The Product Intuition Behind $800M+ In Exits w/ Kamakshi Sivaramakrishnan

Kamakshi Sivaramakrishnan shares her journey from AdMob to founding Drawbridge and her subsequent transition to LinkedIn. She discusses the mobile advertising revolution, the success of AdMob during the iPhone launch, and its acquisition by Google. Kamakshi highlights Drawbridge's unique value proposition and the challenges of navigating ad tech regulations. She offers insights on customer relationships, board management, and acquisition processes. Kamakshi also delves into her transition to LinkedIn, founding Samooha, and the complexities of data collaboration. The episode concludes with advice for founders, especially female entrepreneurs, and the importance of empowering product teams.

Episode Transcript

(00:03):
So I'm delighted today to welcome KamachiShivaratamakrishnan.
Thank you.
And today we're going to talk about her journeyin Silicon Valley through a couple of different
chapters from joining and founding a number ofstartups, and then also kind of the exits
around this.
So welcome to the podcast.
Thank you so much for having me.

(00:24):
Thank you, Peter.
So I'd love to start at the the first companyworked at, AdMob, which just a fascinating
story.
So you joined them around 02/2007, I believe.
And so, I mean, this was fascinating rocketship from my vantage point.

(00:45):
So I got to know Omar a little bit, the CEO.
We were both Sequoia portfolio companies aroundthe same time, and he went to kind of zero to
100 Yes, or seven indeed seven Very, veryquickly.
So I'd love to share a little bit about how yougot that opportunity and kind of how was that
experience for you?
For sure.

(01:07):
My story with AdMob is kind of an interestingone.
I got my PhD.
I'll start maybe a step earlier.
Right.
I got my PhD in a field called informationtheory at Stanford.
So information theory is actually pretty uniqueeven among kind of the engineering community in
the sense that it is basically mathematics forelectrical engineers.

(01:29):
So it's kind of a World War two time era kindof field that actually emerged in that time by
kind of a legendary scientific figure calledClaude Shannon.
It was primarily done for kind of informationencryption coding during the war time.
And then basically, it's a mathematicalframework by which you process information,

(01:53):
quantized information, anything that isrepresented, for example, binary value data.
And that has evolved over the course of theyears.
A lot of the hard problems have been solvedover the last sort of seventy years or so.
So what remains are the very, very hardproblems that are yet they're all kind of very
math in the in the realm of, like, you know,mathematical optimality of said algorithms that

(02:15):
processes information optimally with this kindof performance results.
That's roughly kind of the framework ofinformation theory.
So I got my PhD in a field of by solving someproblems on denoising of continuous valued
data.
So any ambient data that you acquire is goingto be noisy.
How do you kind of denoise the data, extractvalue from the signal, and you know, you're

(02:41):
able to process it for said use caseapplication.
So how do you optimally denoise data?
Now think about kind of it's called universaldenoising techniques of continuous valued data.
Nothing of that says anything about AdMob andkind of mobile advertising and marketing.
But I think this is kind of what the legend ofAdMob is in my mind.

(03:04):
What my PhD prepared me is basically think fromthe ground up.
Don't be afraid of like, you know, black sheetsof paper.
You can think from first principles up anddevelop algorithms, design them, prove
optimality, implement them, and potentiallybuild product and bring them to market.

(03:25):
I didn't do the build product.
I didn't have any idea of bringing them tomarket, etcetera.
I stopped at algorithm design and provingabnormality.
I literally ran into Omar on campus.
He said, you should come talk to us.
And we were like, I don't know, was less than20 people at MOB at the time.
It was a dingy little office in San Mateo, likeatop a number of sushi restaurants.

(03:49):
San Mateo is known for great sushi in the BayArea.
So and so I went in one of the days because heasked me to come over.
And Mark here at this time, I had a number ofacademic offers and the who's who.
At the time, Lehman Brothers was still around,Lehman Brothers, Morgan Stanley, Quants,
because a lot of information theories went intoquantitative finance.

(04:10):
And you were just finishing your
My PhD at this time.
Time, okay, so you were thinking
about So I was what's thinking about what next.
I didn't have anything that was like amarketable idea myself.
Even though you're at Stanford, you cannot missentrepreneurship.
It's everywhere around you.
But I come from such a theoretical field thatthere is nothing marketable or product oriented
thinking.
I'm just like a theoretical kind of, you know,algorithm designer.

(04:32):
And I see the story with sort of one ulteriorpurpose, which I'll come to in a second, that
there is I don't know.
Maybe this resonates with you too, Pete, isthat there is no kind of set recipe for the
founder journey.
Right?
Each one of us has our own journey.
Yeah.
And this journey of mine sparked with, like,this accidental meeting with Omar, him asking
me to come over to the Admab office.

(04:53):
I go into San Mateo.
As I said, it was like this little dingy officeatop a bunch of sushi restaurants.
I went in the morning.
It it can be a bit, like, you know, smellybecause there's a lot of, like, fish cleaning
from the day before.
And I was like, it's a very differentexperience coming from, like, Stanford.

(05:14):
And, like, nevertheless, there was a lure thatwas unexplainable to me even though the red
carpet of Goldman Sachs, Lehman, Citi, MorganStanley, all the kind of university campuses
and tenure track faculty positions completelylike glitzy compared to There's this a lure and
I agreed to give it a try.
It took me some time from that point toactually convert it, to actually going and

(05:37):
being a part of the team.
And what was the, you know, Omar, and you musthave seen that actually this is, you know, the
traditional kind of web desktop basedadvertising systems were getting established at
that point in terms of just the tracking andmatching.
And what was it in what was it in the mobilespace and the data problem that you would that

(06:01):
was sort of intriguing that you needed someoneof your kind of academic and algorithmic
background?
See, I think at the time when they had start,when AdMob had just started, it was basically
trying to create kind of a mobile ad supportedeconomy.
It was not present at the time.
Even if present This
is before the iPhone law.

(06:22):
Yeah, this is before iPhone.
And it was basically kind of, I would say, alittle bit of like ghetto world out there.
There were a few ways to kind of, you know,make market products, ringtones, and
wallpapers.
Yeah.
This is what it was at the time in terms ofmobile marketing.
Mobile advertising was still extremely at itsinfancy.

(06:44):
So AdMob's story was not simply kind of tryingto create a comparable marketplace to at that
time, the kind of the real marketplace wasAdSense, AdWords.
It was very Google dominated, to your point,kind of very desktop forward to kind of
replicate that economy in this fast emergingmobile kind of world.

(07:05):
And I think the timing between AdMob's kind offounding and the launch of the iOS kind of the
iPhones and more broadly speaking iOS class ofdevices soon after basically created an
ecosystem that kind of while standardizing kindof the the the operating system and being able

(07:30):
to create kind of marketplaces, it also fasttracked what does it mean to be to advertise
market in mobile.
Yeah.
And AdMob was at the perfect moment in time.
I just saw this article of this morning aboutkinda NVIDIA being in kinda it's a story of
contradictions.
It's wild luck and great strategy.
Oh, yeah.
I felt like a version of that was true forAdMob as well.

(07:52):
It's wild luck because perfect timing withApple's iPhone launch and great strategy in the
sense of, like, you know, it reminds me of myfirst boss, Kevin Scott, at AdMob telling me,
like, the the sexiest startups or the sexiestideas are the ones that you get started.
Nobody even knows about it.
You're prepared so that when kind of the tidestrikes, you're prepared.
I felt like AdMob had that.
So when Apple struck, we launched.

(08:15):
It's just like investing and stuff.
You have to back companies that are nonconsensus before they become consensus.
Exactly.
It's like today, mobile advertising isobviously and just three years later was very
consensus.
Exactly.
But coming to your question about what was kindof the data problem is that now you have to
predict, say, search, search had an explicitintent where you're going into the toolbar and

(08:38):
saying, I, as a consumer, I'm looking for this.
So the keyword matching to kind of ad supportedhad an explicit signal of intent.
Whereas when you think about contextualadvertising, where it's not search based, that
intent signal isn't there explicitly, Thatmeans you're basically in the web world, it's
very content rich.

(08:58):
So you're able to kind of understand thecontext of the page and do some contextual
matching.
And then there was this whole kind of evolutionto understanding the behavioral intent of a
user based on some of their activity in theweb, etcetera, and the point of tracking,
etcetera.
And that was kind of the web evolution.
But in mobile, problem is very hard becausethat many signals are simply unavailable.

(09:22):
It was a very sparse kind of signalenvironment.
So as a result of which algorithm design ofbeing able to understand monetizability of
advertising in the presence of sparse data isthe algorithm design problem.
So that's kind of intersection between some ofwhat my background and information theory could

(09:43):
be valuable to kind of thinking the problemfrom scratch and not simply saying that we're
gonna take Google's logistic regression thatwas done for search and contextual advertising
and kind of replicate it in mobile because thatis not going to work as well because the signal
profile is quite different.
And so Google acquired AdMob just like a coupleof years after launching?

(10:04):
Yeah.
Was a couple of years.
2009 was when the acquisition actually 2010 waswhen the acquisition was announced.
And then we were stuck in FTC approval, DOJapproval for almost seven, eight months.
I mean, speaks to the fortitude of a founder.
Have such great appreciation for those peoplenow and maybe you of course will have your

(10:25):
perspectives as well.
Eight months of being kind of arguably in kindof this no man's land if you may.
It's a very public kind of, you know, event ofkind of this DOJ approval and cycle.
The company needs to operate as though it werean independent company, but this kind of

(10:47):
overhang exists.
So I have this deep for
I went through exactly the same thing.
Think nine months between the TrullianZillowdeal was announced and it closed and tons of
depositions with the FTC and kind of all thatstuff, and so as a public company, so I kind
of, yeah, I deeply appreciate that you're inthis sort of like period of uncertainty.

(11:09):
You just need to kind of keep your heads down.
I don't know.
I know I'm I don't mean to kind of turn thetables on you, but it's kind of like, you know,
a dichotomy in your mind, right?
You have to kind of almost have two parts toyour brain and kind of really function as
towards the deal, but also kind of operate asthough the deal were not to happen.
I don't know.

