- [John] Welcome to Thursday's class, we have Nirav Tolia who I've known for about a billion
years, I'm really tickled to be able to spend time with him. - You were an undergrad together here. - We were, and one of the things that -- - 20 years ago - 25 years ago, yeah. - Please, don't. - I'm internalizing it,
yeah, I'm the old one. - That's right. - So Nirav, one of the things people don't know about Nirav is he went to the high school that is on
Friday Night Lights. And so, what Nirav will tell
you, that's pretty close to accurate, that's pretty
much what it was like. - We both grew up in Texas,
and so we lived the dream. - That's exactly right, so,
ok, we're gonna talk about networks and growth, and we're
gonna hear Nirav talk about how his management has
changed as we've gone. So why don't we just get started? - Great. - So thanks for coming,
so you want to talk about Nextdoor, and talk about
how it got started? - Sure, so I'm running this
company called Nextdoor which is a private social
network for neighborhoods, and it's got a little bit
of an unorthodox beginning because it's what we
refer to in Silicon Valley as a pivot, I was working
on another company, I was funded with that other company, I'd been working on it for two years, pretty high profile
backers, and about a year in we realized that our big
vision for that company, which was called Fanbase, it
was actually a sports company, we realized that that vision was probably not going to be realized, at least not by our implementation of the product. And so we gnashed our
teeth for about a year, and we tried lots of different things, and when it became
very, very, very obvious that we weren't gonna get there, our first impulse was
actually to return the money to our investors, we were burned out, we were pretty defeated mentally. - [John] Well, you knew
your investors pretty well, so you might want to give
that back to them, too. - Yeah, I was lucky enough to be funded by Benchmark Capital, and the
guy there, Bill Gurley, who I've worked with since 1999, when we started Fanbase it was 2008, and I'd worked with Bill for
almost a decade at that point. And this was our second
go-round, and so I had a high degree of trust and
a high degree of support. It's a real privilege
to work with guys like Bill and Benchmark, and I felt like, to some extent, we'd let them down. And we'd gotten a high
valuation, we'd raised a lot of money, we'd had some
press around the company, and it just wasn't gonna
work, and when you're in the middle of one of
these things, you realize that you're putting 110% of your life into trying to get something to work, and when it's not working, at first you say, well, this is just the
toil of an entrepreneur, and this is just what
it's gonna take, and I'm in the salt mines, and it's
never as easy as it looks on TV. But then there's the
day of reckoning, when you realize that some of your
core assumptions are wrong. And we started to realize
that, and then we tested them, and scratched on them,
and tried to iterate, and we came to a point where
we'd really hit a brick wall. I'll tell you a funny
story, I thought to myself, you know, I'm just maybe too old for this, I'm burned out, I need to
take a little bit of time off, I'm gonna give the money back, and then give this another swing,
so I called Bill and said, let's meet, I think you know
things aren't going well, let's talk about next steps, and we had this breakfast meeting in San Francisco. He brought me a poem, and the poem is "If" by Rudyard Kipling, which
is essentially a poem about becoming a man, or becoming an adult, and the implication of him
giving me this poem was, are you really ready to quit here, or are you to step up and
give this another crack? And so he basically challenged my manhood, which is an effective way
for a venture capitalist to get an entrepreneur to
go back into the salt mines. So we got back together with the team, we contracted it from 12
people to about seven, and we worked on a variety
of different ideas over the summer of 2010, and ultimately,
Nextdoor emerged from that. That was over five years
ago, and it was pretty humble beginning, and it was a beginning that was marked by failure on another idea. We feel incredibly lucky
today that our product is used by over 50% of
American neighborhoods. We've raised over 200 million in funding, our last round a year ago
was at a valuation of over a billion dollars, and so it
didn't start the orthodox way, and it didn't start smoothly,
but we're on a good path now. - Talking to Nirav when
he was doing Fanbase, it wasn't so much about how
did you tell it wasn't working, except that you could
just see Nirav and his co-founder Sarah really just
grinding at it all day long, just trying to figure
out how to make it work. - We did pitch Greylock for
Fanbase, and they said no. That was probably a
leading indicator as well. They came back, and ultimately
said yes to Nextdoor. - We said no to some
things that have worked pretty well, too, just for the record, but sometimes you get it
right, sometimes you don't. It's remarkable to me,
you've had a very, very strong base from which to work from. If you work backwards to
Stanford, even, I remember, I probably first met
you in '98 or '99, and you guys were running
Round Zero, you and Munjal. Could you just tell people
what Round Zero was, and more than that, what
the environment, like, that caused you to want
to start that, was like? - Yeah, and when I come back to Stanford, I realize just how lucky I was to end up on the West Coast, in the Bay Area, and at a place like Stanford,
because I didn't have ambition to work in technology. Both my parents were doctors, and I always thought I would be a
doctor like my parents. I wasn't even technical,
and so the fact that I've now spent my entire
career working for and running consumer internet
companies is really part of a really lucky, lucky path
that started at Stanford. And so I was at Stanford,
was not really so happy pursuing the med school
route, wondering what to do, and sort of at the end of my tenure here, the internet started to pick up, and it started to explode pretty quickly. And it exploded with names
like Netscape, and Yahoo, and Amazon, and many
that you don't know now because they don't exist, but
some that are still around. I was lucky enough to be
an early employee at Yahoo, which was mostly Stanford
folks at that point, so it was like going back to a dorm, started by a couple of
Stanford grad students. And so I was one of the
first 70 employees there, - [John] Wow, I didn't know that. - and joined in '96 just
as the company went public. It was an incredible
ride there, I didn't know what I was doing, I had no training, I was not superstar, I was probably the lowest guy on the totem
pole, but early on, watching Yahoo and watching
the success of the company, I started to think, maybe I
can be an entrepreneur, and maybe someday I can start
something like this, and Silicon Valley fuels that, and even
back then it fueled that. Of course now, it's
happening every single day. So, as part of trying
to figure out how to be an entrepreneur, a
colleague of mine and I, we were both interested
in starting companies. We said, you know, let's
network and try to meet other people in Silicon Valley
that want to start companies. We started by simply having
dinner with four or five of our friends, and talking
about entrepreneurial ideas and experiences, and just
bouncing things off one another. It went so well, because
back then if you were a junior person at one of these companies, you were in the cube farm day and night. You were in the office
working like crazy, grinding, and you had no opportunity to talk about other things externally,
back then, all of the things that we have now, all
of the incubators and TechCrunch type press, and everything else that you guys have to
tune into what's going on, that didn't exist, you were
really just inside your own company, that was the visibility you had. - Mostly engineering, like
liquidation preferences, like nodes, none of that
stuff, there was nothing. - There was no infrastructure
on how to start an internet company, because
they had just been started. There was no real history or record. And so we wanted to go out and meet other entrepreneurs that were
interested in doing it. So we had this dinner,
and it went really well, and we said, okay, let's
have an other dinner a month from now, and
maybe we'll each invite one person, and we did that,
and that went really well, and before a couple of months, we realized that there might be a
need for an organization. It wasn't formal, it wasn't a company, it wasn't really anything,
it was just a loose community of entrepreneurs that
were coming together, and within a variety of
years, because there was such a demand for entrepreneurs
talking to each other and learning from each
other, and networking with each other, within a couple of years, it became the once a month event that most entrepreneurs wanted to be at. We followed a very strict
model of invitation only invites, and you know,
speaking about the class, we carried forward a lot
of the things that we did in those first dinners, even
when we hit a point of scale, where we were having dinners
with many hundreds of people. We did the same kinds
of things to institute quality control, and so
probably at some point during that time, you were
invited to one of those dinners. The dinners, even at scale,
even with hundreds of people, the format was for
everyone to separate into small tables of five,
and so we were trying to replicate that first
dinner, even though we were at a point of scale at that point. - Yeah, it's amazing to
think that, just as a historical footnote, Nirav
did that with Naval Ravikant. - Actually, he was not part
of, one of the founders. - He was one of the
attendees, but there were four co-founders of Round Zero. - And no, that's were I met him. - He became a co-founder
of mine for a company that we started together
in 1999 called Epinions. That was just where I met him. - You want to talk about Epinions briefly? 'Cos this is not your
first rodeo at scaling. - Nextdoor is actually the third company I've started,
you heard about Fanbase, the one that didn't
work, and there was one before then, called
Epinions, and I felt after a couple of years at
Yahoo, and seeing that employee base grow from
70 people to over 3000, and really experiencing
