- So to begin with let's
give a warm CS183C welcome to Yahoo CEO Marissa Mayer. (audience applauds) So one of the great things
about being here at Stanford is that so many of the speakers we've had actually did their undergraduate work here and you, I believe, were class of 1997? - That's right. Yup. - And talk a little bit
about your time at Stanford because I know one of
the interesting things is you studied Symbolic Systems, you were also part of SLE, yay, and did Symbolic Systems
as Reid did as well. So talk to me a little bit
about Symbolic Systems. What did it mean to you? What did you learn from it? - Sure. Well, so I was here, I did my undergraduate, graduating in '97, and then did a Computer Science
Masters, graduating in '99. And, I did Symbolic Systems. Reid and I have that in common, and I did SLE, so I guess all three of us have that in common. It was part of my freshman year. I absolutely loved Stanford, but I really, really loved
the Symbolic Systems program. For me, I came to Stanford assuming I was going to be a doctor, and I got very deep into chemistry and biology my freshman year, and I was good at it, but when I went home that summer, I realized, and I compared notes with a lot of my other pre-med
friends at other schools. I realized that we were all
taking the same classes. We eerily all had the
same atomic, you know like construction kits,
all the same flash cards, and I really wanted to have
an experience here at Stanford that was more unique than that. So, I started looking at things that Stanford was uniquely good at and things that were unusual
for here at Stanford, and obviously, you know
Stanford's always had a very strong Computer Science department, a really strong Psychology department, and so I started learning
a little bit more about Symbolic Systems,
philosophy, psychology, linguistics and computer science and took some linguistics class, liked those too and really liked the
interdisciplinary piece of it, but I realized one of the things that had drawn me to medicine was,
I really was interested in neuroscience and how brains
develop and how they learn. The other interesting
application of that, of course, is artificial intelligence,
and so I ended up switching into Symbolic
Systems and spending my time thinking more
about could we actually build a brain that
operates the way ours does as opposed to how to dissect them. - And I think that you also,
while you were an undergrad. not only were you majoring
in Symbolic Systems, but you also taught at
least a couple of courses in Symbolic Systems as well. Can you talk about that? - Sure well I did. How many people here have-- I think everyone's taken
probably 106 A and B, and I think now it's almost
standard, run for the course, but I started off as a section leader, and then I became one of the head TAs, and then at the end of my time here, I became a lecturer,
so I ended up teaching 106 A, B and X a few times
at the end of my tenure. - One of the interesting
things is our faculty sponsor for the class is Mehron, and
you know Mehron of course, and Mehron actually was
my instructor for CS106A in the ancient days when I
was an undergraduate so-- - You may have been in my first class because Mehron was the first person I was a section leader for. He was actually unusual
because he was the one professor at Stanford
that actually would get more people on his second
day of class than his first. Maybe that was true
here for this class too. I don't know, but he was so popular, people would spread the word about Mehron and get more people to
come to the next class than even to the first one. - Absolutely amazing. I still remember some of the
lectures he actually gave, the snooze bar example, he
said "You still remember that? That's amazing!" So, you were here. Obviously you were very busy academically, but you probably spent your time on other stuff as well as an undergrad. You were very involved in a lot of things. Talk about what you did outside of class. - Well I really, I loved
the residential life, so I was, you know, an RCC, I was an RA, and so I did a bunch of
different work in the residences. I was also on the debate team, something I started doing in high school
and continued to do here. I did some parli debate. Which is a little bit less
intense than some of the national debate tournaments, because it's all extemporaneous speaking. I just really enjoyed my time here and I'm sure I did more
activities than that, but those are the ones
that really stand out. - And then when you finished your masters, and you went to work for what was then a pretty small company. How big was it when you joined? - I went to work for
Google, it's interesting. I assume many people
here know Eric Roberts, so I did a two-year masters because I was doing a co-term. Because I was a head TA I
basically had a half-course load. There's a very good deal
where if you become a head TA, they cover some of your
credits and give you a stipend and all of that, and so I had basically a summer off between the two years
of my masters program. So, I went to Switzerland
and spent the summer building something that looked at which websites you went to. You installed it in your browser, it watched where you went, and then would compare
it to other people's paths on the web and do
collaborative filtering. So if you visited site A, B and C, and someone later visited site B, they would say "well maybe you
want to visit site A and C." And we were aggregating
all this information for the Union Bank in Switzerland to make their traders better at finding relevant information