(11:29):
I'd love to hear any lesson that you feel likeyou learned at that time.
Oh, I think it's really challenging.
I think you kind of, as you say, you have tokind of compartmentalize your kind of mind to
say, look, there's a, there's a, there's a, Idon't know, and this was interesting actually
with a TruliaZilla deal.
At the time we announced the deal, the lawyerssaid, oh, 90% approval rating.

(11:53):
And then as things progressed, sort of like,well,
kind
of like seventythirty, then sixtyforty, andlike And then it went up again to kind of like
95%.
Emotional roller coaster.
Emotional roller coaster.
Like, there was, there were like a lot ofthings that kind of, you know, lots of people
and it's, you know, it's part legal, parteconomic and part political.

(12:18):
And so, but, yeah, it's hard.
I think that's just, I kind of think that's thefounder journey, right?
You kind of, you've got this, like, 17 thingsgoing on at the time.
Absolutely.
There's kind of fires over here, there's firesover here, and you just need to like, okay,
block some of that stuff out and just focus inon kind of what you need to do.
And there are things you control and there'sthings that you don't control.

(12:39):
When you're operating a business, you don'tcontrol what the FTC does, but you kind of can
do control kind of your business, your metrics,your customers, your product, all that sort of
stuff.
But, yeah, I'm sure.
I'm sure that was and and how many with withAdMob, how many people were there at the point
of the acquisition?

(13:00):
Probably about two fifty people.
Okay.
So you went from 20 to two fifty.
Two fifty people by the time This we were wasback in 2010 when the deal closed.
Admob kind of, I think it's mobsters is thekind of like the collective noun for kind of
people that work there.

(13:21):
What's the and a bunch of mobsters went on todo a bunch of different things.
You mentioned the CTO of Kevin.
Like, what was it in the DNA, do you think,that kind of created a sort of despair of
people that went off and do interestingentrepreneurial things?
I think it's a great question.

(13:41):
When I think back among all the companies thateither I founded or I've been a part of, I
think AdMob certainly stands out as one of thatkind of unique kind of moments in time.
The reason I call that not a company but amoment in time is because it brought together
these people who had probably dysfunctional insome ways or rather misfits, dysfunctional,

(14:09):
misfits in some ways, but yet extremelycreative, deeply technical, very good at what
they do and very driven and passionate.
And maybe it reminds me of one of the pointsyou raised earlier is kind of AdMob had such a
blitzing, like start to finish story as wellthat when all these people who are brought

(14:36):
together with such great characteristics thenwent on to do greater things because there was
that remnant energy in them as well, if I may.
And I think I would say that it was a lot aboutthe people.
I think while the cultural element from Omarwas strong, I think there was a very self

(14:57):
directedness among all the people that were inthere as well.
Literally each one of them, the ex CTO ofInstacart, Mark Schaff, CTO of Microsoft, Kevin
Scott.
There's so many founders who have had their ownkind of successful companies as well, including

(15:19):
myself.
All of them, all of us would even kind of just,there was this constant environment of like
iterating and iterating fast.
This was back in 02/1989, right?
This move fast, break things wasn't there atthe time, but we kind of had this kind of get
shit done kind of mentality.
But

(15:39):
I think in those in those high growth startups,people, you know, you see this rapid growth.
You have to have high agency in terms of justlike, okay.
And and the and the faster growing startupsoften, you know, have a clear vision, very high
bar for hiring, but also high autonomy forpeople that say, look, this is the problem

(16:00):
we're trying to solve.
I don't know how to do it as a CEO, but, like,you need to figure it out.
And we're gonna do it in two weeks.
I think this is such an amazing thing.
And that reminds me of kind of my story ofKevin Scott, right, who led all of engineering
at AdMob.
He, to your exactly to your point, hire smartpeople and get out of their way.

(16:23):
I think he embodied that and said some verystrong ways.
And I think to the kind of people like myself,self directed, a lot of agency, It allowed for
a lot of creativity to foster.
And I think disagreement and kind of differentperspectives and different points of view

(16:47):
weren't discouraged in any way, shape or form.
And that I felt was, and in some ways to bealso honest, right, like the longer the company
is independent, a lot of kind of evolutionhappens.
Things go wrong, you know, dynamics change andevolve.
AdMob was also fortunate that it was almostperfect.
It was a little over four years, less than fiveyears.

(17:10):
And so some of the kind of evolution and kindof atrophy deterioration that just happens.
We could avoid that as well.
Yeah.
People weren't trying to kind of be territorialor like some VP trying to get encroach another
because everything was growing so fast.
There's enough opportunity, right?
There's so much opportunity, no fiefdoms.
Yeah, so as a result of that, I think thatenvironment and incredibly smart people was

(17:34):
kind of AdMob's USP.
And I would say that kind of the legend of thecompany is the people it brought together and
the journey that they have since grafted on andAdMob is kind of the early part of that journey
for them.
So I think that's kind of Omar's, in myopinion, kind of true legacy even beyond the

(17:55):
company AdMob, because that was a moment intime.
This kind of continues to linger in the journeythat all of us have continued to have.
What was the feeling with inside the company atthe, you know, when you heard about the
announcement?
Was it, was it, because I imagine these thingsare like, you know, bingo, job done, or like,
really?
We're having so much fun?

(18:16):
We're kind of like changing the world.
We're just getting started.
What was the psychology internally?
I think exactly to your point, it was somewhatmixed with the sense of like, I wish we had
more, like we can do more.
This should be should be we could have gonnastay independent and done more.

(18:38):
But at the same time, $750,000,000 in 2009 is Imean, you the forward multiple of that is a big
number at that time.
It was I think it was the third largestacquisition Google had made at the time.
DoubleClick, YouTube, and AdMob.
So
And it was 2009 where you're like, this is thisin the depths of the financial recession.

(19:04):
Exactly.
Which was like there was no there's noactivity.
Valuations were really down.
Exactly.
So I think there was that, because it also hadexactly to that previous point of it brought
together these people, all of us with like somuch agency and so much self directedness,
there was kind of a desire to still own yourown destiny and we have so much more to kind of

(19:25):
achieve and we can go do.
But there was also a recognition of the factthat exactly to your point, given kind of the
macro environment at the time and just thesheer fact that this is a significant outcome
for the company.
There was an acceptance of that as well.
So that's why I say kind of mixed if you may.
I'm reminded of Omar actually when he kind of,you know, announced it to the team.

(19:45):
His wife was standing at the back of the kindof the team meeting and Omar's story is a very
interesting one.
He's had, like, couple failures, and then thesource gonna AdMob was the big success.
And he dedicated it to his wife.
So I I remember that.
Like, you know, I don't know.
It's been fifteen, eight, sixteen years since.
I still kinda remember that.

(20:05):
So it's that's where kind of the founder storyexactly to your earlier point is kind of this
there's kind of the human element of, like, howwe all go through kind of the the infinite
elasticity that we require so to go throughfailures and then achieve a success and do
this.
I remember a similar thing when we announcedthe Trulia Zillow merger.
There were there were a bunch of people werelike, oh, I get so much more many tools and

(20:31):
data to play with or more customers.
They were like a group of people like, okay,the scope of my opportunity just doubled or
tripled as a function of and there were a bunchof people that felt also that the scope of
their opportunity diminished as a function ofthat.
And I'd imagine the same thing.
It's like if you're if you're kind of anengineer or your job throws off data, Google

(20:55):
has just got more capacity, more compute, moredata than anyone else at that time.
And so that was
was exciting too.
Exciting too.
Absolutely.
And I would say that on the whole, I think itwas one of the best decisions I think maybe
Omar made for the company at the time.
It was a great decision for a lot of kind offolks in the team.

(21:21):
Had stayed had AdMob been independent, would ithave had an outcome as a public company?
Maybe.
But I don't know.
We've also seen through times where privateacquisitions have far outperformed the success
of public companies as well.
So I think we're all, it's a great moment ofsuccess.
So you joined Google for a little bit after theacquisition and

(21:43):
I then joined Google.
I was there for about a hot seven months.
You know, it was a bunch of us.
I think it was Omar who left first, then me andthen maybe a few weeks later, Kevin Scott, who
then went on to be at LinkedIn and the CTO ofMicrosoft and like systematically a lot of
folks thereafter.

(22:04):
It goes back to, again, to that point of this,when you bring such self driven high agency
kind of founder minded people together, andthere is kind of this very early exit that
remnant energy has to be expended.
And that's kind of where the founder journeycontinues into other companies that respond.

(22:25):
And for mine is an example of that, right?
AdMob story had just begun.
The mobile revolution, we were at the veryearly sort of part of it.
2007 is when iOS and iPhone were first Thereleased a app economy was just kind of
burgeoning as we speak.
Mobile monetization, mobile advertising, mobilemarketing had merely but begun.

(22:50):
So many kind of different challenges.
When we think about kind of time and kind ofthe rise of Facebook at the time, mobile was
transformative for the rise of Facebook.
And that happened starting 2011.
So you can imagine 02/2009, this was stillearly.
So for me, I felt like from a data algorithmicperspective, while Google was incredible in

(23:12):
terms of the amount of data available, howtransformative AdMob could be inside of Google,
there's also kind of the mechanics ofacquisition.
You have to integrate it.
There are kind of, you know, between producttechnology business, a number of things that
AdMob had to navigate.
And some of us were were operating with aninfinite sense of urgency.

(23:33):
So we wanted kind of momentum and speed.
We were used to that at AdMob.
So that's why I left to start this companyDrawbridge because it was so evident at the
time that with the rise of mobile, we aretalking not about kind of the single device of
consumption as consumers, we are touching thesemultiple devices.
IPads had broken into the scene, but it's thesame consumer who is across these different

(23:57):
devices.
And when you think about kind of these largeconsumer environments like Google or Facebook
at the time, Amazon less so, they have thenotion of identity established because you're
authenticating yourself into Gmail and toGoogle Maps or whatever be it and certainly
Facebook.
So locked in app experience.
That's not generally true.

(24:21):
Particularly in the advertising markets.
Particularly in the advertising market.
So the value of mobile was still, I would sayit was undervalued.
Mobile was still undervalued.
It was premium in terms of it's a newenvironment to engage from a kind of brand
advertising monetizability potential.
But the impact it driven, it drives was notfully measured.

