00:00it is now my great pleasure to introduce
00:02my longtime colleague at Y Combinator
00:05Paul Buchheit Paul is known for a lot of
00:15things not the least of which is his
00:17wisdom and all things when it comes to
00:19startups but he's also of course the
00:22creator of Gmail the inventor this is
00:27true of don't be evil and has a has had
00:32an incredible career so I just thought
00:34I'd turn it over to Paul for a couple of
00:35minutes for him to describe his journey
00:37from somewhere to in into YC in the and
00:45and what's happened then and we're gonna
00:47have an interesting conversation I hope
00:49on on what it means to build an epic
00:55company Paul Wow all right thank you
00:58how far back do you want me to go
01:00exactly where were you born upstate New
01:09York me too yeah so I yeah I grew up in
01:13the Midwest and I went to college in
01:16Ohio and like in the 90s it was just
01:21different from now probably everything
01:24in the nineties yes everything from now
01:26it was different so yeah I was always
01:29interested in startups and like you know
01:32even as a kid like this idea I think I
01:34just didn't like the idea of working for
01:36someone else and also I liked the idea
01:38that you could make a lot of money and
01:40like create cool new inventions
01:42basically right like you need to read
01:44about inventors like Tesla or someone
01:47like that like wow if only that guy
01:49we're like smarter at business right
01:51like he he invented so much he invented
01:53so much and he made like he died
01:55penniless like he wasn't smart didn't he
01:57start that car company yeah I don't know
01:59if he actually I'm not sure if he
02:01actually got any equity I don't think he
02:02did that so so yeah yeah I was always
02:07covering the back of my mind but you
02:09know as I'm coming up to graduate
02:11college which was in 1998
02:14so I was very interested in startups but
02:16there really was you know there's
02:17obviously no startup school there was no
02:19Y Combinator there was none of was
02:22basically very little the web was
02:24relatively tiny at the time there was
02:27not a lot of resources I didn't really
02:28have any connections I didn't know
02:30anyone who worked at a start-up but I
02:33suspected they were like in California
02:36so just always doing kind of like the
02:38simplest thing first I've been treated
02:41by that what made you think of
02:42California I mean like eBay Netscape
02:47like that first generation Yahoo all of
02:49that first generation of Internet
02:50companies say they had already you know
02:52it's companies are IPO by the time I
02:53graduated so and they were all located
02:56here you know with the exception of
02:57Amazon up in Seattle so you know it was
03:01pretty obvious I didn't have to be that
03:02Silicon Valley to figure out where
03:04Silicon Valley is so so I took a job out
03:08here at Intel kind of with the hope that
03:11I would just like you know there but he
03:14startups everywhere I'll just like find
03:15one or something right so you went to
03:17work for the man say yeah I went to work
03:19at Intel and that was you know it tells
03:22a big company it wasn't it wasn't
03:26awesome it wasn't terrible it wasn't
03:28like a bad job you know I don't want to
03:30like but it wasn't it wasn't something I
03:33look forward to like continuing to do
03:36year after year you used to call working
03:38at HP where I first worked the fur line
03:42it's comfortable it's comfortable but
03:45it's a rut yeah I actually have a
03:47specific memory like hanging out with a
03:48couple of other friends who worked at
03:50Intel and you know I'm like 21 22 years
03:54old something like that and they're
03:57talking about like oh yeah they've got
03:59this really great like retirement plan
04:01where you know if you work there until
04:03for however many years like your age
04:05plus your number of years of service
04:07like you could retire at like 55 or
04:09something I'm like ah like it's like
04:12like oh yeah why don't I just take the
04:14next 30 years of my life and stick it in
04:16a box and bury it yeah I mean it's worse
04:18worse there are people there when you're
04:21sitting there at 22 who have been there
04:23for 30 years who are just kind of doing
04:26time yeah so I wasn't really into doin
04:29time but I was really into Linux so that
04:32was kind of my obsession I'd gotten
04:34interested in Linux like actually really
04:36early like 93 like when it wasn't
04:39actually very good and so I was kind of
04:42like an early kind of Linux elet who was
04:45like oh this is so awesome this is how
04:46you know the future is gonna be and you
04:49know back then businesses didn't really
04:50Linux wasn't like a serious operating
04:52system they're like no serious
04:54businesses will always use Windows you
04:57know you need like an enterprise
04:58operating system not some hobbyist
05:01junkie so I was like looking for
05:04something to work in Linux and because I
05:05you know everything at Intel was like
05:07pretty much windows-based and so that
05:10was my criteria when looking for startup
05:12so it's just that I wanted it to be kind
05:14of technically interesting Linux stuff
05:16again there weren't like you know great
05:19resources on like a huge list of startup
05:21so I I just kind of like scouted out
05:24what I could find basically on Slashdot
05:26because I would I would read slash down
05:29a lot and I came up with a list of like
05:30six startups that seemed like they might
05:34be interesting you know doing stuff with
05:36Linux and so I like just emailed my
05:38resume to all of them sort of like
05:41ironically it immediately bounced from
05:43Google because their mail server was
05:45misconfigured so I didn't hear back from
05:51from most of them ever I interviewed at
05:54one other company and then they I resent
05:57my resume the next day and then it went
05:58to they had fixed like I think the MX
06:00record was broken or something so I only
06:03interviewed at like Google and one other
06:05company and the other company like I
06:07wasn't interested and I think maybe they
06:09weren't interested isn't either one of
06:10us really like contacted each other and
06:13so Google was really the only offer I
06:15had so that made it easy to like choose
06:18and so I went to work there in 1999 and
06:22how many employees did Google have I was
06:24the 23rd at that point and so we were in
06:27we were in Palo Alto on University Ave
06:29kind of on the second floor above what
06:33is I think a book shop or cafe or
06:36something like that what did it feel
06:38like at Google when they were
06:3920 employees there it was really cool
06:41actually like I you know I was really
06:45impressed when I was interviewing with
06:46them that they asked actually like
06:48really smart questions which i think is
06:50a pretty good indication like that was
06:52one of the other reasons I wasn't
06:53excited like the other startup I went to
06:55that I'd interviewed at it was just like
06:58I went into their office and I just sort
06:59of felt the pressed like like you could
07:02just tell people weren't really that
07:03excited to be there and then they'd
07:05asked kind of like dumb questions I