the ride of a lifetime just being there and trying
to soak up everything that was going on, plus
running Round Zero which was what I was essentially
doing with a couple of other people, running
this thing that was scaling, that was a
little bit like a startup. It didn't have funding, it
didn't have revenue targets, it didn't need that kind of measurement. It had a lot of traction,
though, it definitely had a lot of traction, and
it felt like a startup. And it felt like something
that could accomplish great things, and change the
world, and all the things we want to do with startups,
and so at that point, I felt, it's time, even
though in 1999 Yahoo was absolutely thriving and
flourishing, and the stock was at one of its highest
points, many people thought I was an idiot to leave at the time. I left, and with a few people
that I'd met at Round Zero, started this company
called Epinions, which was an idea that was Naval's
idea, it actually was not my idea, it was a great
idea, a brilliant idea. The notion was, looking at
Amazon in those early days, one of the differentiators
on Amazon was the fact that there were product reviews
written by real consumers, which was completely different than the editorially driven product
review press that we had been used to, there
was this old publication, you guys probably don't
even know it, called Consumer Reports, and
that's where most people got their data before
they bought products. Then Amazon came along
and said, you know what, it's good to hear from actual consumers. And so, Naval had this
brilliant idea of building a kind of marketplace
or platform where people who had expertise about
products or experience with products could write reviews about those products that they
both loved and hated, and shoppers who were
looking for information could go to this platform
and get the information they needed to make
better buying decisions. - [John] And it grew quite large, right? - It was a darling of the
dot com boom, initially, it started relatively fast, in fact we had an article in the
New York Times magazine called "Instant Company"
about just how fast Epinions was scaling at the beginning. So it was blitzscaling even
before that was a term. It probably scaled a little
too fast, and at some point we were over 100 employees,
and then the dot com boom and bust happened, and
we busted along with it. We had to contract the company down to less than 20 people, so,
it was boom, it was bust, and then little by little
through some very very hard work over a number of years, it grew to be relatively successful. - [John] Yep, I think
more than relatively, you guys did very well. - It ultimately merged with
a New York based company called DealTime, and the whole thing was re-branded as
Shopping.com, which went public in 2004, and then was bought in 2005 by Ebay for over 600 million in cash. - [John] Overnight
success, after 10 years. - (laughs) You know what was interesting about that is, there was
a lot of research that was done, that I saw, I think
it was done at Stanford, that talked about the
median amount of time from founding to liquidity
event, and I think the median time was something
on the order of 5 1/2 years. If we think about, founded
in 1999, I think Epinions was April '99 or so, and then
public in November 2004, that tracked that median amount of time. Now I look at Nextdoor,
which was really founded the summer of 2010, and here we are, over five years, and we are not going public anytime soon, at least not that I know of. That's one of the things
that actually has changed in Silicon Valley, the
way that we think about scaling companies, we are
more ambitious than ever because the markets are
larger, but our path towards getting public and being truly independent has expanded, interesting. - [John] Yeah, I think the
popular, the conventional wisdom is that it happens fast, 'cos you don't hear about this company, and it shows up in the news, and something that's got billionaires
minted, or whatever, I think the median time to liquidity
has gone up. (unintelligible) - Yeah, and I'll say a
couple things about speed and the phenomenon that
we see, two things. One is, there are companies
that are almost overnight successes, and as an
example I'll say Instagram. Instagram was not an
overnight success because Kevin, the founder of
Instagram, had started something before, but
that product, Instagram, was almost instantly
well regarded, and it was almost instantly one of the
top apps on the App Store. Everyone is looking
for one of those, where you literally have lightning in a bottle. You just happen to have
created something that a lot of people really love, so you want
to have one of those things. The other thing I'd mention about speed is sometimes, fast up can mean fast down. The thing you have to
be a little careful of is, there are lots of
things that get popular really, really, really
quickly, and then they lose that popularity, and they come down almost as quickly as they went up. A sector in general,
not naming any companies but a sector in general that was like that was games, and specifically,
games on Facebook. There was a period of
time when you were getting cows and other things
thrown at you on Facebook, or whatever people were doing in the feed, and it was the main thing that you saw on Facebook, and now you don't see it at all, all those things are gone. So Facebook as a distribution platform for gaming companies has dried up as quickly as it grew into the dominant platform. - [Student] If there's another
bust, what do you think the things that founders
who need to contract their companies are
most likely to mess up? - I don't know that there
will be another bust, but I do believe pretty
strongly in cycles, and I think that we are
at the height, or we have for some time, been at
the height of a cycle that feels like it needs to
come down a little bit. I think probably the
biggest mistake we've made this time around is, we
have once again confused the coolness of an idea with
the true market potential. People are starting things
everywhere, and I think that's great, when they go
out and get venture funding, particularly at the
valuations that you're seeing, there's an implication that whatever that funding valuation was,
the enterprise value of the company will
eventually be worth 10 or 100 times whatever that thing
is, and in most cases, a lot of these companies
are really good companies that could be run by a person who funds it himself or herself, I
think that's been one of the big changes in the internet
with Amazon Web Services and a lot of the tools
that exist, you can be an entrepreneur that's
self-funded or angel funded and then start a cash
flow business, and I feel like a lot of these
venture funded things are essentially cash flow
businesses, they're not bad businesses, but they're not great venture funded businesses, and so when you start to realize, you
know, my actual market is a $100 million market, but I raised money at $100 million valuation,
and I need to get to a billion, you start to do
a lot of unnatural things. I think that's where we're gonna see a lot of trouble for companies. - You usually use venture
because you feel like you need to fund growth
to get network effects that become unassailable over time, which, I think will lead to
the conversation we're about to have in just a
second, I think using venture to fund a cash flow business
is almost never right. You start to think about
growth ahead of revenue, you start to think about
upside down unit economics, it's a whole bunch of problems. - Yeah, specifically, what
people are saying is using venture money to acquire
customers or acquire sales is the wrong way to use
venture funding, unless you believe there's a true network
in there on the other side. - [John] And then, I
think what happens in a contraction is that most
people say, well look, I'm a small cap company,
I'm worth $100 million or $200 million or $10
million, I'm not worth a billion, or a trillion, or
$10 billion, or $50 billion. It won't matter if the top
of the market contracts. I think what surprises people is how fast it kind of cascades down, you start to see tightening at the top and then it tightens down pretty quickly. - I guess, is there a
better way to reducing way down from like 100
to 20? (unintelligible) - Well, you're doing that
because it's the only thing you can do to stay in business. I think, in general, what
founders and CEOs need to think about is,
there's a right way to run a company, and that way
doesn't change that much in a boom cycle or a bust
cycle, it's making sure that every head is
extremely valuable, and that you're not growing too fast
from a head count standpoint. It's making sure that your assumptions are constantly being
re-proven, it's making sure that you have a long
term strategy, and you're taking all the small
steps necessary to get to that strategy, a lot
of the things that you do in a bust are really
good things to do during a boom as well, I think
that's what the companies that survived the bust last
time, that's what they learned. They learned that the
way to operate a company, or the way that we were
forced to operate companies in the bust, are pretty good ways to operate the company even in a boom. - [John] Arright, so
let's move to Nextdoor. Give us a sense of
Nextdoor today, talk about, maybe, let's start with
the products, want to talk a bit about what the product is? - The premise of the product
is that while we have social networks to easily
connect with friends and family, and that's
Facebook, and to people that we're interested
in, and that's Twitter, and finally to our
colleagues, that's Linkedin, there isn't a social
network that will help us easily connect and
communicate with the people right outside our front
doors, and the people right outside our front
doors are typically our neighbors, the people who live in our neighborhoods and our local communities. It's a little different
for college kids, because your neighborhood really
is Stanford, and so Facebook is a perfect
application, or whatever, Instagram, Snapchat, whatever's
being used on campus. But once you leave Stanford,
you move to a new place, you won't know the people around you. When we started the
company in 2010, we were really shocked by this
statistic, which is almost 30% of Americans couldn't name
a single neighbor by name. John and I will tell you, - [John] 30%. - 30% of Americans