(24:43):
And that was behind understanding what doesthat same consumer, how do they interact across
these different devices?
What is their action?
Whether it is online, whether it is in thephysical world, in the digital world, across
these different platforms and devices.
And this was the direction that AdMob was kindof moving into.
And that's kind of the direction with which Ikind of, I took off and said that that story

(25:05):
needed to be completed, but we needed to solvethis identity problem.
And it's And that and that makes sense as anindependent company as well because you Google
obviously wants to understand users acrossdifferent devices, but
Within the Google ecosystem.
Exactly.
Just within the Google ecosystem, whereas allthe other advertisers and all the other
publishers kind of want to understand that kindof that engagement.

(25:27):
So when you think about kind of the like kindof the world versus the open web, if you may,
right?
Like Google and Meta can solve it inside ofkind of their ecosystems, And they do maintain
a significant dominance because of that.
When you think about kind of the open web,you're essentially out of luck whether you're
on the kind of the brand marketing advertisingside or whether you're in the content app

(25:49):
developer or on the media side.
You're out of luck because you've not had thiskind of logged in authenticated experience that
consumers have adopted at scale and that'swhere engagement happens.
So identity was a problem to solve for.
So I left
and I You probably felt that you were uniquelypositioned to solve this problem.

(26:10):
Because of, again, kind of my background indata and the kind of solving for identities
essentially basically, again, going back tokind of this problem of you're solving for
algorithm design in the presence of sparsedata.
So you have to basically kinda do patternrecognition, which is essentially machine
learning.
And you're basically doing machine learning ongraphs because you're talking about

(26:33):
connectivity.
You have devices connected to consumers,consumers connected to each other, those
consumers, connected consumers having their owndevice activity, devices connected with each
other.
Setting aside some of the kind ofconsiderations around privacy, etcetera.
If you think about it, kind of justmathematically speaking, it's a very
interesting problem, right?

(26:53):
You're talking about kind of a highly connecteddense graph and inferring connectivity of kind
of consumers, devices, etcetera.
And this is a machine learning problem at scaleand solving for that in the presence of
extremely sparse data.
So to your point, was kinda, you know, uniquelypositioned to solve for it.
And I'm sure you might have had this experienceat Trulia as well.

(27:16):
After success, you're going to spawn off thesekind of entrepreneurially minded folks.
And the successful outcome also gives kind ofthis little pixie dust.
So I was able to raise capital pretty quicklywhen I decided that I wanted to leave Google
and kind of go solve this for the open web ifyou may.
And that's where I raised money, a seed capitalfrom Sequoia and Kleiner Perkins started this

(27:38):
company Drawbridge.
Hard problem to solve.
It's really not easy.
The data acquisition, it in some ways isbasically it reminds me of like the challenges
OpenAI would have gone through.
It's a different world today, obviously, thankind of work.
But it's data acquisition, whether it's machinelearning or AI.
A lot of it is about the data acquisition to beable to train models.

(28:01):
And would integrate into the ad servingtechnology?
Is that was that
the model?
Yes.
Yes.
So we had basically, had integrated into the adserving technology.
And I don't know.
It's it's an interesting world on its own.
So when you think about the financial marketsand you think about kinda, you know, just no
more, like, you know, traders standing on kindof physical floors and of the stock exchange

(28:26):
and, you know, shouting out prices.
Right?
We are talking about high frequency trading.
We're talking about entirely automatedprogrammatic, kinda highly sophisticated
trading platforms.
That's basically there's an equivalency of thatin kind of the advertising world.
It is highly programmatic.
It's fully automated.
There are these large kind of exchangeenvironments where there's large volumes of

(28:49):
kind of inventory.
Financial markets, must be commodities andequities and bonds.
Here it is basically the attention span of aconsumer in the form of an ad request that is
traded.
And that has signal sparse signal, but it hassignal.
So that we basically tapped into a lot of thesekind of, you know, exchanges to be able to

(29:10):
understand what that kind of connectivitygraph, what can we build out of that
connectivity graph.
And over time we got better and we achievedsome pretty remarkable results of being able to
infer connectivity of devices.
Yeah, and the economics are very attractive,imagine, in the sense that if you're a
publisher and you're making 100,000,000 in adrevenue and you can apply technology that

(29:34):
uplifts that by 10%, 20%, it's like
It's hugely transformative and it's much more
It's huge margin.
Because especially when you think about kind ofthat previous problem of AdMob and kind of the
undervaluing of mobile.
Right?
Because mobile's outcome was completely kind ofin a null space.
Like, what did the what was the effect of kindof exposure of a brand or an ad message in

(29:55):
mobile?
I don't know what action did they they took.
The promise of all of digital was actions aremeasurable.
Right?
So that's why it became so the fundamentalvalue of Google over the course of and Meta now
and even Amazon, these are kind of the thethree horseman in advertising, if you may, is
around the fact that digital and digital ismeasurable.

(30:16):
That's how they've kind of built these massiveenterprises.
So mobile was not measurable and that problemwas tied to identity.
And so you were able to your point reallyincrease value of mobile as well.
So that's kind of the problem that we saw.
At the heart of it, what was very exciting andinteresting for me is is building these, very
sparse.
You start from very sparse graphs, and thenover time you have to build them up to build

(30:39):
the connectivity and increase confidence ofassociativity and edges.
And imagine there's a real network effect thereand there's the more publishers you work with,
the more networks you work with, the more datayou can acquire and the more you can train your
algorithms on the efficacy of that.
Exactly.
It's like AI before AI, if you may, that thisis like a decade and a half before.

(31:02):
What was AI before?
It was basically machine learning is what theearly incarnation of AI has been because it was
not it was more predictive rather thangenerative.
So that's basically this is machine learningdone at scale on large graphs.
And that was kind of the journey and the valueproposition of drawbridge to take kind of the
initial kind of seed companies like AdMob hadraised this kind of creating this potential of

(31:25):
mobile and then kind of that evolving intosaying it's kind of about the consumer no
matter where their engagement and their touchare.
If you have to deliver to that story, you haveto understand identity better.
That's why Meta and Google were really good atit, and this was about being able to create
that opportunity for the open web for anaverage brand, for an even for a premium brand.

(31:45):
The more premium the brand, the less data thatthey had in understanding their consumer in
digital, if you may.
So this was kind of really interesting timeswhere we kinda captured this.
So you started in
2010 '11.
'11.
And you raised in total how much?
So I started with a 6 and a half million seed.
Uh-huh.

(32:05):
And then we went on to raise over 60,000,000 inthe company over about seven years in the
company across three rounds of financing,including some debt financing as well.
And the company had an interesting journey andevolution because this was the time when also

(32:27):
serendipitously Cambridge Analytica and kinda,you know, how consumer consent there was kind
of this rising awareness of what does it meanfor consumers to be participating in consent
around data that kinda footprint that theyleave across the Internet, if you may.
So we went through some really interestingconsiderations.
There was a period of time where it's probablythe Wild West in terms of data was available.

(32:54):
People were obviously had an idea of privacy,but didn't really appreciate perhaps how
valuable a bunch of this data would be.
I mean, today, that's even true today, right?
But to a far lesser extent, there's anappreciation, but are they significantly
participating in that value that is beingcreated by their participation in the Internet
against kinda all these free services that theywe all talk about how much we are the product.

(33:17):
Right?
But exactly to your point, it is kind of thewild wild west at the time.
So there was regulation that was coming in.
As you can imagine, you must have been a partof it as well.
GDPR was one of the most significant ones thatcame in at the time, and it kinda helped us
understand how do we even evolve this.
Like, how do we participate in these kindalarge programmatic, highly automated trading

(33:38):
environments where signals are being inferredto be able to build these graphs?
It helped us evolve how we would do that basedoff of consent.
And then it was kinda, you know, as we wereservicing, we had really large customers,
Disney, ESPN, and I think about largecustomers.
So even kind of platforms such as LinkedIn,etcetera.
It was a kind of a customer relationship thatwe had.

(34:01):
LinkedIn was was a customer.
It was a customer.
It was an early customer relationship.
And it just so happened that I received a callfrom one of the product leads at, LinkedIn
saying that, hey.
We have a problem, because LinkedIn was growingin its mobile presence.
And I don't know if this is the best keptsecret or not, but LinkedIn is a massive b to b

(34:24):
kind of marketing platform.
It's kind of the clean version of, I don'tknow, Facebook, if you may, from kind of even
from a marketing and an advertising.
Oh, think today it's it's quite well it's verywell known, I'd say, b to b sort of, but, like,
five years ago, I don't it was certainlywasn't.
Yeah.
Wasn't it was a it was the size of it's calledLinkedIn marketing solutions.

(34:46):
The size of it was a surprise to me itself kindof when I got to know it more.
And they have this kind of fundamental problemsaying that when you the enterprise buying
cycle is unlike a consumer buying cycle.
It's more consensus and kind of a group baseddecision.
There are multiple touch points.
It's not an individual decision.
You're not, like, clicking and interacting likea one click purchase on mobile.

(35:11):
That's not what enterprise products are about.
So again, going back to that earlier point ofkinda, you know, being undervalued.
LinkedIn had a kind of a fundamental problem toaddress around how to kinda properly value
LinkedIn's mobile engagement because mobileengagement was had output outpaced kind of the

(35:32):
desktop based, browser based engagement.
But yet the inventory had to be better valued.
So this identity problem is still an importantone.
And not unlike the authentication and logged inexperience was is not as pervasive as say
Facebook was at the time.
So there is kind of this problem to solve aswell.
How do we better value our inventory?