07:07didn't think we're like anyway it just
07:10wasn't like a good sign versus like
07:11Google kind of like they were into
07:14whatever it was they were doing and
07:16actually after I was there for maybe
07:18like a week I went back and I'm like hey
07:21I want more equity like sorry you have
07:25to negotiate that before you start so
07:30yeah that might have been like the the
07:32smartest pseudo investment decision
07:34you've ever made even though it didn't I
07:36guess it worked out but yeah I mean it
07:38worked out overall but yeah I do wish
07:40I'd gotten more more so the thing about
07:44the office I mean you asked about the
07:46environment like it just had like a very
07:48kind of like a buzz of productivity
07:51right like because we were doing these
07:53really kind of exciting things and I
07:56think like they had hired people who are
07:57really into building these like big
07:59systems and creating great products and
08:02everything so it was it was an exciting
08:03place to work where you know I would
08:05actually like wake up in the morning and
08:07be kind of like eager to go to work cuz
08:09like I've got stuff I want to do do you
08:12think that was because not just they
08:15were working on exciting things but
08:16because they believed they were going
08:17somewhere yeah I mean there is there's
08:20kind of like an irrational like belief
08:23right like objectively we were just this
08:25tiny little company river but you know
08:28Larry would be talking about whatever
08:29crazy ideas we were just like wow what
08:32like I remember having an early memory
08:34where I was building this system and
08:36then I'm talking to Larry about he's
08:38like why are you even bothering with it
08:39disk we're just going to store
08:40everything in memory or whatever which
08:43at the time was sort of an absurd you
08:45know we didn't have enough RAM to store
08:46all the data but he was always thinking
08:49like so many years ahead in
08:52and it was a little bit insane but it
08:53made it a very like exciting place to be
08:56yeah it seems like there's there's
08:59something fundamental if you're thinking
09:00about whether you have a potential to
09:04build a really big business if you don't
09:06have some sort of vision some sort of
09:08idea of a future it's hard to believe
09:11you're ever gonna get there yeah I think
09:14I think you have to have some notion of
09:16the big picture and it doesn't mean you
09:18look like a detailed step-by-step plan
09:20like occasionally I'll run into a
09:21startup where they're like here's our
09:23five-step plan for how we become
09:25whatever and like it isn't until like
09:27the fifth step that actually they have a
09:29product anyone would want to use if
09:31those are like really bad I mean you
09:33have to you have to have something today
09:34that's good and you have to have
09:35something tomorrow and you have to you
09:37know day after day have like a product
09:39that people actually want you almost
09:42want a certain amount of vagueness
09:43around your vision because if you have
09:44too much detail you just guaranteed to
09:46have wasted your time because it's not
09:47gonna be right so you uh you spend how
09:51much time at Google so I was like Google
09:53for seven years so wow that's just about
09:58how long you've been at YC now I've
10:00actually been at YC from over seven
10:02mm-hmm so what happened during those
10:04seven years stuff happened right so
10:07stuff happened a few things so I yeah
10:10what did you know like when did you
10:12realize that Google was a thing that it
10:14was going to be a like you know it's
10:17this secular juggernaut that was going
10:19to take over the world
10:20earlier this year I I mean it bills I
10:25got I didn't I I think I like many
10:28people I continually underestimated its
10:30potential but you know for very early on
10:32kind of like after being there for a
10:34week I was like ooh there's something
10:36magical going on I didn't think it would
10:40be what it is today as big as it is but
10:43it definitely felt like we were somehow
10:46creating the future and then when we
10:48started really like like when we got the
10:51Yahoo deal and started like really
10:54welcome saying yeah we started really
10:58just like pushing all just crushing all
10:59the other search engines and then
11:02there was like a feeling of
11:03inevitability that goes pretty clear we
11:05were just gonna win because no one else
11:07like they just didn't even know what
11:10they were doing like the competition
11:11this wasn't very good before we move on
11:15to your - leaving Google and starting
11:20your own company it's just maybe say a
11:22word about the founding of Gmail and and
11:26sure so email is actually a thing that
11:29I've been interested in for a long time
11:31I actually first into your resume
11:32bounced yeah no actually before that I
11:34mean in college I I had actually been
11:39thinking a lot about this because like
11:41again back in the 90s like people their
11:44email would download to their computer
11:46they would run like Eudora or something
11:47and it would use pop3 to pull the email
11:49down to their computer and I remember
11:51people be like I need to go to my room
11:53to like check my email and like that's
11:55dumb like that doesn't make any sense
11:57it's information it shouldn't be like
11:58stuck in a box right
12:00and so I I've been thinking about you
12:03know this idea that like everything is
12:05just gonna be online and then we'll just
12:07like be able to access it anywhere
12:09anytime we want and so I actually in the
12:14summer of 1996 I think I kind of had
12:20gotten maybe a little bit excessively
12:21arrogant and had only applied for like a
12:24couple of summer jobs just assuming they
12:26would want to hire me and they didn't so
12:28I sort of ended up without a job and
12:30something like that's fine I'm just
12:32going to start my own company and so I
12:34actually tried to start this thing of
12:36like a web-based email but I that's
12:43funny because that's right when we were
12:44building web-based email yeah mine would
12:46have been better if I have 96 rocket
12:49mail was really good anyway I had like
12:52grand visions for it but I my approach
12:55to developing it was was wrong it was it
12:57was very much like I had this great big
12:59grand vision and then I started trying
13:01to implement it you know piece by piece
13:03the components and then I got bored and
13:06gave up yeah it's actually a
13:09multi-person job to build email it's
13:12it's a yeah it's kind of a lot of work
13:14it's a lot of stuff but the
13:15I went about it I was kind of wrong
13:17because it was it was like I actually
13:19like did sort of this detailed design of
13:21this very intricate system and
13:22everything and I've kind of come at
13:25least for myself I've come to decide
13:28that that doesn't work I need and this
13:31is what we preach to the startups is
13:33like figure out what you can launch
13:35today and so to jump forward to to
13:38Google like it was kind of 2001 there is
13:41sort of a big reorg where they got rid
13:43of all the managers and just gave
13:45engineers more of a free range