quickly in the morning when they first arrive at the job to see what the market conditions were. So, I came back and
Eric Roberts hired me to teach 106B for the first time. So it was my first time lecturing. I came back, sat down with him, he gave me some tips on teaching 106B, and then he asked a little
bit about my summer research, and I told him what I'd been working on. It was so funny because I
still give Eric a hard time about this because he said
"Oh, you know that's so funny because, you know, you were
looking at where people go on the web and trying to
understand where traffic goes and what sites are related. There's these two guys
on the fourth floor. They're building a search engine. They're not looking at
where people go on the web, but they're looking at links
structures between sites to kind of do something similar. You'd probably be really interested in what they're working on." And he's like "but I
can't remember the name." He's like "it's Larry Page, Sergey Brin, but I can't remember the name." And I was like "Well, you
know Eric, I just got back from Switzerland, I'm
teaching for the first time, I'm really overwhelmed." - I don't have time to
meet with those guys. - And I was like "I don't have
time to meet with a start-up or get involved in the
start-up right now." And so I kind of punted
and it was actually good because that was-- I mean, they had literally formed Google about the week before. Google started about middle
of September of 1998, and I had gotten back around
the 20th or so of September. And at that point, they
weren't really hiring and looking at how to scale up at all, and then fast forward to the spring, I got an email from one of
the first employees at Google asking "Would you consider
coming over to interview?" And it was funny because at
the time it was quite late. I had really procrastinated
on picking my first job, and so I said I would consider it, but I need you to interview me on Tuesday because I've absolutely committed, I'm going to make a decision by May 1st. And so it was funny because I went over to interview for the job, and then Larry and Sergey
came in and interviewed me, and then afterwards they walked out and I heard them say "Hey guys," I think they said "we're
going to Kleiner." So they were basically going
to give the big VC pitch that ultimately led to one
of their big funding rounds. And then everyone in the
company went with them, because it was such a
small company at the time. Everyone went with them, and I remember the office
manager came back in, Heather, and she said "I'm sorry, I know it was very important for you to complete your interviews today because you were hoping to make a decision in the next week or so, but everyone just left
to go to this VC pitch, so there's no one here to interview you. Would you mind coming back tomorrow?" So, I had to go home and then come back the next day and finish my interviews the next day, and then ultimately decided to go there. It was a fun time. - I think it ended up being 25 million split by Kleiner and Sequoia. - That's right. - Not bad. - I remember they actually
did their press conference over on the second or third
floor of the Gates Building. Because I remember I had
just signed my offer, and then I had seen the word Google handwritten on a note saying "Google Press Conference this way" one morning when I was walking into Gates. That's when I found out they
got their funding. (laughter) - Now you joined Google as an engineer, but I think over time things evolved, and you became really
focused as a product person. Can you talk a little
bit about that evolution? - Yeah. I mean it was very gradual. So what happened was, I
came in as an engineer, and I started working
on a lot of different artificial intelligence
algorithms at first to suggest things like related
queries or related websites, later to match ads to queries. So we basically were doing
sort of a fuzzy matching and broadening algorithm to basically match up some of our early ads. And then along the way we
realized we had a really hard time finding someone to make
user interface changes. And so I helped with trying to recruit one of my friends from Stanford who ultimately turned down the job. He went on to do his PhD instead. - I'm sure he has no regrets whatsoever. (laughter) - And so then they said OK, and we kept interviewing
for about four months we interviewed people because the idea was we didn't have the much UI work to do at Google at the time, so we were looking for a strong systems coder who was willing to do UI work one day a week. It's almost impossible
to find such a person. It was kind of this crazy job description. So after about four months of trying, my boss pulled me aside and said "Hey, I really appreciate
all your help in trying to find the UI/Systems Engineer. Doesn't seem like it's working out. You had this interesting
thing in your background, Symbolic Systems, there's some philosophy and psychology in there. Would you mind spending one day a week making some of the UI changes that we need to make to the site?" And I said "Okay, sure." And then as I took the
assignment and walked away, I remember he stopped me and
said "Oh, and by the way, what we don't need is more opinions. We just need data." And it was like don't try and think about what the site should look like. Just implement what the