(35:52):
So that product conversation from an earlycustomer engagement, we were able to as I said,
we had really become very good at learning kindof identity or learning these kind of pattern
recognition, which is large scale machinelearning on graphs.
And they were kind of blown away because theyhave kinda as you can imagine, like, learning
is what?
You're basically learning on some sample dataand there are this there is this holdout where

(36:14):
you are going to measure against.
And environments like LinkedIn are can be verygood about validation because they can give you
some training data and they can measure howeffectively you are because there is
authenticated data.
And they were kind of blown away by the kind ofperformance that we had.
Obviously, that's why we had this customerengagement.
They put us through kind of a lot more kind oftesting, aggressive testing to see if this is

(36:36):
really kind of it can be.
And this was just as a customer deployment?
Was
in Customer deployment, but then at some pointit became apparent that, look, I think this is
more than a customer deployment.
They kind of want to really see if this can bea capability that they have to look into kind
of more deeply, right?
And that, and this was kind of my version ofkind of, you know, an M and A environment.

(37:02):
It was not quite an M and A conversation, butthere was an in, there was a
You could see it might go there and you wereokay with
I was okay with the engagement around it,right?
Because it does take away kind of engineeringproduct cycles to be able to engage so deeply
with a big provider who's also a customer andwe have other customers that we have to service

(37:22):
as well.
And that basically post an evaluation,aggressive technical evaluation, etcetera.
Then it became a story of, okay, now we areinterested in doing a transaction here.
And is there any so as you look back at thatexperience, it's sort of, you know, as part of

(37:42):
it is your own psychology and how you thinkabout the kind of future part of the company,
part of it is, you know, navigating a kind ofBD to corp dev relationship, and it's sort of,
you know, often advice I give founders, you seethere's a natural partner, don't call up corp
dev, because you're just, you're going to endin one sort of transition and you don't have

(38:06):
you're one path and you don't have thatinternal champion that thinks your product is
fantastic.
Or even not BD, but but kind of product itself.
Would there be as you look back at the the sortof process you went through, is there any
advice you give founders to navigate this sortof corporate customer relationship that may or
may not go into a M and A environment?

(38:27):
I think the best relationships are the onesthat are formed with, like, product teams.
And I barely interact with dev kind offunctions.
I think if the product team is the key sponsor,the product sponsorship is the most important
one, even if you're seeking actively kind of anM and A transaction.

(38:48):
So my advice to founders would be develop likestrong connections as you have these early
customer issues.
It may or may not develop into kind of an M andthat you may or may not want.
But having the relationships with kind ofproduct sponsorships, key product leads across
kind of your key strategic customers isimportant because those relationships, the

(39:11):
updates that you have, the kind of, you know,strategic opportunities that you could partner
with, whether they are in the context of acustomer or a more strategic engagement, those
that action happens largely on kind of theproduct side, at least in kind of very
technically driven product companies.
So I think those relationships are important.

(39:32):
Sometimes founders can tend to our lives arekinda, you know, busy.
To your point, we do fifteen, twenty things inany given day.
These things can probably take a backseatsometimes, but never compromise on these
relationships with your large strategiccustomers having that kind of relationship and
taking time, having the quarterly kind ofmonthly depending on the customer profile,

(39:55):
depending on your relationship, depending onkind of other factors, keep those
relationships.
I think that's extremely important becausedepending on the company product and time,
these evolve into very kind of probablynaturally whether they are acquisitions or one
of the biggest customer opportunities yourcompany would ever have.

(40:16):
I personally felt like I never had a VP, SVP ora head of corp dev.
I handled it myself.
We couldn't I felt it was we were small to tohave that as a function.
I have no statements about the value of thatfunction.
It's an extremely critical function.

(40:37):
But founders, I think, are best kind of brandmarketers of themselves.
Founders have gone more way more savvy todaythan earlier on.
And similarly, I think cognitive functionshould be kind of very founder led.
Mhmm.
You know, imagine there was a sort of decisionpoint between like joining LinkedIn and kind
of, and seeing, okay, can add significant valueto this perhaps under monetized, high potential

(41:04):
advertising business, and then stayingindependent and working across a range of
different customers.
What what were some of the decision points thatyou had?
Were you were you concerned about competitivethreats or concerned about an independent path?
What what were some of the catalysts?
LinkedIn itself was a catalyst.

(41:25):
I mean, I've in kind of the twenty plus yearsof career at this time, I've had a number of
corporate kind of the big corporate kind ofenvironment work experience, certainly startups
as well, different cultures, different people.
I would say LinkedIn is legendary for the kindof culture that it has.

(41:45):
Kind of thing.
It was under Jeff Weiner then?
Jeff Weiner, under Jeff Weiner then.
Terrific guy, yeah.
Terrific.
And I've had other kind of acquisitions since.
I've had other kind of experiences sinceLinkedIn is special in terms of going into the
company.
I had done a little bit of research just tryingto understand as every founder would.

(42:07):
There was a lot that was talked about kind ofthe LinkedIn culture.
Right?
And how much culture was at the heart and ofkind of the success of LinkedIn itself.
To be honest, I had a little bit of skepticismabout is this a little bit of an
overproductionized story?
Like, is it really this true?
And having spent a couple years there, and evennow I I have the strongest relationships with

(42:32):
folks at LinkedIn, and there's genuinenessabout that.
And I I I think it's one of my fondest storiesof kinda the the the team that acquired us.
The executive sponsor was Ryan Ruslansky whosucceeded now the CEO of LinkedIn.
I think having my conversation with Ryan and atthat time I to obviously talk to Jeff as well.

(42:57):
It was all extremely instrumental.
And to your point, there were considerationsobviously to stay independent and we had a
successful kind of path ahead of us.
But I think being a part of this and to getsome of the platform and wind under our wings
that LinkedIn and LinkedIn a part of Microsoft,right?
Microsoft had receded pretty aggressively fromall things advertising at the time, but still

(43:22):
nevertheless had a sufficiently large searchbased advertising business.
And there are lots of intersects andinterconics between Microsoft and LinkedIn as
well.
So kind of the potential of kind of a largercanvas, LinkedIn itself.
And LinkedIn is a very technology product firstcompany.

(43:42):
And that was a big cultural fit for me and kindof the Rob Rich team as well.
All of that kind of dispelled some of the kindof hesitation that we would have in our minds
to kind of continue our independent journeyversus this.
An interesting backstory is I was literally,this is not even kind of, it happened to be

(44:04):
resonating at the same time was literallydelivering my child at the time of the
acquisition.
So actually I, so one of the final decisionmaking meetings for the acquisition, because
LinkedIn is an extremely high ethics company,does absolutely the right thing by kind of

(44:24):
every member who's at LinkedIn.
And I don't say this because they acquired mycompany.
I say this as an employee who saw LinkedIn inaction.
So privacy and the implications of data anddata technologies was extremely important, even
though there is kind of the undermonetizability problem to solve, but do it the

(44:48):
right way.
So we were very aggressively tested for this.
And there were lots and lots of lots ofmeetings.
It was long process.
The The process of security and
Extremely security, privacy, governance, divinginto every aspect of everything that we did.

(45:08):
It was a long process.
And then after all this, after almost seven,eight months of diligence, it had pretty much
butted up against three days before my due dateto deliver my child.
And that was the only day that Jeff a lot offolks had to come together for the meeting.
And it was about eight hours or so.

(45:30):
I was three hours away from my due date and Oh,sorry, three days away from my due date.
And my family had come to help and they thoughtthat this was highly imprudent of me to like,
you know, disclose to kind of the big day.
Nevertheless, and I still remember the entireLinkedIn cast and crew were so thoughtful about

(45:52):
kind of my situation.
And anyway, I
You gave them an ultimatum.
You need to say, you need to decide by thisday.
Actually, I remember so the funny anecdote is Iwalked in, couple folks didn't know about kind
of my situation.
So they're like, oh my god, maybe we should dothis another time.

(46:14):
And I was like, no, there's no other time afterthis is a long mutton.
I'm going to be focused somewhere else.
So let's get this done now.
So it was kind of an interesting.
So we spent about nine hours and then the nextday, because this is the last week you go on
daily checkups, my doctor was like, your bloodpressure is off the charts.

(46:34):
Like, what did you do yesterday?
You were fine yesterday morning.
As it happened.
And he said, you need to be admitted today.
So then I had to go in and then they didn'tknow they were tech.
Anyway, long story short, the next day Ireceived a text that I think we're going to do
the deal.
Again, LinkedIn is very measured.
Even if they're absolutely certain, it's verymeasured.

(46:56):
Think we are going to do the deal, go deliveryour child that.
So I delivered my child, completed kind of thepost term sheet process, etcetera, through the
early days of my child being born.
So it's.
Got it.
The story is the way everything happens atonce.
Don't you think so?
Like, to your point, to your earlier point ofyou don't there are certain things outside your

(47:18):
control.
This was one of those things I just couldn'tcontrol the
timing Yeah, for of sure.
And I'm curious, you're, I mean, you had somebig name investors.
You said Sequoia and Klein and I like how, youknow, the, you know, the independent path
versus an acquisition, were they, did you feelthere was full alignment between them?

(47:40):
Or was it was it hard to get kind of everyoneon board in the same direction?
There was largely full alignment, on this one.
I would say that, at least to your point, Ifelt like a lot of it is very maybe partner
led.
Yeah.
Legendary has a lot of kinda, you know, It's avery strong culture at Sequoia as well.

(48:08):
But Jim, I think Jim Goetz Jim Goetz at Sequoiawas on the board of Drawbridge.
He led the investment.
So he was strongly supportive of it.
So I'm pretty sure that kind of played.
It was also a big learning experience for me.
I had never managed a board before that.
And kinda to your point of big name investors,big name board members, someone called it at

(48:35):
the time you have like a blue chip kind ofboard for a private company.
Yes.
How do I kind of do that and navigate throughmoments?
And it both was a learning experience, to yourpoint, supported the outcome, really kinda got
behind it, whether from their own respectivekinda, you know, partnerships.
I also had an interesting moment.

(48:55):
The Kleiner Perkins partner, Matt Murphy, hadmoved on from Kleiner Perkins to Menlo, but he
continued to kinda be the like, steward forKleiner's investment even though he was at
Menlo.
Right.
So I had that interesting dynamic as well.
And because that's not uncommon, and sometimesyou can see those board members be somewhat

(49:19):
passive.
I'm sure that's an interesting dynamic tonavigate as a founder.
That and in my case, I think it speaks highlyof Matt that, to your point, there wasn't this
kind of disinterest or unpassive kind ofengagement.