13:47engineering thing gave us all projects
13:48and so they're just like I should build
13:53an email thing like they knew I kind of
13:55had an interest in email Robert but it
13:57was it was just like we want to do
13:58something in email we don't really know
13:59what exactly like just go build it so
14:02having learned from my prior experience
14:03I decided to you know I think I kind of
14:11know I have a short attention span so
14:12the very first version of Gmail I wrote
14:14in a day and it wasn't very good
14:18but I took I previously been working on
14:20Google Groups which was like a Usenet
14:22search at the time now it's something
14:25different but I took the code that I've
14:28been working on for that and then I just
14:30like took all of my email and shoved it
14:32into that you know database into that
14:35indexing engine and then you know I kind
14:37of had an email search and they're like
14:39yeah so your first thing was just an
14:40email you could search yeah and so then
14:42I like email that out to you know the
14:44engineering team I'm like hey I built
14:45this email search you let me know what
14:47you think and then people reply like
14:48yeah it's okay but it would be better
14:50with my email instead of yours so like
14:53all right I have a feature request
14:56so first version zero only searched my
14:59email isn't like you know the next
15:01version I I made it so it would search
15:04other people's email and but it there's
15:06a stunning lesson in there I think which
15:07is like the point wasn't so much that
15:10whose email was being searched it was
15:14that email was being searched like you
15:16guys got to realize back then you
15:18couldn't really search email all you
15:19could do is like find the email you
15:21wanted and click on it but so searching
15:24was like a thing yeah and I figured I
15:26mean a lot of our emails and
15:28and so it sort of works right we're out
15:31in a lot of the same mailing lists or
15:32everyone see seeds so so yeah so then
15:36version two is just like it indexes
15:38other people's email and so I send that
15:40out and they'll be like all right this
15:42is cool now I want to be able to reply
15:44to one of the messages
15:45alright feature request right and so the
15:47whole thing was just iterating the you
15:51know to step by step trying to make
15:54something that would make people happy
15:56and so we actually had as our kind of
15:59like launch gate this idea that we
16:01needed to have a hundred happy users
16:03inside of Google because we you know we
16:04just had we hadn't launched it to the
16:06world obviously and so I actually
16:09embedded in the in the in the interface
16:12this thing of a pop up and be like are
16:14you happy yes or no it was just like a
16:16single question are you happy with you
16:19know Gmail or what was at the time known
16:22as Caribou yes or no and and it would
16:26get a list then I would look at of who
16:28said yes and who said no and then I go
16:30to all the knows and be like what's it
16:32gonna take to make you a happy user
16:34right and actually just get them one at
16:37a time and some of the people they're
16:39like it basically has to be a clone of
16:41Outlook I'm like well you're never gonna
16:43be a happy user now actually I did end
16:46up getting more tips to me I know I I
16:48did end up getting those people and the
16:50way I would get them is that outlook at
16:52the time when you would hit 2 gigabytes
16:54in your in your mail file it would
16:56corrupt the database and they would lose
16:59all of their mail and then I had the
17:00only copy of their email and so then
17:02they would become a gmail user not
17:04always a happy user but they would
17:05become a user anyway other people it
17:09would just be like a minor thing we're
17:10like I just need to add one feature or
17:12like fix a bug or something and then
17:13they'll be happy and so we kind of
17:15slogged through like just one user at a
17:17time until we got to like 100 happy
17:19users is that the genesis of you know
17:22what we call it YC the poo height rule
17:23maybe you can just tell these guys what
17:26like it's better to have 10 right so
17:29that one of the ideas is like when
17:31you're starting out building something
17:33new especially if you're kind of going
17:34into an established category like email
17:36like like literally you know email was
17:38like 30 years old when we started on it
17:41right so there is a lot of history and a
17:43lot of opinions about how email should
17:45be like people would sometimes angrily
17:47tell me I'm like doing it wrong like
17:49because we made the reply on top instead
17:51of the bottom or like weird stuff like
17:53that and so there's like all this
17:56history so it's pretty much impossible
17:57to enter a space like that and make a
18:00thing that appeals to everyone and if
18:02you try to do that what you end up
18:04making is just like a mediocre product
18:06that nobody really loves and so my
18:10philosophy and what we try to get you
18:11know all the stars to do is figure out a
18:14thing that will just have really deep
18:15appeal even if it's to like a tiny
18:17fraction of people if you can make that
18:20small fraction people just like
18:21obsessively love what you're doing it's
18:23easier to then grow that group you know
18:26cuz like there's always people at the
18:27margin where if I just make something
18:29slightly better they're gonna join into
18:31that group so it's easier to start with
18:32that like that deep but narrow appeal
18:36and then broaden it over time then it is
18:38to start with just sort of like broad
18:40men and then try to like convert people
18:43from math to like loving your your thing
18:45in mass yeah I've been thinking about
18:47that lately just in terms of retention
18:49like if you get a hundred people who
18:51never leave that's a great place to
18:52start then if you have thousands of
18:54people all of whom turn eventually yeah
18:57okay so Gmail happened and then you left
19:02yes so happened and yeah a
19:06couple years later I I left the the
19:09company kind of you know it was becoming
19:11a big company again and and I just
19:14decided that wasn't the direction I
19:16personally wanted to go and I just had a
19:19baby who had a pretty traumatic start on
19:22life so I kind of had other things going
19:24on so actually around that same time
19:27actually even before I left Google I was
19:29again reading slash dot because that was
19:31where I got all of my information and I
19:33saw this thing about the summer founders
19:35program which I thought like sounded
19:38super cool because I'm always just like
19:39looking for an idea that's actually
19:41novel for those of you don't know the
19:43summer founders program that's why see
19:45that was the so that was a terrible
19:48write that ended up being YC batch
19:50summer o-5 that's the one that had read
19:52and my chicken visit didn't say Sam's
19:57company Justin and Emmett's company Kiko
20:00we're all in that batch so I saw that
20:03sir Khan is the founder of justin.tv and
20:06and twitch and atrium so yeah a few
20:09things a few things yeah so I was just
20:13like really interested I didn't know
20:14Paul Graham or any of these guys so I