data says to implement. And so it was sort of interesting because in that role then every time we would roll out a new feature at Google, I would go and meet with a team and say "Okay, well what does the feature do? How do you want it to look? How do you want it to work?" And then I would go back and I would code it into the front end,
into the web server of Google, exposing the feature. And I did that for a few
years in addition to some of the AI work I was doing on the side, so I did AI for most of the time and then did this UI
thing and it was funny because then about two to
three years into Google, Larry and Sergey went and toured other Silicon Valley companies. They came back and then
said "You know, all these other companies are kind of like ours, and they're structured like ours. But all these other companies have this role called Product Management. And we don't have that, but we have Marissa, Susan and Salar, and the three of them kind of do this. They go around and they meet with people, and they look at how things should work and look and specify features, and none of them, none of us-- It's funny because I think Salar is still there at Google Ventures, Susan obviously runs YouTube today. But they were like "But none of you actually do what your
job descriptions say." Because it was true; even
though I was an engineer, I was spending more than half my time meeting with teams, working on specifications and UI design. I think Susan was in marketing, but she was spending most of her time actually recruiting on partners, figuring out the features that they needed to be able to use Google search, etc. And so, he rounded us up and started the Product Management
discipline at Google, but as I said, for me it felt very gradual because I was kind of already doing that work and that function. - Yeah, and it really
ties in with something that we talk a lot about in this class. Which is, especially at the early stages of these companies, it's really important to bring in generalists, people who are not just going to say "Well, this is my job
description. That's all I do." But people who are going to do whatever needs to be done. One of the interesting things
about Google obviously, is that it grew incredibly rapidly. It's what we call blitzscaling, so companies that are
growing incredibly rapidly, doubling, tripling every year. What were some of the things that happened at Google that sort of reflected that, and what were some of
the inflection points where maybe you had to
do things differently, you had to change your processes or change the organization to accommodate this incredible growth? - Well, I think there were so many interesting scaling lessons at Google. I think that there are a few of them. So, one thing that happened and happened a little bit later, but
when Eric Schmidt came, he really-- I remember we would always have to be reinventing our processes. Just when you would sort of feel like OK, we got it down. We know what the hiring process should be, or we know what the process for promotion should be or deploying code on the site should be, we would go one step further, and then everything would break, and we'd have site outages, or, you know, people would
be getting really frustrated. And one of the things that
he talked about a lot, which I understand is one of
the principals of this class, is that at every-- he was like "Look, at
every order of magnitude, you should expect every process to break, and you should expect to completely have to reinvent it." So, he's like, It's very different to deal with tens of people versus hundreds of people versus
thousands of people, managing tens, hundreds, thousands. And each time you cross
over one of those barriers, it's very likely that
the system that you used very successfully for the previous stage is going to break, and you're going to have to rethink how do we do this, how do we train people up to deploy code on this site safely, how do we want to scale our hiring process to make sure we keep our quality bar where it is. But it's also important to recognize, one of the things I think
that Eric did at first, and it really frustrated
us, but when Eric-- Eric came in around
March, and we had a plan. We had to close the previous
year at about 200 people, and we had a plan to get to 400 people, so basically double
the size of the company over the coming year. Eric showed up in March,
and he looked at the plan. He just said "There's
just no way you guys are going to be able to double
the number of employees. Obviously, we were going to
more than double traffic, more than double revenues;
there's no way you're going to be able to double
the number of employees and really keep the quality, the culture, the way that you want the company to be, so I'm going to let the
company collectively hire 50 people this year." So, he basically took our hiring plan and scaled it down by about a factor of four. It was actually quite funny. He created these like little dollar bills with Larry and Sergey's faces on them, and he called them Larrys and Sergeys, and he laminated them. He had his assistant laminate them. So, he had these cards and he said "Look for me, and every time you want to hire someone you have to present a hiring packet and a resume, and you have to hand it in with
a Larry and Sergey." And so he took the Larrys and Sergeys, and he distributed them
out to the different VPs, and then of course there became a black market for Larrys and Sergeys, which actually became
surprisingly efficient. Because what would happen is, the head of sales would have one, and he really would need,
in order to make a sale, he would need an engineer
to build something for him, so he'd be like "Look.