(49:41):
There was zero change in kind of the way heengaged and kinda, you know, in many ways kind
of rationalized his position in the board.
And he was deeply engaged, deeply valuable,entirely supportive.
And I'm very grateful for that.
That was my first experience.
I didn't even know how to kind of navigatethat, right?

(50:03):
So my relationship again with Kleiner was a lotchanneled via Matt and I think his kind of
transition from Kleiner to Menlo was sofriendly that even the my kind of exposure into
the Kleiner kind of ecosystem was well kind ofmaintained.

(50:24):
So the relationships were well maintained.
It was an interesting dynamic to navigatenevertheless.
And as you look back on that experience, wasthere you know, we talked earlier about like
the things that you are very happy that you didwas build deep relationships and credibility
within the product organization.
Are there other things through that kind ofprocess that you're happy you did or things

(50:47):
that you you know, if you were to speak tofounders who are navigating a similar
situation, you would do differently?
Maybe the one thing I would say is I think havea more opinionated stance on how you get kind

(51:08):
of, you know, investment banking support andget a transaction.
Sometimes can I think boards can get a push forthat more than
Did you work with an investment banker?
I I did.
And I would say that I think it depends on thefounder size of transaction, the team that the

(51:31):
founder has and the CEO has among themselves,how experienced you are.
Obviously, lots of factors.
But I felt that for me it was among my firsttimes when I did this another time with my next
company.
We didn't engage any banker.
So it does a completely different life cycle aswell with my second company.

(51:53):
But I felt that it was something that I thinkbe a more active participant in some of these
decisions, even though you have highlyexperienced board members, etcetera, at the end
of the day, I think your decision has a kind ofa disproportionate value.
Your perspective has a disproportionate valuethere because everything is you know,

(52:18):
subtracted from the head top the headline valueof the consideration price.
Yeah.
But it's, you know, you're in these decisions,you're like, well, I've
not
done this before.
Let's hire the person that does this kind ofevery single day of the week for their career,
and it's done it for decades.
But I think sometimes the incentives ofinvestment bankers are not always aligned with

(52:44):
the founders, in that they are very expert, butthey there's a danger sometimes that they can
be their repeat customer are the ventureinvestors, not necessarily you as a founder.
And so that or the or LinkedIn, you know, orthe the corporate, the buyer, because they're
gonna maybe sell other things.
And so you know, it's a big generalisation,it's not true, but the watch out is that

(53:07):
they're kind of, they're probably going to dealwith you once, whereas they're likely to deal
with the VCs and the corporate many times withtheir career and so it's, I think.
Synthetise would get in my comment.
A 100%.
I think it's I think it's I think it's verynatural that you're like, well, I need a
professional to help me with this high stakesnegotiation transaction that but I think it's I

(53:31):
think you need to reflect that the incentivesare not always aligned.
I think it often comes down to just theinvestors, like the investors who are kind of
who want to see a sort of successful outcome,but have the founders back in this situation.

(53:52):
Indeed.
Indeed.
And I would say that in my case, the kind ofthe other things that especially when you're
going through this for the first time and inmany founders cases, maybe that's the only time
you would go through.
And it's kind of a high stress moment.

(54:13):
Everything has kind of a magnified kind of yearend.
These are big numbers.
You've not come across them before and you haveto make decisions on them.
It's lots of people's kind of lives and kind oftheir outcomes at stake.
So it feels kind of very high pressure.
And many times there is kind of sensitivity tokind of timing of decisioning, etcetera.

(54:37):
I just like when I did this the second timearound, I think I breezed much more.
And I think as a result of that, I became muchmore, I don't know if I can say much more
objective.
Think hopefully it also impacted outcomes inthe way in which I was able to kind of drive

(54:59):
towards more objective outcomes.
I would feel, I think to your point of advice,would say that while this feels high duress,
high stress, I think it's okay to push back onkind of folks, whether it's the buyer, whether
it's the banker, whether it's your boardmember, all the three Bs, right?
Like, can I just, if you need that extraminute, take it because it's an outcome that

(55:24):
you have to live with?
Your team has to live with inside of kind ofthe buyers.
So it impacts you in ways investors gonna cashout and that's the model, right?
That's the business model.
So I think that is something that I livedthrough it in a much more kind of way where I
began to appreciate it.
And I think I'm better for it today.
When hiring a banker, the, you know, the expertif you hire a banker, then you almost push down

(55:49):
an M and A process, whether you like it or notor you're in danger of.
And then also the incentives can sometimeschange when you start to go through this
process where you move from a situation whereeveryone's aligned to building a great company
to an area of kind of self interest, you know,whether you're the founding team or an
executive or the CFO or the venture investorlike this, and so I think it's things change in

(56:15):
a surprising way, and I think if you if you'renot cognizant of that, it can be, you know, you
can not have the best outcome from thefounders.
Maybe one other piece of advice is that asfounders, the thing that we are breathing,

(56:35):
doing every single moment is focused on thecompany building, Do the right thing, whatever
it is, good things will happen.
We just do the right thing.
And so as a result of that, at least in mycase, I didn't prepare myself for kind of these
events as much because I just didn't focus onthem.
If it had happened, it would happen.
I'm focused on the company building.
Yeah.
I would advise that some, I think founders andtoday's generation of founders, think are

(57:00):
better prepared.
There's so many more sources of information andkind of ways in which forums of discussion are
happening where I think the founder preparationis much higher today than I think, I don't
know, I have felt I was ready for.
Just at least having a preview, hearing otherfounders, their stories, talking to them, at

(57:20):
least learning from their experiences isimportant.
It's important to have that you don't have tospend time thinking about it and planning for
it prematurely and completely kind of randomly.
But doing that is important so that when itcomes to you, you don't feel like you're
starting from a cold price.
So you
joined LinkedIn?

(57:41):
Yes.
And then you joined Microsoft.
Yes.
And then you started a new company.
Yes.
Samooha.
And so what was the what was the sort ofcatalyst behind that?
And I think you've you, I I read elsewhere thatyou said that, starting a founder up is a high
fixed cost, activity.

(58:03):
What what do you what do you mean by that, andwhat inspired you to do it all over again?
For me, the reason I say high fixed cost is thekind of the upfront cost of like, you know, of
like just getting the company off the ground,getting the kind of the early kind of MVP of
the product, finding kind of that early productmarket fit, finding those design customers.

(58:27):
And I miss the biggest point, your team thatcomes together to build this.
I mean, the entire magic is the team that comestogether to do this.
These are all extremely hard problems, right?
And there's a huge cost associated with all ofthis.
And I don't know if late I've started sayingthis more, like when kind of one of the biggest
kind of successful founders of our generation,Jensen Huang sits on stage and talks about how

(58:52):
difficult it is.
There's a deep appreciation for that.
Irrespective of whether it's a3,000,000,000,000 company, 3,000,000,000
company, or 300,000,000 company.
I think what he what did he say?
He said, like, you know, when asked, would youdo this again?
He goes, absolutely not.
And yet he's been, you know, he's changed anddominated the entire industry and he wouldn't

(59:16):
do it again.
So to that point that you just said, I justwanted to say that that's true no matter
whether it's 3,000,000,000,000,300,000,000,000, or 3,000,000,000.
That that's what I mean by irrespective of kindof the 3,000,000,000,000 or three the order of
magnitude, the fixed cost from a founderperspective for kind of to hit that escape
velocity is very high.

(59:37):
Mhmm.
And so there are I that was kind of my way ofrecognizing that, you know, and also probably
kind of giving a sense to founders that yourbest bet is to take a shot at yourself and
invest in yourself, but do it with eyes opened.
That there is a big cost associated kind of towalk down the founder journey, especially to a

(01:00:00):
successful founder journey or the other way.
So it's a big cost to walk down.
So that was kind of my notion of kind of thefixed cost, if you may.
But for you, so with, you know, in any sort ofyou're quite a rational person, I'm sure
there's a cost, but there's also an opportunityassociated with that, and so you must have seen

(01:00:22):
a massive opportunity to kind of to go throughthat cost again, because it's like, you know,
you're you're a mother, you're going throughthat, gone through that journey.
It's like, what was the opportunity that yousaw?
So irrespective of the cost, the reward isincredible.
It's not just simply of the reward I think goesthe spans the multiple dimensions of kind of

(01:00:48):
the value that you create in terms of theproduct that you build for the world around you
and kind of the sheer high.
There's nothing to beat that high.
There's no amount of money that matches thehigh of creating value out of a product that
you've built.
And again, that is a whole kind of spectrum inon itself depending on the kind of product that

(01:01:09):
you do.
So for me that reward is disproportionatelyhigh.
With the recognition that the cost issubstantial, that reward is high.
Of course, there are kind of other more kind ofworldly rewards if you may have like, you know,
financial outcomes, otherwise, etcetera.
That happens, it comes along the way.

(01:01:30):
And for different founders, differentindividuals, for me, those are immaterial.
I like, I'm somewhat addicted to this, the highof creating value.
And I love that kind of zero to one journey.
I think it probably goes back to my kind ofroots, my PhD roots.
It's kind of this building from ground up, thisfirst principles, there's nothing on paper.

(01:01:50):
So I remember Matt Murphy at Kleiner Birkinswhen I started drawbridge.
I literally just walked in with a laptop on thefirst day, like post funding.
He's like, that's it.
I was like, yep, that's it.
We're going to start from here.
So I don't shy away from that.
And I find that to be one of the most rewardingparts.
I don't know if you agree with that and kindayour experience.

(01:02:11):
The early days are kinda the most thrillingdays.
At some point posts kinda certain scale, theunrewarding parts of the founder journey are
kind of just the general managerial kind ofaspects, administrative aspects of running, but
kind of the innovation aspects are highlyconcentrated in kind of the early days.
Yeah.
I I I think that I mean, I look at it likebalancing constituents in some ways, like, at

(01:02:35):
the very beginning, very, very beginning, it'sjust you and the founders, you're kind of, you
have an opportunity to, your only constituentis yourself, and so you're in a position to
optimise for things that you are uniquelycapable to do, and you're in an environment
that you uniquely love.
Like, this is the people you want to theculture you create, the team that you love.