20:17was just kind of like keeping it on my
20:19radar like I'm really excited to see and
20:21part of that was it was my own
20:23experience kind of like being there in
20:25Ohio and thinking like I'm excited about
20:28startups and I wouldn't even know where
20:30to begin right because I don't like if
20:32you go to Stanford like your roommate
20:34probably has a startup or something like
20:35it's just in the water but if you're
20:37from someplace else if you're not from
20:39Silicon Valley if you're not from this
20:41environment it like it feels like you're
20:43just excluded and and I really liked the
20:45idea that it was just this open thing
20:47like anyone in the world you didn't have
20:49to know anyone you could just like
20:50submit an application and they were
20:53together like you know bring you in and
20:56sorry to let startup schools the same
20:58way now yeah exactly I mean what is it
21:00that's the idea that I think like really
21:02resonated with me is that it's like
21:03opening it up to everyone even if you're
21:05not kind of like an insider or something
21:07so I followed them you know pretty
21:10closely I want I was one of the very
21:11first reddit users I think when when
21:13that launched and then did you go to the
21:17demo day no I didn't make it to the very
21:19first that's the only one you ever miss
21:20the only one I missed I I've been to
21:22every demo day except the first how
21:24actually is the distinction of having
21:25been to more demo days than anyone yeah
21:28than anyone including PG and Jessica if
21:33it's a good distinction but as a
21:34distinction yet it's a distinction so so
21:37yeah then that fall they announced like
21:40PG put out this thing hey for the second
21:42batch we're gonna do it in Mountain View
21:43like in this room it turns out and so I
21:46was like oh that's cool I just dropped
21:48them I didn't know my stomache old email
21:50introduced myself and like hey if I can
21:52like help out in any way just let me
21:55so he's like yeah come on by and you
21:57know I started just attending the
21:58dinners and actually made my very first
22:01angel investment in a company in the
22:03second batch winter o 6 which was Kevin
22:06who then later became a YC partner wufu
22:09which is a formed filler you know it
22:13isn't like a company that I pee out but
22:15it actually turned out to be really good
22:16an investment nice first yeah yeah I had
22:20I forget what the multiple is but pretty
22:21pretty high yeah you know I thought my
22:24first investment returned like 10x and
22:26it was actually a little unfortunate
22:28because it made me think I was better at
22:29it than I was I think it's maybe like
22:3250x so yeah hmm so yeah I so that was
22:37kind of how I got hooked into this stuff
22:40so I started at that point just getting
22:43to know the founders in each batch and
22:45I've invested in every batch since then
22:48it's at least a few of the companies and
22:49how did FriendFeed emerge during that
22:52time so yeah after I left Google I sort
22:55of like retired and then we just like
22:59get depressed cuz it turns out tired of
23:02yeah and maybe I have issues with
23:04depression that I don't know so I but
23:06working distracts me so I started
23:09talking some other friends from from
23:11Google who were thinking of leaving and
23:14basically ended up starting a new
23:17company which which was FriendFeed and
23:20that was actually the actual goal of
23:23FriendFeed which was just to build a
23:25place that was really awesome to work at
23:27I was actually just trying to create a
23:30job for myself we tried to recreate what
23:32it felt like when you're at Google in
23:34the early days yeah kind of like the
23:35best of parts because there was also a
23:37lot of stuff about Google that I found
23:39just tremendously frustrating and so I
23:41kind of had this idea that like we could
23:44take like the good parts and not the bad
23:45parts it's like a little bit naive but
23:47that was the hope and you know managed
23:52to get like an incredibly strong team
23:54mostly of a former Google people did you
23:58ever feel like I mean frankly it went
24:01fine but it didn't obviously end up
24:03becoming like a hundred billion dollar
24:04company did you ever feel like you could
24:09pinpoint what was missing from from the
24:13beginnings of FriendFeed that might have
24:15led you on that path or what I mean was
24:20grand vision was that missing was it
24:22it's a explosive growth was missing you
24:25know social social is for those of you
24:28don't know it was it was a social
24:29network we actually created the like
24:31button so like the very first like
24:33button was on FriendFeed and then
24:35mysteriously appeared on Facebook number
24:37of months later they advanced instantly
24:41completely by coincidence ya know
24:44anyway social stuff is just really hard
24:47even and we've seen this Ivan you know
24:50we've had a lot of social stuff come
24:52through YC you know Google has tried it
24:55like it's a really hard space like
24:57somehow some of these things that kind
24:59of like you know Twitter or Facebook
25:01they just kind of like catch and you can
25:04go back in time and make up a bunch of
25:06stories about like why it is and some of
25:08those stories may or may not be true but
25:11it's it's one of these things I think is
25:13like one of the hardest product
25:14categories to like engineer success so
25:18but actually I will answer your question
25:23which is that I I did sort of come to
25:25realize this because in in the process
25:28of you know trying to make friend foetus
25:30success we kept running into this
25:32company Facebook who who actually knew
25:35what they were doing and and part of why
25:37ultimately like I came to believe that
25:42Facebook was just this unstoppable force
25:45in that like we weren't gonna beat them
25:46and Google wasn't going to beat them and
25:48Twitter wasn't gonna beat them like no
25:50it's certainly not myspace right like
25:52none of these got any worse album it's
25:55like having sort of competed with him
25:58you start to realize oh they're really
25:59good at this right and and I kind of
26:04once I was inside so that was why you
26:06know it sold front feed to Facebook cuz
26:08I'm like wow this is gonna be like
26:09another Google sighs there was that as
26:10to what you were doing like Google was
26:12their best in the world at what they
26:13were doing when you were there and they
26:15were the best in the world and you
26:16weren't right and so it's just like I
26:18should go there because it's gonna be
26:20first of all I feel like another one of
26:22these Google size companies and then
26:24like hopefully I can learn how they do
26:26it and I think like being there and kind
26:30of reflecting not at one realization I
26:32is that we kind of brought like Google
26:34always had like a very product oriented
26:36mindset like like we would try to build
26:38products that were like like what the
26:42user wanted so it's like you guys are
26:45users you built product like you said
26:47you built Gmail originally for people
26:50and so with front feet was kind of the
26:52same thing of like we're just thinking