could not name a single, it's actually 28% of Americans could not name a single neighbor by name. John and I know from growing up in Texas, that's not the way we grew up, we weren't necessarily best friends
with our neighbors, but we certainly knew them,
and we relied on them, and I think, if you think
about your parents back home wherever home is, they
are a part of a local community that they feel connected to. When I think about my two young sons, both under three years
of age, I don't want them to grow up in a world
where most of their friends are folks that they're
connecting to virtually on Facebook, or through
Twitter, that they've never met. I want them to feel like they have roots in the community, and
so that was the premise of the company, it was
a pretty obvious premise but completely unproven, because people had lost touch with neighborhoods. So the notion that there's
even any value to connecting with our neighbors, that
was an unproven notion. Now here we are five years later, we have over 80,000 neighborhoods that are using Nextdoor, that's roughly 50%. We think there are 165,000
neighborhoods in America, the median size of those neighborhoods
is about 700 households. - [John] I think one of
the stats, because I think it will be good to give people a sense of the scale we're talking about. - And while he's bringing
this up, almost 50% of American neighborhoods, but over 90% of the neighborhoods in the top 40 cities, so, in almost every major
city, if you want to connect with your
neighbors using technology, you're gonna use Nextdoor,
and what's interesting about Nextdoor, particularly
in the context of this class, is that it was
not an overnight success, and it's still not an overnight success. It was a very long, hard, road, continues to be, but it's picking up steam. We've got a couple of slides
here that we're bringing up, and I'll talk you through
these, but the basic notion is that when we
started the company, because this thing Fanbase
that we'd been working on, because it hadn't worked, and with Fanbase we had this grand
ambition of, can we create bottoms up community,
to create a competitor for ESPN, that was the grand vision for Fanbase, we didn't get anywhere. When we started Nextdoor, it was much less about the grand vision,
now Halloween is a very neighborly holiday, so
we've got the Halloween faces here, I'm not sure if
there's any trick or treating going on here at Stanford, but, you know, once you get out you'll realize
there is trick or treating. But when we started
Nextdoor, we were probably a lot less ambitious with
our long term vision, because we'd just come off a failure. What we were very focused on was, can we just make this thing work in a small set of neighborhoods, so we can get a proof of concept, what people commonly refer to as product/market fit. So now we can go to the next slide. I'm gonna go back as
far as 2013, of course we started the company even before then, and you can see, if we
look at a couple of our key metrics, let's think
of them as almost growth, and engagement, and then
finally, external validation, you see a couple of things,
you see tremendous growth, but look at the scale of growth recently. And so, this is blitzscaling,
from a metric standpoint. This is John's favorite essay,
"Do Things that Don't Scale" at the beginning, and if
you get those things right, - [John] Can you talk
about how you do that? - Yeah, if you go to the next slide, I think I have it in the next slide. Our first year of Nextdoor, we had 176 neighborhoods. It took us an entire year
to get 176 neighborhoods to use the service, and
my co-founders and I, we talked to most of
the people that started every one of these neighborhoods. To start a neighborhood on Nextdoor, you have to put in your
address, you have to draw the boundaries of the neighborhood, you have to name the neighborhood, and then, in some period
of time, 14 to 21 days, you have to get nine of your neighbors to join, and all verify their addresses. So, starting a neighborhood is not easy. If you think about, 30%
of Americans don't know a single neighbor, we
were looking for those neighbors that knew at least nine others. That's how we could
create that critical mass. - [John] This is a
little counterintuitive, because if you didn't get that done, you'd send a note and say,
good try, we're going to turn off your neighborhood for a while. - If you started a
neighborhood and you didn't have 10 people, it was called a pilot, and the pilot would expire,
originally in 14 days and then eventually we
extended it to 21 days, and of the 80,000 neighborhoods
we have, we probably have another 80 to 90,000
pilots that have expired. It was a quality control mechanism, almost a game mechanic, because
we said to ourselves, look, when we're building these neighborhoods, we need to build the right foundation. If someone comes along,
like John, and says, Oh, I'm gonna use this
for my own personal gain, I'm gonna call this John
and the Boys in the Hood, I'm gonna draw a boundary
that's not very sensible but that makes a lot of sense for me, and I'm just gonna invite
my friends that agree with everything that I
say, well, that wouldn't make for a very useful neighborhood. The people who typically know nine or more of their neighbors are
the folks who care deeply about their local
communities, they're the ones who care about Neighborhood
Watch, they're the ones who care about neighbors
getting to know each other, they're the ones who
welcome the new neighbors that move into the
neighborhood, they're the ones that act as the social
glue, and those are the ones that we were seeking, but
it was very counterintuitive because, if you think
about it, it takes us an entire year to get 176 neighborhoods, and we know that there are
165,000 in the country, well, you can do the
math, how many years would it take us, at that run
rate, to get penetration? It would take us a lifetime, right. But we were focused on something different than establishing the run rate for growth. We were focused on quality, and that was very, very important for us,
if we go to the next slide, - [John] Wait, so one
thing, you keep using this word neighborhood,
which I think we all think we understand in our heads, but, neighborhood's not at all an obvious thing when you try to draw the lines, right? So you almost needed to figure
out what they meant, right? - Yeah, there's sort of
an interesting fork that we were confronted with
early on in the genesis of this idea, and that
fork was, do we try to model ourselves around Twitter? The notion of Twitter
is, you can tune into conversations based on who you like, and who you want to follow, and it's not even symmetric in terms of you
needing to get approval to follow someone's tweets, and we said, if Nextdoor were like
that, you would just put in an address, and then
you could tune into all the conversations going
on in some geographic radius or proximity to that address. And you could change
that, like if I'm coming down here, I could say, show
me all the conversations that neighbors are having
in Palo Alto, as an example. The Facebook way,
particularly the way that Mark started Facebook if you remember, you had to have a university email address to join the Facebook for that university. I had to use my Stanford
alumni address to join stanford.facebook.com, and
that was very different, because it made me feel like
I was part of a community. It was harder to get
critical mass, because it was just people who
had gone to Stanford, or were at Stanford, that
were part of that community, but it made me feel some identity. And so we decided early
on, every one of these neighborhoods is gonna
have a discrete boundary. That meant we needed to
build a geospatial platform in addition to a social
networking platform. We decided that every one
of these neighborhoods is gonna have a distinct
name, and every one of these neighborhoods needs
to have identity behind it, because when you ask
someone where they grew up, tell me again the city where
you grew up, San Antonio? - [John] San Antonio. - So he grew up in San Antonio, but if he met someone in San Antonio, he would say, this is the neighborhood I grew up in, whatever it was. - Well, in fact I did, I
sat with Josh earlier, and we talked about how my wife
grew up in Alamo Heights, and which part of San Antonio,
yeah, that's exactly right. - And so, it's a very
important part, in fact, the interesting sociological
experiment we did, is we said, when you meet
someone for the first time, say, at a party, what are
the things you talk about? Well, I go up to John,
and I say, Hey John, nice to meet you, why are you here? And he says, Oh, this is Peter's party, and Peter's a friend
of mine, and I say, Oh, Peter's a friend of mine,
too, that's sort of Facebook. Now I notice that he's
wearing a Stanford shirt. And I say, Oh, you like Stanford? And he says, Yeah, I went to Stanford. And I say, Oh, I went
to Stanford, too, and that's a little bit of Facebook,
a little bit of Twitter. Then, the next thing, if
you're in Silicon Valley, is you say, Where do you
work, what do you do? That's always a part of your identity. And he says, Oh, I'm a venture capitalist. I say, Oh, interesting,
that's sort of Linkedin. But almost every single
time when you meet someone, and think back to when you met people your freshman year and when
you meet people now, you say, Where are you from, right? And the person says, Oh, I'm from Texas. Where in Texas, this little
city in west Texas, Odessa, where in Odessa, and
it just keeps tunneling all the way down until you
typically hit the neighborhood. And that's a big part of
knowing who someone is. We felt that notion of identity
was extremely important. It was the hardest decision we made for a couple of different
reasons, one is, there is no objective encyclopedia of
neighborhood boundaries. People argue about neighborhood
boundaries all the time. We needed to create a
Wikipedia type system where we would have the community create a set of boundaries, that was hard. The second thing we said
is we're gonna make the conversations private, unless
you live in the neighborhood. Everyone who joins Nextdoor
has to verify where they live. Very counterintuitive, people talk about reducing friction, we increase friction. Finally, we said, there isn't
even gonna be a neighborhood unless 10 neighbors can come
together in 21 days, question? - How did you find these
people who are so integrated into their community,