I'm going to give you this Larry and Sergey, but you have to promise you're going to use it to hire someone who's going to build this feature to secure this revenue etc." And so it actually became
really interesting, and I think, as painful as it was because we were just like
"Wait, there's just way too much work here to slow down
our hiring like this." It was actually a really
good moment for the company because it made us be
really thoughtful about how we were scaling, where we
were putting our resources. We had to be that much more thoughtful about where to prioritize, and
where the opportunities were. And so, yes hyper growth is really fun, but you also want to realize when you want that hyper growth to happen
in terms of users and revenue, and not necessarily in terms
of the size of the company. - [Chris] One of the
interesting things is, you institute something
like the Larry and Sergey black market system, but
then at some point in time you decide, presumably you guys
decided to stop doing that. I assume there aren't
still laminated Larrys and Sergeys being handed
out, so how did that happen? - Eric was always very principled, and we did annual planning cycles, and so we ran that way through
the remainder of the year. So, I think the last Larry
and Sergey got turned in right around Christmas or New Years, and then the next year we
went with a formal hiring plan that we had all agreed to as part of our overall strategic process. It was sort of an interesting
way for him to come and add his own bit to Google's
culture that very first year. - Right, and so there's
also a natural expiration of the program as well. Now, speaking of programs at Google, one of the most successful
programs at Google is the APM program which you launched. It's now become probably
one of the most incredible sources of talent in Silicon Valley, and I think a lot of the
students in the classroom are interested in that program. Can you talk about where
that program came from? Why hadn't somebody done
something like that before, and what made it so special? - Again, it was one of these things where it was just sort of organic. So, what happened was
in, I think it was 2002, Jonathan Rosenberg joined, so we had just formed the
Product Management Group the previous summer, and Jonathan joined eight or nine months later. So this Product Management
Group was really small. It was me doing various consumer work, basically everything
you saw in Google.com, Susan doing work with partners, and Salar doing work with advertisers. So each of us had our own
different constituent group that we were designing
products and features for. And at the time we had more
than 100 engineers etc., and Jonathan was hired as
the VP of Product Management, so he became all of our
boss, and we started working with him to try and recruit
new product managers in. And we just had this
very interesting issue that we needed people who
were really technical, because the Google engineers
really pride themselves on being very technical
and not really wanting to interact with people who didn't deeply understand the technology. Jonathan, of course, was a really experienced product
manager and had his view of who makes a good product manager. But the end game of it was that Jonathan had been there
for about four months, and we had hired two people
into Product Management. We had hired John Piscitello
and Pearl Renaker, so there was basically
a team of five of us. At the same time, Wayne Rosing
who was running Engineering, had hired eight people a week
for the last eight weeks. So he had hired like 64 people, and one of the things you
learn is that you actually want to scale engineering
and product management in a certain ratio. It basically ends up,
depending on how complicated the products are, you want to kind of hold product management at an eight to one or like a 12 to one ratio, 12 engineers, eight to 12 engineers
for each product manager. And at that kind of rate,
we're basically getting one new product manager
for every 32 new engineers, and it's very hard to keep up. And further, which was
very upsetting to me, Jonathan took, I think
he took John and assigned John to work with Salar,
and he took Pearl, and assigned Pearl to work with Susan, and I had no one to work with,
and so I went to Jonathan, and I said "Okay, well, this
is great, but when do I get some help on the product management?" And he said "Well you
don't have any revenue." And I was like "I don't have any revenue?" Because he was like "Well,
Susan does partners, and Salaar does advertisers.