26:53about like what do I want and I think
26:56like one of the really powerful insights
26:58that they had at Facebook was that it's
27:02almost like the it's not the end user
27:05who you have to please it's the network
27:08because they understood that the most
27:11important feature of a social network is
27:13that your friends are there and so
27:15sometimes there's actually a tension
27:17between like what's good for the
27:19individual and what's good for the
27:20network and they would always do what's
27:22good for the network do you think that's
27:27just specific to network effect
27:31businesses social media or is that a
27:33more general rule do you think like an
27:37enterprise product probably doesn't need
27:39to think yeah I mean an enterprise kind
27:41of has some of the similar thing where
27:43you have to make the buyer happy not the
27:45end user which is I think a lot of times
27:47why enterprise software is not really
27:48lovely to use is because the person
27:51using it isn't the person with making
27:52the binder so both cases you sort of
27:54have to have a meta level where you're
27:56like we always talked about making
27:58something people want but people you
28:00have to think more broader than just the
28:01end user it's it's everyone who's
28:03involved in the in the system right
28:05right and it's people get this wrong a
28:08lot of times where they think like
28:09there's this meme out there that like oh
28:12if you're not paying you're not you're
28:14the product not the customer wherever
28:16but it really isn't right like you have
28:17to identify whose happiness makes you
28:20successful so like for Google it really
28:23is like the end like the people using
28:24Google search like if they go to a
28:26different search engine like there's
28:27Google screwed right like you have to
28:29keep those people happy and you know
28:34with the Enterprise example its whoever
28:36the buyer is in the I think the sort of
28:39subtle thing with with Facebook was was
28:41that what makes those people happy is
28:43that you have this like huge network
28:45and so work that's that's efficient and
28:48easy to use and right and so and so you
28:50really have to think in terms of like
28:51almost what the network wants which is a
28:53different way of thinking yeah so when
29:00you when you were you friend Pete was
29:03bought by Facebook and so you you landed
29:06in facebook did it have a similar feel
29:08to their I guess Facebook was a little
29:11bigger by the time you got bought but
29:12did it have a have a Google like feel
29:14then I mean there are a lot of people
29:16from Google hey hi you know former
29:21co-workers similarities and differences
29:25so that was again one of the things one
29:28of the reasons I was excited to go there
29:29it was just like I had had this
29:30experience going through Google and it's
29:32very easy to over generalize you know
29:35like oh wow the reason we were
29:36successful is we had like you know
29:38brightly colored balls or something like
29:39that like you can make up all of these
29:41like the rationalizations for what made
29:44you a success but in retrospect
29:45everything looks the truth is I think
29:47like the majority of the reasons you
29:49come up with are probably wrong there's
29:51there's like a few key things I will say
29:54though that like going into the office
29:56it kind of had that vibe of like this is
29:59a place that's going somewhere right and
30:02and I think it's like a really powerful
30:03indicator and it's a subtle thing and
30:05it's a little bit almost metaphysical or
30:08everybody like you walk into an office
30:09and you're just like how is the energy
30:10how does it make me feel does it when
30:13I'm here is that like when I watch my
30:16own energy levels am I going up is this
30:19making me feel like oh wow or is this
30:20like pulling me down am I like depressed
30:23because and I think the reason that
30:25works is because the employees actually
30:27know what's going on right and so if
30:29you're just there cuz like you're
30:31collecting your paycheck while you wait
30:32for the company to go bust like you're
30:35gonna have a different energy from yeah
30:36you feel that if it versus if you think
30:39like oh my god I need to launch this
30:40thing it's gonna be so exciting like
30:42it's gonna take over the world right and
30:43I guess it's just a very exciting
30:46feeling to have like these products like
30:48like when we were building Gmail and
30:51maps and stuff like there was this
30:53feeling this palpable feeling of like
30:55this is gonna blow people's mind so
30:57buddy put it out there right like
30:59that's a really exciting thing do you
31:01think so part of the energy comes from
31:03these amazing products that you're
31:05building but it seems like part of that
31:06has to come from the founders and the
31:08just the attitude that they bring the
31:11belief that they bring like whether it's
31:13soccer or Larry and Sergey yeah
31:15absolutely I mean I think that there has
31:18to be a founder conviction like they
31:21have to believe in what they're doing
31:23and not that they're just chasing after
31:25you know some trend like oh I read an
31:27article and TechCrunch I'm gonna go off
31:29in a new direction I mean I actually
31:31think like Zuck is probably the most
31:34extreme example of this I've seen I
31:36almost feel like the whole world could
31:37tell him he's wrong and be like no and
31:42and and this is really reflected and
31:45some of the I think internal mythology
31:47of the company there's a really
31:49interesting and this is before my time
31:51but you know back when they were just a
31:55college social network they had
31:57something like 10 million users which
32:00first social network isn't that much and
32:01Yahoo wanted to buy them for like a
32:04billion dollars and everyone is like
32:06used to take the money he actually took
32:12the money but then a change that they
32:13like they're gonna changed theirs and he
32:16didn't want it there was pressure on him
32:17right so like everyone around it was
32:19like pressuring him to take the deal and
32:21then by the way if you get offered a
32:22billion dollars take the deal right
32:25in case you have any doubts right and so
32:28finally yeah I mean ultimately turns it
32:30down basically replaces an entire
32:32management team after that because they
32:34were all kind of like we should sell we
32:35should sell I mean so he kind of brought
32:38in people who who maybe believed a
32:39little bit more oh and then there's kind
32:42of actually the story I meant to tell
32:44wasn't even that it was that when they
32:46launched the newsfeed the which was
32:49around the same time and people right
32:53hey they were furious and and so again
32:55they were like something like 10 million
32:56users unless something like 8 million of
32:58them joined a group protesting it like
33:01like it's impossible to get that
33:03percentage of your users to do anything
33:04even if you want them to so to like
33:06suddenly have the majority of your users
33:09like a revoltin or whatever and they're
33:11just like no I think people
33:12we're gonna come around on this in there
33:14you know right obviously look at the