like the 170 individuals across the country that
you were working with? - We started in the Bay Area,
and we started by asking all of our friends, it was basically
like a game of Telephone. We would ask all of our
friends, do you know anyone who's really involved in the community? And they would say, well,
I know these people, and we'd talk to them,
and if they were people, then we'd say, Can you
start it in your community, if they weren't, we'd
say, Do you know someone, and it would just feather out that way. At this point, and this is
really one of the lessons, I think, around scaling,
all of those things that we did manually, we
initially manually verified people's addresses, we would say, Can you send me a picture
of your driver's license? Or, can we get a Homeowner's Association roster, and go through there one by one, and get people to be verified? We would go door to door
to get people to join. There are all these
things we did completely manually, where we rolled up our sleeves, and it was very dirty
work, and it was never going to be scalable, and we knew it. We were just looking
for signal in the noise. Then systematically,
since then, we have made those unscalable things
scalable with technology. That's what has enabled that growth. Systematically, we've taken
verification, made it automated. We've taken neighborhood boundaries, the whole thing's automated. Now, if you want to join
Nextdoor, you come in, you put in your address, you're put into an automated application
that then dumps you into the pilot program which
is completely automated, all the reminders in the early
days, my co-founder Sarah would call these neighborhoods
that were getting started, that only had five people,
and she'd call up the person and say, C'mon, you can
do it, you can do it! Just invite one more person,
they might know another person. Now, we have all those hooks
built into the platform. So, we just looked for
signal, we looked for fit, we looked for some
fidelity, and some notion that what we built was
real we didn't worry about how are we gonna scale this, we had to worry about that next,
and we've been worrying about it ever since, but
initally, we focused on quality. And as we think about
the organization today, back then it was seven
people, now it's over 100, we try to do the same thing today. We actually separate the company, we have organizations like
engineering, product, design, et cetera, but we sit
people, typically, in project groups where we have a designer, a product person, five or
six engineers, an analyst, and we tell them, Let's try to simulate what Nextdoor was like in the early days. It's okay if you do scalable things. To take an example, recently
we launched a feature on Nextdoor, which is finding
a babysitter on demand. One of the things that you
do in your neighborhood all the time, is you ask for a babysitter. So wouldn't it be cool if
you could find a babysitter as easily as you can find a car on Uber? - [John] You guys all have
this problem all the time. - Well, you guys could be the supply side if you wanted, not the demand side. When we piloted this,
which we've been piloting in Palo Alto and a couple
of other neighborhoods, we had the team members go and literally walk around Palo Alto to try to find babysitters to sign up for the service. Clearly, that's not the
way we're gonna do it long term, but we got supply to a point where we could test
demand, demand and supply were matched, and now
we're looking at ways how can we actually productize this. - [John] I'm struck by
all the similarities, there's so many similarities between Nextdoor and Linkedin,
and a lot of different points in the thing, probably it's because by the end, you're gonna have this asset that looks like Linkedin,
which is very, very hard to assail (drowned out by Nirav). - Well, let me show you some of the data, because, I think, the data is
very interesting to look at. This is where we are now, all
50 states, we've got Alaska and Hawaii in there, go
to the next slide please, we've made it a goal to
get highly penetrated in the major cities,
because as we think about building a competitive moat
against any competitor, we want to own the most
valuable cities in America, and these are some of them, and these are the percentages of
neighborhoods in these cities. You'll also see something
interesting, which is Nextdoor is a true mass market phenomenon, it's not a Silicon Valley phenomenon. You're gonna see cities all
over the map here, literally. You got Manhattan, Los Angeles, Houston, Washington D.C., go to the next one. Now, to get a little bit more interesting, this is engagement, how frequently do people use the product,
and the solid line is, how frequently do our members, and the denominator's all registered
members, so this is everyone who's ever
signed up for Nextdoor, 41% is actually the jagged
line, that's the percentage of our members who use the
service every other day. It's remarkably consistent,
the two dips are Thanksgiving and Christmas, and what's interesting about it is,
the reason it's consistent is because, while we've
had to work to scale the number of neighborhoods,
we have not had to work to scale the
quality of the experience, because the quality of the experience was figured out in those early days. Now, we just need more
neighborhoods using this thing. - [John] Allen said that
too, Allen Blue said, we're patient, we're content to go slow, until we figure out what
made people really love it, and then you figure out
ways to put it out there. - There's a time to
actually step on the gas, and that time is when you're
sure of where you're going, and you can be sure of where you're going you're going when your users tell you that they love the product, go to the next one. - [John] Every other day
usage is a pretty good signal. - The ultimate goal for
consumer internet product is what Google calls the toothbrush test. We use our toothbrush once a day, hopefully some of us twice a day, and so can you create a product
that people are using every single day, that's
the dream, question? - [Student] Going back a
little bit, you talked about how you piloted these
170 neighborhoods to see if there was any signal,
but how did you know these neighborhoods were representative of the rest of the U.S.
(drowned out by Nirav)? - That's a really, really good question. We worked hard to try
to get a cross section of geographic diversity,
socioeconomic diversity, we wanted rural and urban and suburban, we wanted senior folks
and young folks, and what was pretty interesting
is, the use cases did have some variation from
neighborhood to neighborhood. In some neighborhoods, people
were looking for babysitters. In the Marina in San
Francisco, which is where many people go after they
graduate from Stanford, no one's looking for a
babysitter, but they do want to know what's
going on on Friday night. So it's a different kind of
conversation that you have with your neighbors, and
senior citizen communities in Florida, there are
neighbors that want to actually do bridge nights
and solitaire nights, and that sort of thing, so
the conversation differed, but the quality of engagement
and the basic human need to connect with people around me, that was incredibly consistent,
but it's a good question, because we did, because
we were coming off of a failed startup, have a lot
of skepticism about ourselves. So we did the first one, we said, well, this is just in Menlo
Park, anything would work in Menlo Park, so then we did another one, (John laughs) and we said, well, that's
just Portola Valley, okay now, all we've got is a little coincidence, Menlo Park
and Portola Valley. So we said, let's force ourselves to go outside of California,
so we went to Seattle. And then it worked in
Seattle, and we said, Well, maybe that's still
just a tech kind of thing. Ultimately, we made our way in that first 175 neighborhoods to Hamilton, New York, which is about as different a neighborhood as we've probably ever
lived in, because we would see postings in that
neighborhood of things like, I've lost my cow, my cow is wandering around the neighborhood, can a neighbor please tell me if they've seen my cow. That's not what you see
in either Menlo Park, or Portola Valley, or any
San Francisco neighborhood. But the conversation was still rich there, and it was still
something that we felt had very strong signal, so
we again, systematically, sometimes failure can
be a very very powerful motivational tool, almost always, I think. We had this spectre of, yeah, but is it really working, is it really working? My wife asked me, when we
crossed 80,000 neighborhoods, which was just this week, she said, "Can you believe it,
80,000 neighborhoods!" I remember when you told the company, that first summer, if we get to 100 neighborhoods, we'll
bring a dim sum cart into the company and give everyone dim sum, and we'll call it the Dum Sum 100. And our engineers looked
at me like I was crazy. We'll never get 100 neighborhoods
in 90 days, that's crazy! We got to 100, and I
remember telling them, if we get to 10,000, we're
gonna do the Tuscany 10,000. Well, we never went to
Tuscany, because we're scaling a lot more quickly now,
but even when my wife said, (laughter) we've run out of money,
even though we've raised $200 million, we can't take
all the employees to Tuscany. I did tell, I think we
said, we do have some goal about founders shaving
their heads if we hit some extreme number of
million, double digit millions users, or something like that. The idea was, and still
is, we haven't made it. We're working hard every single day, and we're motivated by the fact that every single day, we get some
positive data points, and we're just gonna keep going. There's this old quote,
I'm not sure, it may be like a Sun Tzu quote,
but the journey of a thousand miles begins with a single step or something, I'm not sure
who it's attributed to, - [Student] Mao. - Okay, so it's a Mao quote,
sorry, I was a little off. (laughter) Don't quite know my Asian historians. - [John] Very super similar, oh yeah. - But that is very true
for us, it's not true for all businesses, by the way. It's not true for all businesses, but for us, it's absolutely true. The only way for us to get
there, and get there for us is, we want to be used by every neighbor, in every neighborhood,
every day, around the world. - [John] This story you told
about, maybe this is just Palo Alto, or maybe it's just
