CDP is
requiring kind of, it's tryingto be the source of truth and,
and in the absence of a clouddata warehouse, you basically
have to do everything that acloud data warehouse does and
then more, and whereas, In awarehouse native approach is
you're, you're saying, okay, youhave a data warehouse, so I'm
(36:54):
not gonna do all the backendprocessing, I'm not gonna do all
the computation, the ingestion,and everything I would have to
do as a cdp.
Um, I'll just do kind of the,almost the UI layer or the
orchestration layer, so thesegmentation, the business logic
and the actual activation.
And so it's kind of decouplingthe, the database part from.
(37:19):
The interaction and the UI andthe application side.
A CDP kind of combines the twoand one because in a world where
you didn't have, you know, clouddata warehouses, you had to do
both.
Michael Hartmann (37:30):
Yeah.
Well, and so reason I was tryingto clarify is where my head went
to, cuz as I've already likeshared, right?
I'm old enough to remember backwhen it was like mainframes and
terminals and client server wasa thing like this sounds very
much like that.
I mean, it's.
That's kind of was, it's notquite the same, uh, but
certainly, definitely notmainframe terminal, but the
(37:52):
client server.
That concept of where you'resort of splitting some of the,
the, um, processing effortsright between different sides
of.
The, the, it sounds somethinglike that, which, so it's
interesting to me to see kindathe evolution has been to go to
all these sort of, um, specificapplications with their own
(38:14):
databases and their purposes toone that's more like kind of
going back to old school withnewer technology to support the
ability to distribute the, thedata and all that.
So, To me that's interesting,right?
Like I think I finally, thathelped me click on how to
differentiate this from thingslike cdp.
So tell me if I'm I'll totallyoff here.
George Xing (38:36):
yeah.
No, no.
I think, no, I think that'sexactly right.
I mean, you, you need, um, youneed one database at the, at the
end of the day, right?
And.
Your customer data should bestored in, in one place.
You don't need multiple copiesof it.
If you can have everythingpointed to that single copy and,
um, you know, in a world whereyou have a cloud data warehouse
(39:00):
that is your copy of the data,uh, you don't need a CDP plus
another tool, plus another toolthat all have separate copies of
your data.
Um, and uh, and, and I think,you know, that's why.
You know, you're starting to seea lot of companies move from,
you know, their CDPs over tokind of, uh, adopting cloud data
(39:22):
warehouses,
Mike Rizzo (39:26):
Makes a ton of sense.
I'm just like, I don't know,heart man.
You might have one more questionand, and so we could go with
yours.
Michael Hartmann (39:34):
I'm, I'm, I'm already realizing we, like, we
could have, we, we may need tofollow up on this one.
Mike Rizzo (39:39):
right.
Um, my, my question is, is like,you know, do you feel.
Who's responsible for this,like, transformation around how
to think about the activation ofthis data and like, is it, is it
a collective responsibility?
Like, like, you know, peoplejust need to get in a room and
(40:02):
talk and like, and like episodeslike this can help educate them
on this is why we need to talkabout it.
Or is it it like, I don't, who'sowning
Michael Hartmann (40:12):
Is it?
Is it a business operations?
Is it, yeah.
No, I think that's actually areally, really good question.
So I'm curious to hear whatyou're seeing out there, George.
George Xing (40:21):
Yeah, I like the short answer is where we think
about this question a lot.
Um, cause in our go to market,it obviously kind of ha has a
big impact on.
And where, you know, who we goafter and who we talk to, um,
and how we position ourselves.
I think one thing that, whatI'll say is, and I don't think I
(40:42):
have like a silver bullet answerhere either, but I, you know, I
can share some things that weobserve.
Um, one is that I think, uh, youknow, obviously the adoption of
cloud, cloud data warehousesmakes it easier for companies
earlier.
To get their data into a bettershape.
So what I mean by that is, uh,if you talk to me two years ago,
(41:07):
right, the average series Bcompany would probably not have
their data in a good place.
Like, um, they might not evenhave a data team.
Whereas today, the averageseries a company that we talk.
Has a cloud data warehouse andhas at least one person who's
managing it.
(41:28):
So just kind of the shift ofdata maturity, um, you
Mike Rizzo (41:33):
that's awesome.
George Xing (41:34):
Yeah, yeah.
And, and what that means isthat, you know, um, the people
who really benefit from this arethe downstream stakeholders.
So, uh, marketing teams go tomarket teams that consume.
I think they're just gonna haveaccess to higher quality data
faster in the life cycles.
So you're gonna start to seeearlier stage startups have more
(41:55):
sophisticated life cycle andmarketing programs.
I think that's, that's thedirection that we're headed and
that reflects a need generallyin the market where companies
are collecting more and moreproduct data earlier, earlier in
their life cycles and doing morepersonalization and targeting
and workflows based on thatproduct data.
Um, I think on the other, talkto marketers.
(42:17):
Um, you know, we also kind ofsee a similar type of
convergence where, um, you know,I think there's this stereotype,
uh, at least, um, and I thinkit's kind of wrongly placed
that, you know, marketers arenot technical or data savvy.
(42:37):
I think it's actually theopposite.
Um, I think marketers areactually very data savvy and we
start.
Marketers are actually quitetechnical, very proficient, able
to speak about, um, technicalconcepts and understand the way
that their data is structuredand interface very directly with
their data team and say, no,this is, I need the data in this
(42:58):
format.