33:16newsfeed is like a fundamental thing but
33:18that receipt is Facebook for most people
33:20right right there's nothing else but I
33:22mean that requires I think like a pretty
33:24high level of conviction when
33:26essentially all of your users are
33:27furious at you and you're like no I
33:29think you're going to come around on
33:30this the for somewhat unclear reasons
33:35this reminds me of another bukit ism
33:39which is that that sam actually repeated
33:43in his introduction to the course which
33:45was that you know what a company really
33:47needs is frugality focus obsession and
33:50love and it seems like like if you
33:53encapsulate those things even in the
33:54very beginning and you think about
33:57leaders like zack that they kind of they
34:00kind of go together and make sense yeah
34:02yeah i i think like focus is one of the
34:07most important things because like as a
34:09start-up it's actually i think your most
34:11powerful weapon right like the reason
34:12that you're able to take on like these
34:15big companies or whoever is because
34:16they're doing a thousand different
34:17things right and so the only way you can
34:19win is if you take everything you have
34:22and you focus it at one point right
34:23because you you are underpowered in
34:26every way versus these other companies
34:28so the one thing you do have that they
34:29don't have is you can be like absolutely
34:32100% focused on the thing that you're
34:34doing and that was like you know with
34:36google that was part of the reason that
34:38we won it search was i was all we were
34:40focus on like early on right is and you
34:42became the best in the world at it like
34:44you versus like everyone else's ago we
34:46want to be a portal right we need all
34:48this other so like that that's
34:51really key frugality i think is also
34:54really essential I think like one of the
34:59best indicators of if you're doing it
35:02right is essentially your input to
35:03output ratio so you want to do more with
35:07unfortunately most startups do less with
35:10more so like you give them money and
35:12they just burn it up and there's like no
35:14results so like being able to to produce
35:17a lot of outputs with a minimum amount
35:20looks like amplify time and money right
35:22but you have to like amplify the mean
35:24get that ratio where it's
35:26I give you a dollar and you turn into
35:27ten dollars my god that's a good
35:29business right like we just scale that
35:30thing up even if it's ten dollars worth
35:32of something like ten dollars worth of
35:34product $10 worth of innovation and so
35:37and this goes back to always figuring
35:39out how to do things in the absolutely
35:41minimal way and minoo it means this is
35:45one of the key startup things where
35:47they're like it you know if you if
35:48you're not embarrassed of your product
35:50you've waited too long to launch like
35:52these things are at the root there
35:55there's a lot of like frugality there
35:56right because you don't want to spend
35:58too much time or too much money like you
36:00want to get it out there as quickly as
36:02possible in as little cost and well so
36:05you don't want to spend too much time
36:06running on the wrong thing and you just
36:08don't know enough right right a builds
36:09way too much email and you wouldn't have
36:11learned what to actually build right
36:13it's easy to go completely off the rails
36:15and it's one of the most surprising
36:16things like I've learned just working
36:18with so many startups is how often the
36:20cause of death is too much money
36:22like I think there are companies that if
36:24they had not been so well funded could
36:26have been successful in the company I
36:28like to make fun of not because they may
36:31have been successful but just because
36:32it's funny to laugh at is juice arrow
36:35like the the $700 juice bag squeezing
36:39machine and the thing that made that I
36:42think such a like tremendous trainwreck
36:44is that they gave them a hundred and
36:46twenty million dollars for that thing
36:48before they ever like talked to any
36:50customer like if if juice arrow had gone
36:52through YC and we gave them one hundred
36:54and twenty thousand dollars they would
36:56have had to go talk to a customer and be
36:58like hey like we want to sell you the
37:00seven hundred dollar juice bags right
37:02they would have had to talk to customers
37:04like like and that's the thing like we
37:07push so you have to go talk to your
37:13customers is so easy to become detached
37:17and off in this like bubble of delusion
37:20in the problem with having lots of
37:22funding before you're really ready for
37:24it is that's exactly what happens so you
37:26can you can sell yourself on this idea
37:29oh yeah once we build the perfect juice
37:31bag squeezer like we're gonna take over
37:33the world everyone's gonna want to spend
37:35$700 I mean plus is like you get this
37:38giant box in the mail
37:39has tiny bags of juice in it is awesome
37:42like everyone's gonna want this right
37:43and they were able to live in that
37:44delusion for years because they never
37:47had to go out and talk to customers but
37:49if I refer a call you could actually
37:51just recreate everything you did by just
37:52going like that right the reason for it
37:54so I cut open one of the bags
37:56the reason is I think to actually get it
37:58to work they had to really heavily
37:59macerate the fruit so that's why I call
38:01it a juice bag squeezer because it was
38:03actually already juiced it was so is
38:05really just separating the pulp from the
38:07juice which is why also it's no better
38:09than just like going to the store and
38:10buying juice like the whole thing was
38:12like built on a number of flawed
38:14premises but but like in particular you
38:17weren't really even juicing it has
38:19already been juiced you're just
38:20straining the juice but there's a deep
38:23lesson there which is no matter what you
38:25raise if you raise fifty thousand
38:28dollars or five million you you should
38:32think of the amplification you need if
38:34you raise five million dollars that
38:36doesn't mean you can create 150 thousand
38:39dollars worth of value and be happy you
38:40have to create 50 million dollars worth
38:43of value yeah but you can't like it's
38:46easy again I think to come up with
38:48stories in your mind about how you're
38:49doing it because like once we release
38:50this you know juice squeezer it and and
38:55that's why I think like you just have to
38:57stay close to the customers you have to
38:58be out there selling because you can
39:00just become delusional otherwise and so
39:02that's why you know with startups like I
39:04I've just seen it again and again like
39:06if they get a huge round of funding
39:08before they have product market fit it
39:10kills the company like it because
39:12because it's just too hard for the
39:14founders to resist that because actually
39:18going out and dealing with the fact that
39:19you don't have product market fit it's
39:21really painful it's easier to tell
39:22yourself stories about how oh yeah we're
39:25killing it I'm hiring I just hired a
39:27whole new team we've got this cool