Portola Valley, whatever, it reminds me of my very, very
favorite thing that anybody in Silicon Valley has ever
said, it's Andy Grove saying "Fight like you're right, and
listen like you're wrong." What he meant is that
he wanted everybody to come out of the strong
product point of view about what's right, and
he wanted them to be relentless about looking
for data to figure out whether it was wrong or not. - The interesting thing
about data is, your instinct and your intuition can be wrong. Data, if it's interpreted correctly, is very, very rarely wrong,
and so what you see up here is what I commonly refer to as one of my favorite slides,
what this shows here is a very systematic,
consistent, predictable growth across neighborhoods, growth
in size of members joining. What we've done here is we've cohorted the total number of neighborhoods we have, which is changing over time because we're adding more neighborhoods,
we've cohorted them and we've said, What
percentage of the entire cohort corresponds to what size of
members in a neighborhood? You go back about 2 1/2
years ago, Q1 of 2014, almost 60% of our neighborhoods had less than 100 neighbors using Nextdoor We were just getting
started, it's literally like growing a tree, the
first couple years you grow a tree you don't see anything. But if you go to Muir Woods and you see some of those incredible
trees, those are trees that have been around for a long time. So what you see here,
with the green slowly but surely, systematically
growing, is that all of our neighborhoods
get larger over time. The denominator here, 2013, is 7000. Over in 2015, it's 80,000. So all these neighborhoods
are getting bigger. Let's to to the next slide, I think? - [John] Yeah, Dropbox had a, I love this idea, because you're getting every cohort getting stronger over time. Dropbox had a chart
like this which is every cohort is making more money
than the month before. - Do you have increasing marginal utility? - [John] That's right,
when you start stacking these things, you get
a very, very fast ramp. - You know you have a
network business, I mean, what's the definition of a network effect? Every node that's added adds
value to the entire network. That's a little bit what you're seeing. I'll just show you this, and
then I'll take some questions. This is pretty interesting,
this is looking at household penetration by neighborhood, and the neighborhoods that
have launched on Nextdoor cohorted by their launch year date. The green line here is
all of our neighborhoods that were launched in 2012, so these are neighborhoods that are three years old, and they
are now 30% penetrated. So, 30% of the people
in those neighborhoods, and it's the entire
basket of neighborhoods, is using Nextdoor, what's
interesting is that the next year, 2013, which is the orange line, it's closely tracking. We're doing a little better, but it's very closely tracking 2012. Same for 2014, it's very
closely tracking 2013 and 2012, and then you look at
2015, it's about the same. So what I told John is,
if you see this kind of predictability, and you can
take a step back as a CEO and say, okay, as long as
this continues, I know that today, we have 50% of the
neighborhoods in the country. I also know that at some point
in time, that's years away, not many years away, all
of those neighborhoods will have at least 30%
penetration, and more, actually, because it doesn't
plateau, it keeps going. What that essentially says
for me is, and the board, this is what VC's do, they saw this chart and they said, Sir, are you telling me we could fire everyone in the company? (John laughs) And at some point, in
the next three years, 50% of the neighborhoods in America would have 30% of their population on Nextdoor? That's not actually true,
you couldn't fire everyone, you'd have to keep some of the engineers to keep the site up,
but it's largely true. Which is awesome, not the firing part, but it's awesome from a
predictability part because a lot of businesses that
we see are what I would refer to as treadmill businesses. Every morning you wake up, and you gotta go take a run on the treadmill. Those miles that you're
accruing, that distance that you're travelling, you don't
get any credit for it tomorrow. It's just one day, you do
everything you can, and then the next day, you've gotta
earn the whole thing again. The really good networks
businesses that accrue value over time, it's
not a treadmill business. Every single step you take is adding incremental value to the whole. That is Linkedin, that is
the definitional quality of Linkedin, and it's something that we hope is gonna be true of Nextdoor. - [John] I see a question from over here. - [Student] So how are you gonna keep people on the platform once they've made the connection there,
and (too quiet to hear) - We should repeat the
question, (unintelligible). - So, sorry, how do we
keep people on the platform once they're there, and
then the second part? - [Student] Right, so because
there's a lot of cases where, you know, people go on the platform and make the connection, (not clear) - You know, we've been pretty lucky. When you have a platform that the average of all users is that 40% of those users are using the platform every other day, there's something that's so
powerful in the core engagement of being able to tune in
to what's going on in your neighborhood, and there
really is no substitute. That's not actually a problem
that we think about that much. We don't have a lot of
churn, we don't have a lot of people who use
the platform a lot and then go off and decay over
time, which sometimes happens in gaming, you use the thing a lot and then you're like,
I'm sick of it, I'm done with Farmville, I can't do it again. What we see, on the other
hand, is the opposite. People, they go in a little
carefully, and then over time, as more neighbors join
and as the information is richer and as their
utility actually delivers more value, they use the product more. I didn't show you a cohort
where we take our neighborhoods and we say, Which
neighborhoods have the highest engagement, and what you would see is the neighborhoods that are
the largest have the highest engagement, which is also great. 'Cos all the neighborhoods
are getting larger. The one other thing I'd
say is, we're very lucky in that there isn't an obvious substitute. John and I were talking about
it, there is no internet competitor, really, nor is
there an offline competitor. It used to be, when I was at Stanford, I actually subscribed to the Chronicle, as a student at Stanford,
a lot of people did. - [John] Is that a newspaper? - That is a newspaper. (laughter) Dead trees, the whole thing. - [John] Words on paper. - And so, I would get informtion
about the local community. I'd read the Daily, I'd read
the Chronicle, and that was it. I wouldn't go to CNN.com or
whatever, I'd maybe watch the news, but that's where
I'd get the information. Today, I don't know anyone
that reads a newspaper. So, how do you tune in to what's going on around you, we have to crowdsource that. - [Student] You said you
had 40% retention rate now? Do you remember how it was for the first neighborhood in Menlo Park? - Incredibly consistent,
and you can actually, if you go back one slide,
you'll be able to see it, because, one more slide,
what's interesting here is, this goes all the
way back to June of 2014, and it's the same basic average. What that means is, back then
we had incredible retention, and today we do, now I always tell people, two sides to the coin, the
good news is, 40% are using it every other day, the bad
news is it's not 60%. It turns out that when
you slice the cohorts a little more finely, you
find that the more established neighborhoods, which are
older and larger, they do have increased engagement
rates, so it does get better. - [John] As a product thinker,
when you look at that chart, it's nice that you have it,
'cos it provides you a base, does it seem hard to increase it? As a product thinker, you
said, maybe there's nothing we can do in the product,
it's just natural. - I would say, and the
company would agree with me, our product is very
coarse, it's not a well refined product, and
the reason for that is, we're actually not a product company. We're a community
company, and so the thing that we've focused on, and
the thing that we've built and the thing that we obsessed about
all the time, is community, and community is why we
have 80,000 neighborhoods. The product that services
those communities, there's a lot of upside, a ton of upside. All we have to do is
look at the neighborhoods that have the most penetration to see rates that are much higher than 41%. It's really about
getting all neighborhoods to be like the best neighborhoods. - [Student] What are your
biggest growth bottlenecks, and how did you figure out what they were? - We have a core growth
bottleneck that is, it's sort of an interesting
conceit, because the bottleneck is also the
reason for the service. Social networks typically
grow through virality, and virality is, I invite
you to the service because I'm gonna get something
out of inviting you. Google is not a viral
service, you tell people about Google because you just like
it, it's word of mouth driven. But Linkedin, you send
a connection request to someone because you want to be connected to them, so we're the latter. Interestingly, though, because Americans don't know their neighbors, they have no mechanisms to invite their neighbors. The reason it's an interesting conceit is, if they did know their
neighbors, there would be no reason for Nextdoor to exist. The reason that growth
is so difficult for us is because of the value
proposition we're trying to create, which is, people
don't know their neighbors, so they can't easily invite them. If you think about Snapchat,
and Snapchat's growth, and we think about
Instagram's growth as well, the social graph is something that Facebook doesn't have a monopoly on. All of us who have phones
are carrying around our own social graph on our phone. It's not actually unique to Facebook. So, when a new service
starts, if it's a service that relies on your
friends, you can typically hit one button, and bang,
that whole thing is populated. There is no easy way for
us to go neighborhood by neighborhood and
populate the whole thing. We have to find one neighbor
who knows two neighbors, and those two neighbors
know two neighbors, so we sort of systematically
grow that way. We accelerated or heightened