Um, I need you to get me thisdata, point a certain way, talk
to their product teams, talk toengineers, and be able to ask
for what they need in a veryspecific way.
That wasn't the case.
Three, five years ago when I wasworking, you know, at Lift.
And, and I think what that meansis there's just a much tighter
(43:21):
collaboration between thoseteams, and I think it means that
they're able to get what theyneed faster.
I think there's just betterdialogue and better
collaboration.
Mike Rizzo (43:32):
Mm-hmm.
George Xing (43:32):
Um, so those are two things that I see from my
kind of like selfish standpoint.
I hope that this trend acceler.
Because it'll make our job a loteasier in terms of educating the
respective people, um, gettingdata people to appreciate the
go-to-market challenges and the,and the business use cases, but
(43:53):
also getting kind of the, thego-to-market teams to appreciate
the technology considerations.
Um, that's a lot of where we sitand trying to kind of pull those
two teams together and get themto talk to each other.
Michael Hartmann (44:07):
Yeah,
Mike Rizzo (44:08):
I really like that, um, George, and I appreciate
your answer.
Um, hopefully people tune intothis episode and we can
accelerate that, that learningas well, uh, with you.
So thank
Michael Hartmann (44:18):
Well, I, I think what was really, really
caught my attention when wefirst started talking about this
was just, I think in the back ofmy mind and probably said it out
loud and public too a few times.
Like it feels like there's been,even with all the volume of new.
Marketing, revenue technologycompanies out there.
(44:39):
They seem to be really niche.
There's not really beensomething that seemed really
sort of a, a a, I hate to usethis paradigm shift, but like a
significant change in direction,um, like this.
So that, um, it's, it's reallyinteresting and I, I'm gonna be
really curious to see how itplays out.
(44:59):
So, George, thanks for joiningus.
Anything, anything last minute.
And truly, like I have a wholeset of new questions that
weren't even like, we didn't getto everything we had planned on
even here.
So, um, anything like last bitof nuggets you wanna share with
our listeners before we, we, wesign off here.
George Xing (45:20):
Um, I, you know, I think that the, the one thing
that I, I'll say that I learnedis, uh, or, or, you know,
learned from working with ourcustomers is that, um, Almost
universally, you tend to kind ofunderestimate the benefits of,
um, being able to use betterdata to run the same campaigns.
(45:43):
It's like, you know, we talk topeople that say, okay, yeah,
like we have data that we're notusing product data that we're
not using, but we haven't reallyprioritized improving this
onboarding flow because, youknow, we got like 50 other
things that we're working on.
And I, and I say, okay, wellyou.
We can run a really likelightweight, simple poc.
(46:05):
It'll take like a couple hoursto get started and then, you
know, at that point, if you cantell, if you see no results or
uh, then no harm, no foul.
Right?
But why not give it a try?
And inevitably, you know, Ithink people under estimate, and
it's not because we have amagical tool or anything, right?
(46:25):
It's just because people alreadyhave data that they're not
leveraging and.
The impact of personalizing thesame emails with just a little
bit more, or, um, being able todo a little bit of more routing
logic to send a specific segmentof users different messages or
messages that they wouldn't havereceived before, um, will drive
(46:49):
conversion at like a verycritical step of their
activation funnel.
I think especially for foundersand, and people that haven't run
a lot of these programs before,your intuition is that, hey, it
might not matter that much inthe beginning, but you know, uh,
it turns out that in generalpeople underestimate the impact
(47:10):
that it could have.
So, um, I would encourage peopleto give it a shot.
Michael Hartmann (47:15):
no, I think, I mean that's what's something
I've been preaching for a whileis to focus not on like these
big, easy to measure conversionfor things, but looking at.
Micro conversions through acustomer's journey and, and
really focusing on improvingthose incrementally and they
have a multiplier effect, right?
Can really make a difference.
So, fascinating stuff, George.
Thanks for, thanks for joiningus today.
(47:36):
Um, if folks want to keep upwith you, connect with you or
learn more about Super Grain,what's the best way for them to
do that?
George Xing (47:44):
You can go to super grand.com, which is our website.
Uh, also feel free to shoot me anote at George Super Grand dot.
Always, uh, excited to haveconversations like this, chat,
marketing, uh, chat data.
So, um, don't hesitate to reachout.
Thank you.
Michael Hartmann (48:00):
Yeah.
Great.
George has been a pleasure.
I my mind is really now, Mike,thank you as always
Mike Rizzo (48:07):
Thank you, George.
Michael Hartmann (48:08):
we'll get, Naomi on next time as well,
hopefully, and to all you outthere listening and, uh,
continuing to support us, wethank you and continue to give
us your feedback and support andideas and suggestions for topics
and or.
Guess so.
Until next time, we'll talk toyou later.
Thanks everyone.
Bye.
Mike Rizzo (48:26):
Bye.
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
24/7 News: The Latest
The latest news in 4 minutes updated every hour, every day.
Therapy Gecko
An unlicensed lizard psychologist travels the universe talking to strangers about absolutely nothing. TO CALL THE GECKO: follow me on https://www.twitch.tv/lyleforever to get a notification for when I am taking calls. I am usually live Mondays, Wednesdays, and Fridays but lately a lot of other times too. I am a gecko.