39:29office like there's all these like a
39:31cool technology yeah we're doing all
39:33this amazing stuff right I'm in the
39:35press all the time I go to conferences
39:37I'm up on stage you know like there
39:40knows like it's easy to just kind of
39:43like be completely detached from reality
39:45because because you're not out there
39:48McCrea ting a thing that people actually
39:50want to use we were talking earlier it
39:53seems like there's this is this
39:54interesting juxtaposition that that if
39:56you really want to be creating an epic
39:58company you have to be thinking about
40:00the future but you better be paying
40:01attention to the past at the same time
40:03because there's a lot of lessons to be
40:04learned about what goes wrong more
40:07things go wrong than go right with
40:08startups and yeah III mean that's just
40:10the past but I think more importantly
40:12the present like that's the thing like
40:14you have to have some some you have to
40:17maintain kind of like a dual like a
40:19dream or a vision of like what that
40:22future is of you know something that
40:23gives you an orientation like I need to
40:25head in that direction like like I think
40:28about you know the Lewis and Clark
40:30expedition or ever they're like we're
40:31going west we're going that direction
40:33but like you don't really know all the
40:36streams and rivers and forests and
40:38mountains you're gonna have to cross
40:39along the way right you but so you have
40:43to be very focused in the moment so that
40:45you don't get like eaten by a bear or
40:46something like that but at the same time
40:49you have to have that vision of like
40:50where you're headed and so I think those
40:52are kind of the two things I being a
40:54visionary alright yes like between you
40:56know just paying incredible attention to
40:59your current customers right and having
41:02take over the world right and I think
41:04the thing that happens with too much
41:06money is you just spend all of your time
41:07like dreaming of like the Promised Land
41:09but you're not actually dealing with the
41:11problem that's in front of you we may
41:14have I want to turn over to Q&A and we
41:17may have covered most of this already
41:18which is fine but you've been spending a
41:21lot of time with YC batch companies
41:23doing I think what you're called a
41:26hundred billion dollar office hours
41:28which is which is my understanding is a
41:32way for you to help them think about
41:35what it might mean to actually become an
41:37epic company what are the like you know
41:39there two or three things maybe that we
41:40haven't talked about that you cover in
41:41those office hours I thought this would
41:43be like a group office hour a hundred
41:44billion dollar office I yeah all of our
41:46startup school companies yeah so I
41:47thought a lot about you know what is it
41:49that makes these these companies that
41:51are you know Google Netflix Amazon like
41:56all of these like just huge huge
41:58companies like what is the
42:02that gets you there in it's not this I I
42:06don't think it's that you're smarter
42:07necessarily it helps to be smart it
42:10isn't that you're harder working there's
42:12a limit to how hard you can work what I
42:16think necessarily to get to that point
42:19you're sitting on top of a big
42:22exponential change in reality or in the
42:27world so like you know Intel and
42:28Microsoft in in the early days Apple
42:32we're kind of sitting on top of this
42:34like the rise of micro computers right
42:36and so like what made Microsoft a big
42:40company is that they got in early on
42:42micro computers and then they kind of
42:44owned that space right where the micro
42:46computer operating system you know and
42:49they were they were able to ride that if
42:51we weren't from microcomputers like they
42:53wouldn't have been able to do that and
42:55likewise you know Google obviously was
42:57sitting on the growth of the Internet
42:58that the reason that Google made sense
43:00in the in the thing that part of the
43:02reason other people got it wrong was
43:04they weren't thinking about the fact
43:06that the amount of information online
43:08was increasing exponentially and so we
43:10were gonna have to get a lot better at
43:12like or organizing it right like googles
43:15mission from the beginning is to like
43:16organize and make useful all the world's
43:19information and people are always
43:21surprised that they're aggressively
43:23gathering information it's like right
43:24there in the missions David all the
43:29world's information game sososo I what I
43:32try to have the the founders I kind of
43:35give them a two-part question and so the
43:38first part is like let's say I hop in my
43:40time machine which is hopefully a
43:43DeLorean and I like travel ten years
43:47into the future so I get out it's the
43:49Year 2028 you know I take a look around
43:52I like grab some nearby people and talk
43:55to them or whatever I ask questions like
43:57what about the world of 2028 is
44:01fundamentally different from 2018 like
44:03what's what's the thing that's totally
44:06and so the Trump will no longer be
44:11yeah yeah we don't really know the
44:18future you just completely derailed me
44:26that was the first thing out of two
44:29anyway so the thing is what I want them
44:31to do is identify the change and this
44:33should be as it pertains to their
44:35startup because obviously there's gonna
44:36be a bunch of changes but as it pertains
44:38to your startup but factoring out your
44:40startup because even if like Facebook
44:42didn't exist I guarantee someone would
44:44have built a giant social network right
44:46like Facebook was was was sitting on top
44:48of that exponential change and they were
44:50the ones who were most able to like
44:52capture it that position and hold on to
44:55it but if it wasn't them I guarantee
44:57someone else would be in that ecological
44:59and estimated ability there's a Nevitt
45:02ability these companies will exist in
45:04some form and maybe you're making it a
45:06little bit faster or a little bit better
45:08or something like that but that position
45:10is gonna exist and then the second part
45:12of the question is simply like how are
45:13you gonna be the one who captures it
45:15right so let's do some Q&A
45:20any questions what's the tipping point
45:36where you really approach the market and
45:39you know know that you're going to
45:41convert a lot of it for us and customer
45:43interviews about 80% of the people want
45:46to try what our product does and we even
45:49have people pay 40 bucks for a cup of
45:51coffee I just take my money so what's
45:55your question so how do you identify the
45:57tipping point and really how you execute
45:59that tipping point that gets a market to
46:02that so the question is how do you
46:05identify a tipping point for a market I
46:13almost feel like this is being over
46:16thought if you've got people who will
46:17pay you $40 for a cup of coffee I would
46:19just start selling $40 cups of coffee
46:22like that sounds like a really good
46:25and very profit the tipping point is
46:28when you can't keep up with demand
46:29right and so again like you don't you
46:32don't find this by thinking about it if