our growth problem by making a decision early
on that our conversations would be private, so that means no SEO, no social sharing, so both the virality and the discoverability of Nextdoor, both of those things
are severely hampered. Which, is why, if he looked
at this company before, and we were just sort
of white boarding it, he'd say, you've picked
a very, very difficult execution path, because most
of these things will grow either because they're
intrinsically viral, or because they're easily discoverable. We've actually done the
opposite in both cases. Now, how do we fight those things? We've done some pretty unique things. On the invitation side,
we have many members, our primary way of growing is neighbors invite their neighbors
to the service, but if they don't know their
neighbors, they don't know their email addresses, how do they do it? We pay for them to look at
a map of their neighbors, and click on the houses
they want to invite, and then we will actually send
a postcard on their behalf. - [John] You'll still do that? - We will send, this year,
many millions of dollars of postcards, I think on the
order of 20 to 25,000 a day. The nice thing is, every
one of these postcards, one of our members has said, I want to invite that neighbor to the service. In the early days, you'd
see all these tweets where someone would say, Wow,
this is really old school, and they'd be showing a picture
of a postcard from Nextdoor. It's like, do you really use
a postcard to get someone to join an online service,
right, a little counterintuitive. - I wanted to talk about
nearby neighborhoods, that was a (Nirov's voice drowns him out) - One of the things you
also want to look at is that virality, is there a way to increase the viral vector that
you could possibly have? Because, if I say to you,
I live in San Francisco, you live in Palo Alto, there's
no reason for me to invite you to Nextdoor because
we're never gonna interact. But maybe if we can connect
Palo Alto to Menlo Park in some way, 'cos there
are some conversations, the neighbors in Palo
Alto and the neighbors in Menlo Park may want
to have, classifieds, for example, where you
have an extended market. We came up with a feature
called Nearby Neighborhoods, where today, we not only connect you with your immediate neighbors, we allow you to also communicate with about the 10,000 closest households
in other neighborhoods. Those can be contiguous neighborhoods, that are on your border, or they can be neighborhoods that are not exactly on your border, but very close to you. That did two things, it created a better viral vector for growth,
but the other thing it did, is, it increased the
utility of the service, because, for example,
if you have a lost dog, and you want to tell your
neighbors you have a lost dog, your dog doesn't know
that he's now leaving your neighborhood, and so no one else is going to be able to find him, because he's going to the next neighborhood over. We had many people who said, the utility of the service will be
better if I can communicate with a broader set of
people, and there were other people who said, I
live right on a boundary, I feel like I'm really
part of both of these neighborhoods, and how can you proxy that? So it wasn't just a
growth driver, it was a quality of experience driver, and I think old school internet is the best growth driver, is a great product. You hear a lot about growth hacking and that sort of thing, but ultimately,
for a service like ours, every time we do something
that really delights our members, we see a puff of growth. - [John] It seems to me, it
also increased engagement a little bit, so it increased the density of the people who were around you? - It got to the point
where, so, as an example, the number one use case on Nextdoor is, I'm looking for a service
provider, so it could be a babysitter, or a
painter, or a contractor, a doctor, dentist, et
cetera, and that's 26% of our 11 or 12 million messages a day. Those typically have
about 15 to 20 minutes before you get your first
response from a neighbor. Over a 24 hour period, you typically will get a median of 5 1/2 replies. That's pretty incredible,
if you think about it, there's a platform out
there where you can ask your neighbors for a
service provider, and within half an hour, you'll
get one answer, and in the next 23 1/2 hours,
you're gonna get four more. That's pretty powerful,
but we needed to increase the density of people that
could respond to that query, and that's something
Nearby Neighborhoods did. - [John] I'm just gonna
put this thing on autoplay for the rest of the time,
there's plenty of questions. - [student] How did you think
about revenue (unintelligible) - Good question! Money! - How am I thinking about revenue? - Yeah, how'd you solve
the money equation? - I'm not sure if Reid
has copped to this, but in the early days, when
we were starting a company and I was leaning on
Reid heavily for advice, because he's built something like this, or something that we aspire
to be, to follow in the footsteps of, he said, one
way I think about it is, you build user scale,
you build usage scale, and then you build revenue scale. It's this notion that,
when building a platform with indirect monetization,
which is to say, we will not charge our members directly, we need to monetize through
ads, or some services that we provide them,
or something like that, the path that you do
those things in, because each one of these problems, growth, engagement, and monetization, each one of these problems are very difficult. The only way to solve them is to break them into sequential steps. The first thing you do is you figure out a reason for people to simply join. Once you've figured that
out, you can move on to a reason for people
to stay on the service once they've joined,
and use it, once you've figured that out, then you
can move to monetization. That's actually the path
that Twitter and Facebook have followed, I think
both of those companies were probably four to
six years in before they started to truly monetize,
and that's where we are today. 2016 is a really big year
for us, I have an offsite tomorrow morning, where,
as a management team, we're gonna talk about
2016 and how it's really gonna be a great year
for monetization for us. We feel like the platform from
a user and usage standpoint is at scale, so now it's
time to capture value, yes. - [Student] What product insights can you derive from the continued
usage of Craigslist, given that it's still the same? (John laughs) - That's a great, great
question, I think marketplaces are typically driven by
liquidity over product experience, that's the number one. - That's the question, was,
more or less, what can you figure out because
Craigslist is so strong. - Or, said a different
way, why is Craigslist still around when a thousand companies have tried to take it down, and from a superficial standpoint,
it hasn't, there's no mobile app that's great for Craigslist, it's not intrinsically
built for a mobile world, all these things that you
see in great products. I think, ultimately, you
go to buy things where the sellers are, and you go to sell things where the buyers are, and
Craigslist still has a lock in a couple of categories
like real estate, and personals, and others where it's still the best marketplace out there. We have a slightly different
take on it, because Craigslist is a marketplace where you can go there anonymously, and you
can buy and sell anonymously. That's not the way Nextdoor works. Nextdoor, you have to verify your address, you have to put your real name in. Classifieds, for us and
Craigslist, that's not the direct competitor, we don't believe that there's a lot of
business value for us there. We think they're doing a pretty good job, and we don't have a way
of getting that kind of scale because of our model. I think, for other startups
that are going after Craigslist, it's not about creating a prettier UI, it's about getting those
audiences to shift, and that is very, very hard to do. - Marketplaces are all about whether the transaction clears or not, if you put something up for sale,
you want to make sure it sells, and nobody has liquidity like Craigslist, it's brutal to beat them. - Sure. - [Student] (too quiet) - Yeah, that's a great question,
so, I use this framework because it has been very
helpful in the companies that I've been part of,
which is, at any given time, you have five objectives in a company. You have growth, you have engagement, you have monetization,
you have infrastructure, and you have people, so those are five distinct things that you can be worrying about at any given time. If you think about your total resources, or your total focus, as a pie
chart of all of those things, you have to figure out,
what are the size of the slices at any given point in time. Sometimes, that point in
time is annual, sometimes it can be quarterly,
sometimes it's even briefer. If you have an outage, for example. Interestingly, a lot
of these services that get scale quickly, they go from spending no resources on infrastructure,
and we did that, to spending a lot of
resources on infrastructure. For us, if I go a little forward, 2014, towards the end of 2013, 2013 was all about growth for us, and we haven't put any percentage of the
resources on monetization for a long time, now
we're about to get there. We've actually put
surprisingly few resources on engagement, because in
the first instantiation, 2011, we spent a lot of time,
176 neighborhoods, we were not focused on growth, we
were focused on engagement. The next year, 2012,
was more about growth. 2013 was, again, about
growth, and then 2014, we needed to actually invest in the service, because it was coming down. We couldn't handle the
number of concurrent messages that were being
sent, and that sort of thing. Again, as we move forward to 2016, I am thinking along
with the management team about, across these five
things, what's gonna get dialed up, what's
gonna get dialed down. The pie itself gets
bigger, because you raise more money, you hire more people, you have more resources, assuming that things are going
well, but ultimately, those are the five knobs that
I believe we can control. As we look toward 2016, 'cos my memory, I probably don't remember
2012 at this point, it's been overwritten, I have
limited hard drive space, 2016 certainly we're
gonna dial up monetization quite a bit, growth