You don't have any revenue." And I said "But if we
didn't have any users using the site, if we didn't
have any users on Google.com, we would have no revenue. There would be no revenue opportunities." And, I was like "Okay" and I just decided I wasn't going to get frustrated about it, and I said "Okay. I want to bet." And he said "Well, what
do you want to bet?" And I said "I want to bet
that I can hire and grow new product managers
better and faster than you can hire experienced ones." So, I was like "I want the
license to be able to go and hire some people to help me, and I'm more than happy,
if we hire good people, I'm more than happy to share the wealth and spread them around through the whole product management group, but let me go." And he said "Okay, I'll take that bet, but what are you going to do?" And I said "Well, I'm going to go and hire great computer science and
Symbolic Systems majors and computer science-related
majors out of school who haven't been product
managers anywhere. And I'm going to teach them to be great product managers in the Google style. We'll give them a lot of mentoring and we'll kind of follow the template." Salar and I both came from
Stanford right into Google, and there it was kind
of a baptism-by-fire. We got a lot of responsibility, we got big projects to work on, and Larry and Sergey just yelled at us until we became the people
that we needed to be to do those jobs well, and so I said "I'm just going to repeat that process." - Very technical management process. - I'm going to just repeat that process and hopefully with less yelling. (laughs) And so, he said "Okay" and
within about a month or so I had hired the first APM, Brian Rakowski, who now runs the program, and things just took off from there. We did a lot of different things in terms of helping build a community across the APMs, build a rotation, because we fully acknowledged and expected that, like a lot of people
when they are graduating, you don't know exactly
what you want to do. And so the odds that you get assigned the perfect job inside of Google is low, and so we said OK. One of the scary things about that is usually for you to try different types of product management,
consumer product management. advertiser product management,
publisher product management, enterprise product management, you'd have to change companies a lot. We actually said look, that's
one of the great things about Google is you can actually
stay at the same company and try a lot of these different roles. So we're going to make that really easy for the associates early
on in their career, and then, after they
graduate from the program, we decided we'd make
it a two-year program. Then they can go and work on what they ultimately want to work on, and end up in the group
they want to be in. They can stay in their
group or they can move on. Yeah. - [Voiceover] Can you talk about
some of the personal traits that you selected for with
that initial cohort of APMs? - Sure, so what did we select
for for the first cohort. - So one quick thing, because
we don't have the microphone, we'll just repeat the question. - Yeah so, I was just going to paraphrase. So the question was what
traits do we look for, personal traits do we look for
in the first class of APMs. So I wanted people who are
really technically excellent, so that was very clear,
but there were a bunch of other traits and things
that I was looking for. One, I wanted people who really understood how to apply technology. Not just people who are
great at coding technology, but who would basically say,
"Hey I see this trend happening with FM transmitters, and I
think it means that we're going to be able to do X, Y, or Z." And so we designed a lot
of questions to ask people, about like you know what's
the coolest thing you've seen in the last six months, and
what long term implications do you think that it has,
and people who got good at answering those kinds of questions, ended up being better APMs than people who might be able to tell
you just very technically what's possible or not possible. The other thing is, we
had to acknowledge that we were hiring people who
were really inexperienced in this role, but they were
yet taking a leadership role. And so you needed somebody
who was very humble, and also was a really good listener. And the humble piece comes in because, to really win the
respect of the engineers, and people that they were working with, it was a matter of saying well
wait, this isn't just about you walking in and running
a meeting on day one, that's not going to happen, the other people on your
team are 15, 20-year, very experienced, very
accomplished engineers. The right way to win them over is to say, "Hey, sure, of course I'll
take notes for this meeting, of course I'll schedule the meeting." Like, "Oh of course, you
need me to go and get you some machines to expand your