(01:02:57):
And then you bring in investors, and you bringin investors that hopefully you love to work
with and inspire you every day.
And then that, you know, is generally, youknow, additive.
And then you but the more constituents youbring on, bring customers, you bring partners,
bring more employees, it's like there's morecomplexity, and you can't actually delight all

(01:03:17):
those constituents consistently all the time.
But it's that early phase where you have a, youknow, small number of constituents that you can
delight them is just it's such a thrilling partof part of the journey.
You said something that's, like, really struckme that you can't you can't delight all of them
at at the same time equally, etcetera.

(01:03:38):
Is that one of your realizations that youdeveloped as a founder CEO yourself that, look,
I'm not going to be able to satisfy all myconstituencies by an equal amount, certainly
not at the same time, certainly not all ofthem.
And was that liberating for you once yourealized that?
Yeah.
No, I think so.
Mean, I ended up like, because you end up inany, particularly in marketplace businesses,

(01:04:01):
but also ecosystem, but you end up, you know,what I ended up doing was stat ranking my
constituents, you know, what is number one incertain organizations, your customers need to
be number one, in certain organizations, youremployees need to be number one.
And then actually, if you have it's a veryrational way to do it, but otherwise you're in
this decision paralysis, because you can'tdecide, well, do I delight consumers or delight

(01:04:28):
employees?
Like, and making that kind of like reallyforced decision to kind of, this is the step,
right?
And of course, it's not black and white, but asa framework, helps us choose between, okay, are
we gonna make a decision to, you know, it couldbe we're gonna do things at the obsession of
our customer delight, and we're gonnadeteriorate our financial performance, which

(01:04:54):
may make short term investors disappointed.
So be it.
That's fine.
But that trade but I think if you're notconscious of those trade offs, then you're
gonna fall into a trap, and it's gonna beproblematic for you
as a is one the nuggets I hope the listeners ofthe podcast get.
This is an important nugget, this kind offramework.
Otherwise, to your point, I think decisioningis one of the cardinal things that we as

(01:05:19):
founders do, in kind of in the presence of aninsufficient, incomplete information.
And we are not taught.
We are not gonna professional decision makers.
I don't know who is, but we are certainly notkind of bred to do that.
And then we're thrown into an environment wherethat's something that we have to do routinely,
right?
So the framework that you laid out is kind ofone of the very important ones that should have

(01:05:43):
some framework.
Otherwise, you're going to be like in decisionparalysis and emotional duress.
And it's also, I think it's very helpful tohave that transparency with your board and your
employees as well.
This is our priority.
This is what we're going to do.
And, you know, absent of that, you're going toget knocked around by different participants to

(01:06:03):
make those decisions.
Absolutely.
So anyway, coming to my yeah, I joinedMicrosoft.
I spent time with Satya and Kevin Scott.
Kevin Scott has this group called Office of theCTO.
It was called Octo.
So it was where all kind of a lot of thestrategic projects including OpenAI at the
time, etcetera, excuse me, came out of.

(01:06:24):
So it was really amazing times.
I mean, the biggest learning that I had beforeI talk about kind of my next company is seeing
Satya in action.
I mean, it's a delight.
It's a delight to be in his meetings and yousee what a CEO AtScale of kind of operates as.
And it was amazing to see a leader of his kindnot insist on the mic time in a meeting.

(01:06:50):
And so that was kind of one of my biggestlearnings I had that.
So anyway, that was a great experience.
And then Kevin, I joined Kevin back again afteralmost over a decade because Kevin You were
with an AdMob.
At AdMob.
And then when I joined his team, it kind offelt that it was kind of the super SWAT team of

(01:07:11):
specialists, they're called technical advisorswho basically work on like super special
projects at Microsoft.
And I think I may have had a conversation withKevin.
There was a sense that I'm kind of this personwho probably will jump into my next
entrepreneurial pursuit.
But this is, again, going back to kind ofKevin's kind of USP.

(01:07:33):
Right?
Again, going back to kind of that mob USP andthe leadership the ability to let be and the
ability to know that I'll take whatever time Ican with smart people because they're bound to
go ahead and do other things.
I think that kind of acceptance is also hugelyvaluable at creating fostering talent and not

(01:07:53):
having to kind of hold on to them for becausesometimes it's human tendency to want to hold
on to people.
I certainly kind of went through that.
I think I learned that as well from Kevin.
And so then when I was thinking about this ideaabout like, I think this maybe resonates more
with you, Pete, is that this kind of this wholenotion of, you know, escrow and kind of the

(01:08:17):
world that the product, that legendary productyou've built and kind of we all have gone
through an escrow transaction at one point inreal estate at one point or another in our
lives.
That notion around data doesn't exist becausedata has this highly sensitive nature to it,

(01:08:38):
especially enterprise data has some sovereignvalue to it.
It has competitive value to it.
And yet businesses have to transact with datawith each other.
It's not just a consumer at an enterprise.
It's business and business as well.
And, again, a world that you, of course, havegone through and kind of the products that you
have built yourself as well.
So to think about like products, data products,data clouds that provide that kind of secure

(01:09:00):
governance collaboration environment whereenterprises can share data without making
copies of the data across kind of cloudreplication environments, etcetera.
I think that notionality is extremely criticalacross all industry use cases, but certainly in
the media marketing industry with, again, goingback to kind of the evolution of regulation,

(01:09:22):
etcetera, circa post AdMob, post drawbridgethat kind of got me to this like, will there be
kind of a world where Slack or Dropbox,etcetera, less like cloud storage collaboration
environments will evolve to kind of datacollaboration while keeping the security

(01:09:43):
governance of the underlying data.
And you can actually do good things out of datarather than simply generate user generated
communication content.
And that was kind of the heart of this company,Samuha, that I founded.
And while it has, and I'll be the first personto say this, it has a very reductionist way of

(01:10:04):
thinking about it as a clean room because itcomes as kind of clean room from kind of the
fab clean rooms you think about or even kind ofthe data clean rooms when you think about
transactions, the finance capital environments.
That's kind of behind what the positioning ofit as a clean room.
But my thinking of it is, could we create kindof these intelligent environments where there

(01:10:27):
will there be a generation of datacollaboration where you're not simply talking
and collaborating with your colleagues andpeople within your enterprise or across
enterprises, you'll actually conduct under theparameters of governance security policies that
you set up in your data.
So will there be a world where we as anenterprise employee, not only do we get our

(01:10:48):
badge, our laptop, and all the kind ofassociated enterprise devices, we get policy
data depending on the function, the role,etcetera, that we are in a given enterprise.
And we will become citizens of data, stewardsof data within the enterprise.
We become empowered to do that.
That requires data collaboration to be doneright within an enterprise and across

(01:11:10):
enterprises.
So just so I understand it, all enterprise dataand consumer data is now in the cloud.
And then the opportunity is enabling data to beor companies to analyze and understand data

(01:11:33):
that sits in multiple clouds, in multiplecompanies, in multiple environments, and to
make decisions, analysis, recommendations fromthat data without sharing that data, without
without combining that data in a but to do itin a way which is maintaining security,

(01:11:53):
maintaining privacy, primacy of that data, butto enabling that in this safe environment is a
massive unlock, particularly in the world of AItoday where you see opportunity is really
fueled by data.
Exactly.
And I think to your point, it's a massiveopportunity.

(01:12:13):
And exactly to the point that you said,irrespective of kind of the cloud environment
that you're in.
So when you think about that, at least when Iwas thinking about the idea and then that's
where kind of for me, I don't know, there was acompany that came to my mind when I was
thinking about this idea.
It's when I thought about the cloud agnosticcity, it was Snowflake.
It's like it is because they have the cloudagnostic city.

(01:12:35):
It's a highly government governance is a bigpart of what Snowflake offers.
But kind of this building, this kind ofcollaboration capability for enterprises is a
great unlock.
And kind of that's where I felt like as anindependent startup, we can start building
maybe for certain industry use cases.
That's why clean rooms and media marketing,etcetera.

(01:12:56):
But when you think about this and I don't know,healthcare and how healthcare data can be
between providers, pharma, payers, hugelyregulated compliance requirements around all
these data and how data can be more empowering,enabling.
And similarly, you think about, I don't know,you think about kind of the way in which I

(01:13:20):
think government is moving, we sayinterdepartmental agencies on data sharing.
How do you kind of create this kind ofenvironment for agencies to share data with
each other without having to, again, createcopies of the data.
Migration and replication of data is at theheart of like, you know, crumbles of data that
you leave at the heart of privacy, primacy, andto your point sovereignty of data, the breakage

(01:13:44):
of that.
So that was kind of the idea and the productvalue prop under which I wanted to start the
company.
I even thought about it should be kind of dothis at Microsoft.
But again, kind of there is this kind of thereason we live in the area that we live in is
because there are people who want to move acertain pace.
And that's something that has kind of been kindof constant for me.

(01:14:04):
So I felt the only way I could do it was kindof start a company and kind of move with
purpose and intent.
And I got together with my co founder and weboth started this company and
And your co founder background is
Is in encryption and cryptography from Apple.

(01:14:26):
And so he and I got to meet each other.
His name is Abhishek Bamik.
I got to meet him through a friend of mine andkind of really, really get a meeting of the
minds very early on.
It's deeply technical problem.
It's a deeply technical problem to solve.
And so in some ways, the it's even it'stechnical in a kind of a different dimension

(01:14:51):
when you think about cryptographic techniques,secure multiparty computation.
When you think about cloud based technologies,you think about that dimension versus kind of
machine learning on graphs, if you may.
But again, the undercurrent is kind ofalgorithm development at the heart of kind of
both of it.
So that's where some of kind of, you know, mycommon value exists.

(01:15:14):
And my personal development is that I think I'mnot purely a technologist or I don't know, not
even a techno, an algorithm kind of person.
I think I started understanding the value ofproduct, product market fit.
Does it mean to have kind of sales and businessdevelopment and partnership and raising capital

(01:15:35):
and being able to communicate kind of customervalue, deeply care about customer value.
And all of those things makes me a better kindof more rounded founder versus being kind of
the technical founder.
Yeah.
I mean, the marketing industries are highlycommercial by design and so you and they have a
very economic they have a clear economicimperative and, you know, the product will live

(01:15:59):
or die by its efficacy in the market ratherthan sort of a core engineering And so you are,
nature, those industries, they're tough.
They're tough industries.
You can win and lose in the same year justbecause things could change so quickly.
I'm so
glad you said creates a very commercial.