for us in 2016 is about international expansion, so we're not necessarily dialing growth down, but we're directing it towards a different target, whereas we've been domestic to date. And then, I think that
there should be, I mean, interestingly, of the 50 or
60 engineers that we have, 30% at all times are
dedicated to infrastructure, what we call core engineering. That's a lot of resources, that's a big slice of the pie that's
on core engineering. Then people is something
that, you guys talk about this in your class,
as you start to scale the organization, you
need to spend more of your resources thinking about how to take care of your people, how
to create career paths for them, how to think
about organizational structures that maximize the talent that you have in the organization. So we're dialing up
organization and people, we're dialing up monetization,
we have more resources, so in some cases, even though I say we're dialing down growth,
we still have the same number of resources towards
growth, it's just as a percentage of the whole,
it's gone down a little bit. - [John] Can you talk about
how the planning process, you talk about setting these priorities, you've got an offsite, how has that changed as you've gotten bigger? You're up over 100 people now, heading toward Dunbar's number. - I think what happens
as you get larger, is you have a clearer grasp on the near term, which gives you the freedom to
spend a little bit more time planning the near and the
mid to long term, I'd say. In the early days of
the company, think about 2010-2011, we were not
doing annual planning. It just didn't make any sense. We were literally having
standups every week thinking about, what are we
trying to accomplish this week. So now we're at a point,
and I've just hired a great executive for HR to come in, because, again, we want
to dial up our investment on people, and one of the
things she told me was, we're doing this 2016 planning, and that's all fine, but what's the bigger picture? Where are we at the end of 2017? Where are we at the end of 2018? And for so long, I thought, I can barely keep track of what's
going on next quarter. I can't think about next year. But then I started to
realize the reason it's so important at our stage to think about 2017 and even beyond is, if we
don't set that big goal and then work backwards, we may not choose the right things to do in 2016. What happens over time is, you have more people involved, that's sort of obvious. We try to create real bottoms up planning of the real initiatives,
and so again, as a company gets larger, you want
to get as many people involved as possible, but it's hard to do. You have to develop processes to do that, so it's not just chaotic,
so we, typically, as a management team say, in
these five areas, here's an OKR, an objective
and a key result in each of these areas, but how,
that's sort of the what, and maybe, at a higher
level for me as CEO, I have to articulate why,
why is this important. So I start with why, the team comes in and helps us understand what, but then the how is something that we want
everyone in the company to participate in,
because they're the ones that are actually creating this value. When you're small,
you're doing that all in one room, in a very
sort of haphazard, like, you don't want that overhead,
offsites and thinking about strategic planning,
and, you know, this is sort of paralysis by
analysis in the early days. But as you get a little further on, you need to dial that up a little bit. - [Student] Could you
talk about one or two of the strategies or tactics
you're (unintelligible)? - Yeah, so I was very lucky,
with Epinions, to have a guy named Bill Campbell on my board, who's my CEO coach, and he's probably the most incredible mentor, advisor, coach, management kind of
expert in Silicon Valley. In fact, I used to joke,
when he was coaching me, he was also coaching Steve
Jobs, and Eric Schmidt at Google, and Jeff Bezos,
and he'd come visit me and I'd say, Time for
the scrub team, right? 'Cos you know, you've
been with the real CEOs, and now you gotta come with me. He's like a business father to me. I would say, to most people, that most of the good things I've
been able to accomplish, I've learned things from
him that have enabled those things, and all the
bad things that I've done, he gets none of the
blame for any of those. A couple of things, he ran
into it for a long time, he was the CEO, and I
think he's still the Chair, and they had a set of core values, and the first one was something along the lines of, it's all about the people. So, the first and probably
most important thing I learned from Bill is, it
is all about the people. These things that we're
building, these companies, these products, these businesses, it all comes down to talented people. Getting talented people
in a room, making sure they're aligned, getting
the team together, and then pointing at the mountain and taking the mountain together, that's what he's exceptional at,
and he's done it in his own companies, and he's helped many incredible companies do the same thing. He was a football coach for a long time, but he's the ultimate coach in general, and I mean that term in as glorious a way as you could mean it,
he's the ultimate coach. - [John] He doesn't get
a lot of ink these days, he's a behind the scenes guy, but he's probably the best mentor, like, one of the best mentors in the world. What he's really special
about, is he can give you specific situational guidance, but he also recognizes that you're not the same guy as Eric, wasn't the same
guy as Steve, I think he coaches, he sees the
genius of you, and he puts his advice and hangs
that around who you are. - He's got, I would say,
three incredible, incredible and unique skills, the first is, he's just an incredible leader himself, an unbelievable leader himself. His instincts are incredible, and he's someone that people want to follow. He always used to tell
me, your title doesn't make you a leader, your
people make you a leader. Being a leader is about being someone that people will follow, and he's
someone people will follow. So you want to soak up as
much as you possibly can. That's the first thing, second thing is, the guy's actually done
it, he's not a professor who's never been in the field, he's not a management consultant who's helped with a few people but has never had to make the decisions himself,
this is a guy who's been in the boiler room and has achieved tremendous success doing it, he knows and empathizes with
everything that you have to do as a CEO or as a
management team member. The third thing is,
think about the companies that he's been around,
Apple, Google, Amazon, I mean, what has he seen, he's literally seen the history of our industry. I don't think there's
anyone like him, no way. - [Student] Taking a
step back, what is the conversion rate of the
postcards you guys send? (laughter) - Back to growth. - They won't let me divulge the numbers, but what I will say
is, it's not as high as we would like it to be,
but particularly in 2015, we've been able to significantly increase the conversion rate, which then allows us to send out more postcards, because we're looking at a cost per
acquisition, and we're sort of backing into that,
based on the conversion rate and so, we're very
strategic about postcards. I'll give you two vectors
upon which we're strategic. The first is, we don't
give every neighborhood the same number of
postcards, if a neighborhood has 2000 members, the
neighborhood as a whole gets very few postcards,
'cos it doesn't need them. But if a neighborhood
is trying to get to 10, we may give those early people who are starting the neighborhood 100 postcards. That's the first thing that's
actually pretty interesting, in fact, moving into
2016, I think most of our postcards will be going to neighborhoods that have fewer than 50 members. We know from all the
data that, neighborhoods above 50, they're gonna grow on their own. The second vector that's
interesting is, we've now built a reward system, where the number of postcards you get as
an individual is driven by the number of acceptances from postcards that you've sent in the past. If we've given you 10 postcards, and you've sent them out, and five of your neighbors have joined, you get more. If we've given you 10 postcards, and you've sent them out, and none of your neighbors joined, you
may not get any more. This is done on a monthly
basis, and it's actually pretty interesting to look at the data, because the first day
of the month, we have this giant spike in postcards sent, because there are
literally people, they want so badly for their Nextdoor neighborhoods to get penetrated, that
they're waiting until the clock strikes midnight on
the first day of the month, and then they're sending
all their postcards out. - Can I ask a dumb question, maybe? How do you figure out what's a reasonable CPA for a pre-monetization service? - You have to have some
assumptions of, ultimately, how much you can monetize those people on the other end, let
me just hypothetically say that these postcards, the CPA ended up being a dollar, so it's a dollar for us to use postcards to acquire members. We should be able to generate a dollar per member in some
reasonable number of time, and we've gotta back
out the churn members, and we've gotta back out any other things, sort of, you can do, you
can write fiction in Excel as they say, so you've gotta
be very diligent about it. I think for us, it was a
combination of thinking about value capture, but it was also a combination of thinking
about value creation, which is, what is it worth to us to have 80,000 neighborhoods that have launched, and how much would we pay for that? We look at it a number
of different ways, but because, as a management team, many of us have laid off people, we
are exceptionally careful with our funds, exceptionally
careful, we have to be. - This is a really difficult,
just for the record, this is a really difficult
calculus, as you can, fiction in Excel, right,
you can convince yourself of almost anything, you can justify almost any spend, (unintelligible) - I'll tell you in
general, as a CEO, I'm okay with spot spending for experimentation. If it fails, no problem,
because you don't have to do it again, the things
that I'm very wary of, and over-hiring is one of these people, one of these things, I'm