code or expand your test on, I'll go and apply to get the hardware and wrangle some of the processes." And so, we hired people who were willing to sort of roll up their sleeves
and get their hands dirty in terms of the overall process. And really help listen to the engineers, because the engineers had
some of the best ideas of where the products should go. And then also really winning people over by their ability to help
the team get organized. But also, because of
the lack of experience, they would have to be very data-driven. Because the thing is you
can't make your point in terms of how you want to evolve a product based on instinct. So I've seen this work elsewhere, I've seen a feature be
added here or there, the way a lot of times
they would have to win over those engineers was by
being very data-driven and saying look I've looked at the logs, it looks like all of our users do this and then they follow on and they do that, and therefore we think
we should have a feature that makes that follow-on step easy. Things like that, and
so basically listening really well, being humble,
being very data-driven and being really technically excellent were probably the four personality traits we looked for the most. But I do remember, because to
the FM transmitter example, there was one night,
when I forget if it was Larry or Sergey that
came down and had dinner with the first class of APMs. And we were all sitting around a table, I was sitting on one end, they were sitting on the other. And I think maybe it was Larry, he said "Lately I've been
doing all these experiments with FM transmitters to try
and understand which one can broadcast my iPod
from sitting in the trunk to the radio in my car
the best, and which one has the highest range and all this--" And it was really
interesting because it was sort of this dinner
party, but you can imagine normally at a dinner if someone
said something like that people might lean back, start
another side conversation, but the moment he said that, it's amazing, because I remember like
all eight or ten people literally leaning in to the table, and saying "Which brands have you tried? Which one's the best? Does it work? Is there interference?" And people just started peppering him with questions about these FM transmitters he had been trying and
how well they worked. And it was really interesting and obvious to me in that moment that we had picked people
who were really curious. Not just about search, or
what we were doing at Google, but just about applications of
new technologies and trends, how they worked and what
they could ultimately mean in terms of how life was gonna change or what people were gonna try and do with different parts of technology. - It sounds like one of the things that helped make the program what it is is really getting that close attention from senior people like yourself or Larry and Sergey coming down. I think that, and we're skipping ahead a little bit in this sense, but I think you've also instituted
the APM program at Yahoo. Can you talk a little bit about that, and how do you get the
senior people involved there? - Yeah we have, there were
a few APMs from Google who ultimately had also ended up at Yahoo over that period of time and so, they said "Look, it was such a great experience for us and we want to have a great product management discipline here." And Yahoo had a good product management discipline, but we really
felt like we wanted to grow people in that same
way, in that same vein. And so I had enjoyed
building the program so much myself at Google, that I said
OK, that was sort of my turn, so Enrique Torres and
Fernando Delgado stepped up and they were former
Google APMs who stepped up to basically run the APM program at Yahoo. And so they've done it
their own way, and it's fun, and it's really worked nicely there too. - And do you feel like
this is the kind of program where it could work in a
lot of different companies? Is it very special to just Silicon Valley, or do you see broader applicability? - It's actually based
on a lot of different rotational programs that you actually do see in broader business. A bunch of the decisions we made on it, I knew that Viacom, the
parent company of MTV, had a really special rotational program where they would hire
you in, it was two years, you would do six-month rotations through MTV and a bunch of their
different subsidiaries learning a lot about
different parts of business. McKinsey obviously and a lot
of the management consulting firms have a two-year analyst program. And so I had really
drawn on that notion of trying to give people really
broad exposure in a program that was really designed for training, but obviously because
our needs were so acute, you couldn't just have people dedicated to small projects or experimental projects. So we basically were able to give people really big charters so they could immediately start to make a difference while they were learning on the job. - One of the things that has happened, obviously eventually you