(01:16:19):
I'm so glad you said that because in one of myconversations at Stoflego saying that, look, I
mean, one of the industries that I think theworld at large has kind of a love hate
relationship with advertising marketing for allthe right reasons.
Right?
I mean, what it symbolizes or kind of means.

(01:16:40):
But I would also say that it's one of the mostelastic industries where kind of if you have a
product that works exactly to the point, it'llhit faster.
The industry will absorb it faster than otherindustries.
But by the same token, if it doesn't, it'll bedropped the very same day as well.
So I I think very well said in what you did.

(01:17:01):
So that was kind of the product that I gottogether with my co founder.
This is an industry which has some of theseproblems that present itself.
And you think about kind of how the bigplatforms, etcetera, are stepping away from
kind of trackability and that's a constantlyevolving topic.
And even most recently Google with all theantitrust, are positions that Google has

(01:17:21):
reversed itself, etcetera.
So it's a very dynamic environment.
And this kind of handling customer data,consumer data, and doing that in a compliant
privacy safe fashion is an extremely importantproblem.
It's as important in health care.
The problem is health care moves so much slowerthan this industry does.
So that's why you're able to make moves in thisindustry faster.

(01:17:43):
So we got started, we got some really earlykind of wins like big names like T Mobile,
etcetera, that we kind of got to win.
And then it was interesting.
And then you raised initial capital from?
From Altimeter.
And I met Brad Gerstner at, I sit on the boardof a company called iHeartMedia.
And so Brad is my fellow board member there.

(01:18:08):
He was a board member at Zillow early on.
Oh, really?
Yeah, yeah, He was like he's way back a travelguy.
That's true.
That's true.
You know, an online travel guy and then, well,he's now a generalist investor with a But,
yeah, he was on the board of Zillow before themerger between the companies, but now I've

(01:18:30):
never He stayed on after the merger?
No, he left before, but he was always, like,very engaging and we stayed in touch, so like,
he's, yeah, he's a terrific and very, very
sharp Very sharp.
Sometimes say his superpower is his ability tolisten and kind of synthesize information.

(01:18:51):
Anyway, I met Brad and then he has been veryactive with Snowflake and kind of once I just
called him and thought it was gonna be an Iheart call or something.
And I was like, hey, I've been getting, youknow, working on this and wanted to see, I
think they intersect with Snowflake is verystrong here because this has a cloud agnostic
city and this kind of we should partner withSnowflake here on this one.

(01:19:14):
He's like, this makes a ton of sense.
And one thing led to another.
I had funded the company, got started, then heput in an early check and an ultimator and
Snowflake.
And I got connected with my current boss who'sChristian Kleinman.
He's the EVP of Product at Snowflake, anamazing guy by the way.

(01:19:36):
So he saw some of kind of what we had put in.
We had developed as an MVP product.
Things were extremely synergistic withSnowflake.
And we had very quick early customer wins aswell.
And within a year, Snowflake put a prettycompelling offer to buy the company.

(01:19:58):
This was about eleven months or so since thecompany was.
Did I read right that you took money fromSnowflake Ventures?
Yes.
And that was with Altimeter at
the With Altimeter.
Because that's probably an interesting decisionbecause you're like, you know that Snowflake is
a phenomenal platform to build this applicationon top of, to give you distribution and access

(01:20:22):
to their ecosystem and customers, but takingthe money is also may taking the money from
from Snowflake Ventures may create lack ofalignment or kind of or potential conflicts How
of did you think about that?
A 100%.
I think great question, Pete.

(01:20:42):
So I interesting story around that as well.
So I've been kind of a fan girl of Snowflakes,you know, enterprise success from its kind of
early days.
So there is a moment during all of this kind ofmachinations.
I was on a customer call.
I got a number, a call from a 408 number, and Ididn't answer.
I didn't recognize the number.
And then the text popped up saying, hi,Chemukshi, this is Mike Scarpelli, CFO of

(01:21:07):
Snowflake.
And I had like a total fan girl moment.
He was like, can you call me back?
I was like, yeah, sure.
I'll call you back after.
And this was when you were just conceiving ofthe idea?
No, this has been the machinations of exactlyto your question of like, does it make sense
for Oh, snowflake make your
own discussions with different ventures andthen
Yes.
So I got it.
The CFO comes in.
Yeah.
Okay.
So anyway, was, I think it's a great point andI did grapple with it a fair bit.

(01:21:34):
I had kind of a personal kind of mantra afterDrawbridge to say that I never do corporate
kind of investor ever.
It comes with a lot of strings attached.
And it's, I don't know, usually painful.
I don't wanna be too kind of blanket statementabout it, but this was kind of my personal

(01:21:54):
experience.
And I kind of violated it with with SnowflakeVentures, if you may.
And I think exactly to your point, it had kindof a lot of kind of considerations for me.
At the time of company inception formation,this could be many things across even competing

(01:22:18):
kind of companies, etcetera.
It doesn't have to be and to your point, it canwe could distribute this across the
hyperscaling kind of CSPs, the cloud serviceproviders themselves as well.
So there are so many directions in which wecould go.
So Snowflake Ventures, does it come with toomany strings attached?
So I would say that Snowflake, and especially alot of kudos to to Christian himself and the

(01:22:42):
Snowflake team, they did a great job of like itdepends on each founder and kind of their
personal kind of line that they would like totread.
But they set a line that was fairly independentto say that, look, you go ahead and like, you

(01:23:03):
know, obviously there were there were therewere kinda us initially to see if we could do a
like, you know, more like snowflake leaning,but we we had a desire to kind of explore the
full spectrum of possibilities.
And the snowflake ventures check, I was quitespecific that should not keep us bound to kind

(01:23:26):
of, you know, one environment platform oranother.
And I give a lot of kudos to Snowflake Venturesthat it was not heavy handed in the sense that
it comes with an exclusivity and that youcannot work with someone else.
Yes, I've seen it many different ways, butcertainly seen some excellent corporate VCs

(01:23:48):
that and in theory, they could be some of thesingle best investors, because they're like,
they understand your market, they are alignedwith the outcome in terms of success, in terms
of they want to help you succeed, they can helpyou succeed, and they're going to be rewarded
for it.
And so I've seen some scenarios where it's beenincredibly beneficial.

(01:24:11):
I think Shopify Ventures have done a good jobrecently.
They're not trying to, you know, that manycompanies think worry about platform risk.
And in some ways, like having Snowflake, aninvestor, kind of mitigates that platform risk,
because as long as they're cognizant, they'renot trying to pollute that platform.
I've also seen the opposite, where they kind oflike, you know, use investments for kind of

(01:24:36):
rights of first refusals, for information, forcorporate learning and development and so.
But it sounds like this was a great So
So mean, exactly to the point Pete, you saidthat to my earlier comment of corporate VCs, my
personal experience wasn't very pleasant wherewe are on.

(01:24:56):
So that's why kind of I swore off of it.
To your point, there are many corporate VCs whoare excellent and great investors in their own
right.
Depending on the corporate and exactly to yourpoint, Scott could be a match made in heaven.
I think I ended up in that kind of fortunatepoint on the Snowflake side as probably a

(01:25:18):
special case because there are many other kindof situations where it's probably gone wrong as
well.
But in this case, would say that I ended up asa special case as an example of how it went.
We had the opportunity whether we didn't havethe shackles around us to kinda, you know, bind
us to one or the other.
Honestly, the kinda early customer wins that wegot, how does T Mobile work with like an eleven

(01:25:43):
month less than eleven month old startup?
Like, you know, like big names, Comcast, TMobile, these were all the kind of things that
I would say the go to market vehicle thatSnowflake can expose you or Snowflake, not
vehicle, but kind of the go to market ecosystemthat can be opened up if you are a good product
for kind of the fit around Snowflake and itscustomer base is tremendous.

(01:26:07):
And that's kind of, I would say in our casebetween the ventures and Snowflake's kinda, you
know, value prop itself of opening up anecosystem.
If you are able to again create the rightalliances with their sales team, etcetera, it
becomes like a real vehicle.
Snowflake sales team does a great job of kindof partner sales, like how they bring effective

(01:26:31):
partners into customer conversations.
I think especially in kind of today's kind ofpost AI world where we are about, like, lower
cost of sales.
I think alignment with the likes of Snowflake,etcetera, can be really interesting
considerations for founders.
Where is there a way in which you are able toagain, depending on the product, depending on

(01:26:52):
the category and the industry, I don't wannamake blanket statements.
I think leveraging that those kind of dynamicswith companies such as Snowflake and can
accelerate the time for initial go to market.
I'm not saying that it's a replacement forscale good market motion.
But for initial go to market by compressingthat and realizing kind of initial kind of

(01:27:12):
success, I think companies like Snowflake canbe extremely valuable there.
And certainly, Simuha was kind of one such kindof player and product.
Then so Snowflake acquired Samuha, would yousay eleven months after the initial financing?
Technically, I think the acquisition closedslightly after the twelfth month, after the

(01:27:35):
year.
But and and so what was the I mean, was it thefact that you could see this one plus one
equals four?
It kinda you can obviously, you're building inpart on their platform, and then you kind of
can leverage their go to market and customerrelationships.
Was it was it as as simple as that?
Okay.
We're we got to this team.
We really like this team.
They can accelerate our vision five, ten x.

(01:27:58):
Is it was that
that's basically they can accelerate our visionfive, ten x.
And then there's see the the there's alsoanother kind of product reality or a go to
market reality I was contending with.
Again, going back to kinda I'm a bet I'm a morerounded founder today versus just being a
technical founder.
Right?
While the product of collaboration is extremelycritical one, I was quite aware that I'm

(01:28:22):
looking at kind of a polynomial scale ofcomplexity of sales.
I'm not selling to one customer at a time.
I'm selling to multiple customers at a time.
Because you need to kind of to It's datacollaboration.
Data collaboration.
Yeah.
Selling to one customer is hard enough.
Selling to multiple is supremely harder.
So it was a dynamic that I was quite aware of.