46:34you think about it too much you end up
46:36just like in delusional land you find
46:38out by talking to the market and having
46:40them give you dollars and and and so
46:42it's it's like literally when you're
46:44just like scrambling to keep up with
46:46with demand and that and that's the
46:48thing that you get into in these like
46:49hyper growth companies like one of the
46:52biggest problems inside of Google is
46:53like we couldn't build datacenters fast
46:56enough right like things are just
46:57exploding so quickly that you can't even
47:00scale out that's that's what tells you
47:02that that you have that yeah I kind of
47:04hit you over the head right it's not
47:05like you don't you don't have to give
47:07instruction for when you're in hyper
47:08growth you know you're there in the back
47:37practices for enterprise companies where
47:39maybe you can't launch something that's
47:41really really bad because there's
47:42requirements around security etc so the
47:48danger I feel like with a lot of these
47:50enterprise things is that you end up not
47:54really solving an urgent problem because
47:56like when you talk to people they're
47:58like oh yeah this is it would be great
47:59to have this solution or whatever and so
48:02then you go off and you build this thing
48:04you spend years maybe building this
48:07product and then you go to them you're
48:08like here it is we got it and like you
48:11know we're really busy this quarter and
48:13actually we don't have it in the budget
48:14until maybe next year so can we talk
48:16again in nine months or something right
48:18like and that's just deadly for a
48:22start-up like so the way I try to think
48:25about it is you want to find someone who
48:28has a really really urgent problem right
48:32and so like the analogy I like to use is
48:34like imagine that your arm is pinned
48:37underneath a boulder
48:38and you have two options one option is
48:42to like knopf your own arm like with
48:44your teeth or the other option is to
48:47sign an LOI with this promising young
48:49startup like right right like that's the
48:53kind of customer you want is like
48:54someone who's like desperate right and
48:57because they're desperate they're
48:59willing to put up with kind of a shitty
49:02product maybe they're willing to spend a
49:04lot of time working with you right and
49:08so like actually one of the things that
49:09works really well like is is is if the
49:12founders are actually able to just go
49:14into the organization and actually work
49:16there so like that was how like the very
49:19early stripe installs Patrick Collison
49:21would show up and integrate stripe into
49:22your code you just show up and do the
49:25work and so like that's a good thing and
49:28if they're not really eager to have you
49:30come in and integrate into their systems
49:32or talk to their people or whatever like
49:33maybe they're not in enough pain and
49:36like you need to go find a different
49:38customer who's in more pain and if it's
49:40a big enough market there's got to be
49:41someone who's desperate back there you
49:49believe in using personas to represent
49:51your customers do I have that right not
49:55don't believe in using people to
49:58represent my customers sorry common
50:17denominators for the founders hundred
50:20billion dollar companies so this was
50:23actually where the focused frugality
50:25obsession and love thing came from is I
50:27was actually trying to distill it down
50:28into a small enough number of words and
50:31then I was here to try to translate it
50:32into emoji but I failed at that part I
50:35couldn't figure out the I couldn't
50:37figure out what is the emoji for
50:38obsession like love I like okay a heart
50:43but there is I think like this obsessive
50:46focus there's there's almost an
50:47irrationality in some of these founders
50:52example that comes up as like the most
50:54insane version of it is probably like
50:57so I actually he came into Google gave
51:02I want to say 2003 or something like
51:04that and it was it was after he had
51:07started SpaceX but before they had done
51:09the first rocket launch and someone is
51:12like what are you gonna do if the rocket
51:17explodes when you launch it he's like
51:19well I have enough money for three
51:21launches so I hope one of them succeeds
51:23like he literally ploughed his entire
51:26fortune the first three blew up - I
51:28think yeah the first three did Lobby
51:30that he managed to scrape together
51:31enough money for like a fourth rocket
51:32yeah but that fourth rocket hadn't gone
51:34there would be no SpaceX and Elon would
51:36be bankrupt right like he like that is
51:39insane I would not take all of my money
51:41and plow it into like some harebrained
51:44rocket scheme right like that's like a
51:46level of irrationality that is I think
51:49like wonderful so I think the question
52:28is how come Google never made it in
52:30social networking you know it just
52:34wasn't I think really in the DNA so we
52:37actually like had even before Facebook
52:40do you remember Orkut like we like it
52:42was we actually could have or you know
52:44at the time like in Brazil right yeah
52:46well it became big in Brazil but I
52:48actually tell you like orchid actually
52:50was taking off and you can make up a
52:52bunch of reasons why maybe it didn't end
52:54up being the social network but it was
52:55exploding and then it became it was
52:59growing so fast it became unresponsive
53:01so like it just that was kind of why it
53:04died off in at least the United States
53:06so slow but part of the thing is that
53:10within the company I think like it just
53:13didn't resonate with people enough like
53:15it wasn't really it didn't become this
53:17crisis we're like oh my god we have to
53:18make this thing succeed it was kind of
53:20like this thing that's over here to the
53:21side right and so like Google is very
53:25much about like these really interesting
53:26big computer science problems and stuff
53:29like that like that's what really would
53:30get people excited and this stuff kind
53:33of seemed like a distraction I mean like
53:35the founders of Twitter or at least like
53:37two of the three founders of Twitter
53:38where Google employees right like like
53:41and they never would have built that at
53:43Google because like if you had built
53:44it'd be like what is this thing there's
53:45nothing technically interesting about
53:46Twitter right and so I think it just
53:49wasn't really in the DNA and then when
53:51they did try to do it it wasn't because
53:53like it wasn't because like you know
53:56Larry has a deep love for social
53:58networking or whatever right it was
53:59because they wanted to be competitive
54:01with Facebook but by then Facebook was
54:05way ahead of them and had a much deeper
54:07understanding of like what makes this
54:09stuff go so if you're behind and not as
54:12good you're guaranteed to lose it's not
54:14a good outcome so I think we're going to
54:16call it there thank you very much Paul
![How To Succeed As A NEW & YOUNG Realtor [Deals Every Month + Luxury Listings]](/_next/image?url=https%3A%2F%2Fi.ytimg.com%2Fvi%2FrtrRghgTayc%2Fhqdefault.jpg&w=3840&q=75)