very wary of spending money where I'm locked into spending over the long term, without any terminal point. When you hire someone, you're giving that person a job, and if things work out you want that person to
work with the company for as long as possible, but
if you decide to do a test where we give twice as many postcards to some group of people, and we see what happens, we measure the conversion, we don't have to do that next month. I'm a big fan of taking risks, because you have to take risks to
actually have successes, but I like to bound those, I like to be disciplined and bound them. - Another way to say
this, structural things that increase your burden are dangerous. - Very dangerous. - Long term leases, office space. - Office leases are
terrible, they're terrible. In fact, today, the last
three leases that I've signed have been subleases,
they cost a little more, but they are not long
term, and I don't know how to predict the size
of the company and its square footage needs
over a three year period. I shouldn't be able to,
because if things are going well, we should be growing a lot. I'm willing to pay more per square foot to not lock myself into a
12 year, in San Francisco now, the landlords are looking
for a 12 year commitment. (laughter) And they're getting
it, they're getting it, and the math people are using it, they're saying, well,
there's such a huge demand out there for office space that, even if I grow out of this space in three years, I'll find someone else
to sublease it from me. Well, there are two problems, one is, what if they contract, and they don't need the space themselves, and number two, if the whole market contracts,
that demand's gonna shrink. It's smart on the landlord's
part, because they want to lock in all the good stuff
that's happening today, but it's very dangerous for
startups, very, very dangerous. - [Student] (unintelligible) - Sixty, probably, yeah. - [Student] Have you had any
surprises (unintelligible) The second question is, I
appreciate the discipline of getting new members
now, as you think about moving into a product
base, or revenue base beyond advertising, could we get some overview of how you think
about shared services, or other forms of revenue (drowned out) - On the first, I used to tease the team because they would come to me initially and they'd say, our member is 35 to 55, married, kids, and I'd
say, no no no no no! I don't want to create
that kind of product that has that kind of narrow band. And I thought about it, and I said, but, it sort of is the data, so
you can't dispute the data. Ultimately, what we
realized is, what was the age band of the first
million users of Facebook? It was, like, 17 to 21. What's the age band today, well, I know my 72 or 73 year old mom is on that thing every single day, and
so, you pick an initial market, and you over serve that market. That, for us, was people
who own their homes, and so you have to be a little older because you've generated some capital to do it, who had
children, they cared about safety, they cared about
schools, those kinds of things, they were
living in suburban areas because that's where you can afford homes and send your kids to school. That's where we saw real
penetration early on. Here we are today, 90%
plus of the neighborhoods in the top 40 cities,
there's no rhyme or reason. We have incredible penetration in Florida senior citizen communities, and we have incredible penetration
in the most urban areas like Potrero HIll in San Francisco, where you have a much younger base. There's even some
neighborhood at Stanford. I don't know exactly what's happening with that neighborhood, you should try to join it, I think it
might be some professors, or some folks who live
on the Stanford campus. The hope was always that we would pick a group initially, and
then widen the aperture, either through product features, or just because we understood
the needs of the users. What it turned out happened for us is, our core value proposition,
which is to speak with your neighbors, or communicate and connect with your
neighbors in a trusted way, that's not an age driven
value proposition. It started as, it's more
important for a certain set of people, but it turns
out that it's marginally important to everyone who
lives in a neighborhood. So now the ambition, it knows no bounds. We believe, and we said
this in a board meeting a couple of months ago,
100% of the neighborhoods in the country should be using Nextdoor at some point in the
future, it shouldn't be just the neighborhoods
that have married people. The second question, how do
we think about monetization? I think there's
interesting learnings here, particularly for consumer
internet companies that have indirect models, where you're not gonna ask your members to pay. Uber and Amazon, they ask you to pay. It's a direct transaction, you know what you're getting into, Google and Facebook, they're free services, and
they're monetizing through you. Let's talk about those two as a way of illustrating the way we think about it. For Google, the ads, in many cases, are a part of the
product, if you do a query on some commercial thing, you may think of the ads as content, and I think they've even done all kinds of testing and analysis that shows
that when they remove the ads on certain
queries, trust goes down, it doesn't go up, because they're useful. Now, let's think about Facebook, which by the way, is a juggernaut, they're doing incredibly well, and I
have tremendous respect in every way for what they're doing. However, I have never
clicked on an ad on Facebook. I've never thought to
myself, I'm going to Facebook and I hope I see an ad that I like. They have said, we have
such incredible scale, and they do, a billion concurrent users in a single day, that we can just put a couple of ads in there and we can drive what they refer to as demand generation, whereas Google has focused
on demand fulfillment. When you go to Google and you type in "digital camera" it's
because you want to buy a digital camera, and they're
gonna help you fulfill. When you go to Facebook, and they show you a picture of the newest camera, and 100 of your friends like it,
you may actually say, cool, I want to go buy this camera. But, that's demand generation
versus demand fulfillment. That's the way we're thinking
about Nextdoor as well. I'll give you a couple of examples. I mentioned that 26% of
our use cases are around finding a service provider,
that's demand fulfillment. Neighbors, does anyone
know of a great contractor? Well, we should be able to, in a very straightforward way, monetize
that, and we plan to. On the other hand, real estate agents are clamoring to get into Nextdoor, because they either want to tell you how much your home is worth so you can sell it, or to tell you that there's
another home around you that's bigger and nicer
that you may want to buy. But, they want to catch people who aren't necessarily thinking
about buying or selling their homes, in that case, it might be demand generation for us, so we're thinking about both of those, but I think the important thing is, over half of the conversations that go on on Nextdoor are what we would call transactional. There aren't status
updates and photo sharing. It's not like you come back from Hawaii and you say, hey neighbors, look at these killer waves that I just surfed on. It's much more, you go
to your neighbors and you say, hey, I'm thinking
about where to send my kids to school, do you
guys have any feedback? If it's a private school, that's a very large high economic value decision. The good news for us is
there are conversations about cleaning up graffiti, and about getting together to have a barbeque, that aren't easily monetizable, but in Google's case, only 30% of the queries are where they generate capital, and we might think of
Nextdoor the same way. That 50% that's
transactional, we wanna drive both demand fulfillment, which is very straightforward,
but we also wanna think critically about demand generation. - [John] One more question,
and then we should wrap up. - Who, who, okay this one, right there. - [Student] (unintelligible) - There's been a lot of
discussion recently about how do we make these conversations
as civil as possible. Because we do require
verification, we also screen out all sex offenders, because they're required to
register where they live, so you cannot join Nextdoor
if you're a sex offender, and because of this core
intrinsic thing, which is, if you're mean to your neighbors, you may see them in the real world the next day. If you're mean to someone
on a message board on the internet, and that
happens all the time, message boards are
basically, message boards and comments tend to be cesspools of meanness in many cases, but it happens much less on Nextdoor because you have to see these people, if
you raised your hand, and you said something really
inappropriate and mean, you come back to class a week from now, and all these people see you. That's a different kind of regulation. That being said, whether
it's education around the right way to report criminal activity, or whether it's the
neighbors talking about some political issue where
there's a disagreement, we have to remember that civility and constructive dialogue is the way these neighborhoods get better. That's actually, as we think about 2016, gonna be one of our big focuses, because before we had these massively penetrated neighborhoods, where there were thousands of neighbors communicating,
it was mostly all utility. There's no bad actors when someone says, hey neighbors, do you know
of a great dog walker? That's not a good opportunity to be mean. But in San Francisco
now, if someone posts, hey, here's the ballot
for the upcoming election on November third, and here's who I think we should vote for,
there's a very spirited conversation that goes on after that, and people disagree, and that's okay. That's good, because we want that conversation to happen, but we want it to happen in a constructive way, and that's something we're focused on. I think every large platform, Facebook went through this a little bit, where it was used as a platform for bullying, and sometimes it still can be, stalking, that sort of thing, any large platform that gains scale, and Facebook is at the very top of the list,
has a responsibility towards making sure the conversations and interactions are as safe and civil as possible, and we take that seriously. - Arright, I think we'll
wind up, thank you Nirav. - Thank you guys. (applause)