(01:28:43):
So I could have said there was a time when inthe kind of company inception, was like, look,
it's an interesting problem.
It is going to be solved maybe by thehyperscalers themselves or maybe by an
independent company.
It just pro it can be prohibitively complexfrom a go to market perspective because of
which maybe I should step back from this.

(01:29:04):
Like, I won't be able to, there's not enoughkind of, you know, venture dollars available to
fund building kind of a team of the size to beand at the time span to be able to bring all
these kind of the time to sales for all ofthese participants to customers to come
together.
But I still went forward with it.
And for me, the kind of ecosystem dynamics, thelikes of Snowflake played was a big part of

(01:29:31):
being able to get that lift off.
So for me, this one plus one greater than two,I lived it even in those eleven months and I
felt like it is extremely critical for thesuccess of this product.
So,
and having kind of seen the patterns of it andhow like at least kind of the sponsor Christian
in this case was supremely supportive.

(01:29:53):
The sales team was great in terms of theinteractions that we had across product sales,
partnerships.
It was kind of a great combination.
Mike Scoutelli is a compelling negotiator,compelling personality himself.
A lot of things came together.
I felt like, okay, this is a moment in timesupremely premature, to be with you.

(01:30:14):
So one path is raise an enormous amount ofmoney essentially to kind of to actually go
through this multiple sales process,complicated build out with, you know,
significant potential, but like a ton ofexecution risk and financial risk against that.
And the other path is to sort of build thatwithin Snowflake itself.

(01:30:38):
That And Snowflake constructed it very welltoo.
And I give Snowflake a lot of kudos in the factthat, I mean, they were recognizing of the fact
that, look, this could be a big independentcompany for ourselves and like, we don't have
to do this.
So, I mean, without going into the specifics ofit, Snowflake, even as a part of like construct
of the transaction itself, Snowflake is verycreative in terms of doing this the right way.

(01:31:00):
And I think over the years they've gotten verycreative, very compelling, very aggressive in
terms of identifying targets and making it kindof very compelling for whether it's the
founder, the product, etcetera.
So I would it I think a combination of thesefactors were all very kinda interesting.
Interesting.
There's the earlier version of this, I don'tknow, Pete, you know, Cisco had this kinda had

(01:31:24):
championed this alpha model of kind of the spinins, you know, like these cap outs, the spin
outs, and then there's this kind of the spinins bring them bring them back in.
I don't know if that ever happened at scale.
There were certainly some kind of noteworthysuccesses there.
Right?
I think that this is not quite that, but Ithink and I don't know.

(01:31:48):
The real version of that is called venture.
That's what you do.
There's no other outcome for it.
But I think Snowflake at least maybe in one ortwo instances has done it kind of successfully.
I don't know if this is necessarily qualifiesas that, but maybe this has flavors of it, if
I may.
Yeah.
I think, I mean, it's recognizing that to buildtransformative technology, it can be really

(01:32:10):
hard to do that internally.
And so bringing in, you know, other startupsperhaps prematurely and scale that within that
environment is the the more likely path tosuccess.
And it may appear on paper very expensive forthe kind of for the sort of incumbent team, But
the opportunity cost is enormous because theworld Exactly. Is

(01:32:32):
Is moving so
And I think it also depends on the executivesponsorship, etcetera.
Exactly to your point of the incumbent team onpaper may be expensive, their conviction, their
belief, how they're able to kind of rally anorganization, a team behind it.
So I think there is some kind of the rightperson, right time, right company, right

(01:32:55):
product, right founder team.
All of these have to come together.
I think Simuha had that kind of as a moment intime kind of thing.
So when asked me, so Drawbridge was AdMob wasfour years, Drawbridge was eight years.
And then Simuha was like less than a year.
So all I said is you live and you learn.

(01:33:18):
Don't know.
They were like, what happened?
So, but again, it's my takeaway is that I thinkmy journey has evolved through And I think I'm
a fuller person who's able to identify thingsand opportunities and fits better.
Again, going back to this question that youasked about kind of what is the advice you

(01:33:42):
would have for a founder?
See my that corporate development done withproduct teams.
I did that relentlessly from kind of my secondhalf of drawbridge into Samoa.
Like at Samoa, my connections into AWS orSnowflake or a bunch of other big hyperscalers

(01:34:04):
was very deep from pretty early on.
So we were working with a lot of these guysfrom pretty much inception because I think that
product partnership is something that I havehas played out well for me for the class of
products and especially enterprise deep techdata machine learning and not yet AI products
that interface is extremely critical becausethe big kind of incumbents are they have

(01:34:31):
strategies in this direction.
They have not fully built out kind of products.
So they are paying attention to it.
They have some leverage in one way, shape orform either in terms of an opportunity like
LinkedIn marketing solutions and Snowflake'scase, you know, collaboration products
themselves.
To your point, the opportunity cost is high forthem because they have existing momentum.

(01:34:52):
So I think some of identifying these things,finding the right partners, interfacing with
them, getting started early on is invaluable.
Yeah.
Incredible opportunity.
Well, I just have one last question for you,and it's been amazing hearing your story.
So you've spent the last fifteen years in theheart of data in Silicon Valley, building and

(01:35:16):
in startups and corporates, and here we are injust an incredibly dynamic, exciting
environment where data is fueling a lot of theAI innovation.
And so curious if you were to give advice bothto early stage founders who are looking at this
space today, and then also for corporates andenterprises thinking about this.

(01:35:39):
What advice would you give to those twoconstituents, knowing what you know today?
To founders, I hope what I'm about to saysounds appropriately measured and not like too
dreamy.

(01:36:01):
If you have a chance to give a shot toanything, give it to yourself.
You are your best investment.
And I think there is a so I say this when I saythis to founders.
I know it sounds oxymoronic.
Haven't they done that already?
That's why they're founders.
But I'm saying this to kinda, you know, goodengineers, good product managers, good

(01:36:24):
salespeople.
Like, kinda, you know, employees themselves whojust are basically contemplating ideas but are
not risk forward.
Then for some reason, the risk forward kind ofweighs in on them.
It's a tough one because it seems infinite.
So I would say that I think you should giveyourself that chance.

(01:36:47):
And as I said, that's why I hope I'm not toodreamy.
To invest in yourself.
Invest in
yourself.
Take a risk.
To take a risk and invest yourself.
It's literally the best bet any human being cando for themselves.
It's the most amount of justice we can do toourselves is invest in ourselves.
And it does involve a moment in time when therisk seems infinite and the reward seems

(01:37:08):
unknown, in fact, zero or negative at best orat worst.
But I still believe that it is worth the riskthat one should take for oneself.
And the earlier you do it, the more chances youhave for kind of course correction and
recalibration and re pivoting and pivotingyourself in the right direction.
And when data startups, especially forfounders, it's a tough one unlike, you know,

(01:37:35):
founder journey products are stuff across theboard.
For machine learning AI and data infrastructureand data startups, there's this chicken and egg
problem at drawbridge at some of that.
There was one common, Where is the data?
I I I'm just getting started.
I have no data to access.
So this is where understanding how do Iidentify some design customers, you know, big

(01:38:00):
large corporate incumbents completely out oftheir realm.
Have to find the right size of Big like enoughbe Big enough big enough to be interested, but
not too big.
Yeah.
Not too small where they themselves are kind offiguring this out.
So I think finding that as kind of the earlyinsertion point.
This this this is all art that is as muchnecessary as the science of being able to solve

(01:38:21):
for this.
Right?
I think data startups come up against that.
AI also comes up against this.
Where's, you know, access to data?
So I I don't know.
Take a shot at yourself and understand how tocalibrate data acquisition.
It's extremely critical.
And maybe for kind of, and I have kind of this,even in the world of I have to give a shout out

(01:38:42):
here, given that I'm a female founder.
I have this like this thing of WTFF, where'sthe female founder?
I want more female founders to kind of give a,take a shot at yourself and risk forward tends
to have some correlation with gender.
There are other factors as well.
I don't know, maybe that's kind of my call outthere.

(01:39:06):
And to corporates and incumbents, I'd say thatI think having the product teams empowered.
Like I really like the Snowflake model in Vish.
Like Christian has kind of an interestingphilosophy where like it's huge amount of

(01:39:27):
autonomy, huge amount of empowerment, productleaders, product leads work actively with
market like, and with startups to be able toidentify kind of new innovative ideas,
companies, teams engage with them deeply,understand how to bring them into kind of the
platform, become early adoptees.
This is not generally true across every typeof, you know, tech product, but I think having

(01:39:52):
this kind of culture kind of making kind offounder friendly, deep, high amount of
empowerment for product leads within incumbentskind of creates a good dynamic where people
become resourceful and identifying kind ofpockets of talent and opportunity that can

(01:40:14):
create acceleration paths for the incumbents.
That is not very top heavy, like, oh, the CEOis like, I'm sure that there were times when
there are certain companies that you gotexposed to, you identified, etcetera, but there
are probably equal other opportunities foracceleration at your time at Trulia, Zillow,
etcetera, where, you know, others were able to,and your team were able to identify and bring

(01:40:36):
that to you.
I think particularly now when the rate ofchange is accelerating so quickly, access of
information if you're a CEO is, at a bigcompany is just like, you just don't get the
signal, it's you get this and then the kind ofthe rate of change is so fast, you need someone
who is like inside the organisation, able toaccess the information, understands the

(01:41:02):
business needs and are able to act on that veryquickly.
And that then product creates the rightinterface, whether it's with kind of venture as
a vehicle of investment or whether it isSculptev as a vehicle of acquisition to
accelerate kind of your innovation.
Think that product function, I think it's animportant one to be able to do that.
So at least kind of the companies that I haveeither been acquired by or worked with, I think

(01:41:28):
this function is an important one.
So I hope that incumbents are more forward toyour point of this level of rate of change that
we are living in so that like the leadership'snull space doesn't become the company's null
space.
Thank you so much for joining us today.
Thank you very much.
It was great to have this conversation.