spent a lot of time at Google, but you are now the CEO
of Yahoo, big step up, and you must've taken some of the lessons from what you saw with both Eric and Larry running the company. What are some of the things
that you learned from them while they were at Google that
you're now applying today? - Sure, I was really lucky
to have had great mentors and got to have seen such
great business leaders over the years and just in
so many different situations. I was at Google for about 13 years, and I will say not a day
goes by that I don't at least hear them in my head and
in a different scenario, in terms of different
advice they've given. Eric was always an amazing
fountain of insights and advice, you know Larry's
much more instinctual, but sometimes you'd learn so
much from what his instinct was in terms of how he might run a meeting, how he would redirect a product team. And so, there are so many different things that have come with it. But I think one of the key
learnings from Eric is, I came to Yahoo at a moment
where it was very tumultuous. Everyone liked to count how
many CEOs there were in certain periods of time and I always
refused to play that game. Now I'm far enough way from it
that it's not acutely painful but I was basically the
7th CEO in 61 months when I walked in the door. So it had been a very tumultuous time. It became very clear to
me, one of the things that Eric had said which was
a really humbling statement, as you had gotten closer and
closer to the executive level, is he said, "Good executives
confuse themselves when they convince themselves they actually get to do things." (laughs) And he's like, "You don't
get to design web pages, you don't get to design
apps, you don't get to code, right, what you get to
do is set a direction and then your job is defense,
get things out of the way, get the stuff that's
gonna slow the team down, distract the team, get in the way of the team, out of the way." And so he's like, "You
know your job really becomes listening to
them about what's gonna get in the way and how to get
that cleared out of the path so they can move as
fast as they can move." Because you've got to acknowledge yourself that you're not gonna do the designs, you're not gonna do the implementation, you're not gonna do the coding, and so your job is to help
make them as effective as you can make them. And it was funny because
one of the first things that people said to me when I got there, it was my 2nd day on
the job, and they said "OK, so, when are you going to have your big strategy rollout meeting?" And I said, "My what?" And they said, "Well you know
your grand vision for Yahoo. All the new CEOs show
up and sometime between day two and day five, they
call the whole company together and roll out their big
new vision for Yahoo." And I said, "Well I
don't think I'm going to have a big dog and pony
show, I think I'm gonna kind of go and sit in the
cafeteria and listen to people because you guys have all
been here for a lot longer than I have and I've got some
ideas of what we should do, but you've got a lot more
ideas about what we should do, and you know what's
worked and what hasn't, so I'm going to spend a lot
more time listening to you before we do that." So I was like, I think we'll
maybe have a strategy meeting, I had come in 2012, maybe we'll have a 2013 strategy meeting
sometime later this fall, but I'm gonna need your help
in shaping the strategy. I think that that was actually
a much more comfortable way for the people at Yahoo to
accept me, to view me, and like I really view my role there is to listen and then to get things out of the way. And you know, you overall
set some direction but then you ultimately
spend a lot of time really just trying to get rid
of any distractions or things that are gonna ultimately
make the team less effective. - So it sounds like you
were really coming in and trying to learn from
what was already going on, trying to listen, what kinds
of questions did you ask that were most effective at
getting useful information? What can our students learn
from the time that you had? - It's funny because I am kind of shy and I'm sure I did ask a lot of questions, but what is funny is that
there were so many people who were eager to come and
bring their viewpoint. (laughs) And so things you would
learn, I remember I was in the cafeteria one day,
and this guy came up to me and kind of tapped my tray, and said, "Is it time to go?" And I thought he meant go, like leave, like leave the company,
because at the time in that tumultuous period the company had had a lot of attrition, and I said, "No no, please don't go,
please give me a chance I've only been here for a few days." And I was like "We might end up doing something really good here." And he was like "No no no." He's like "Look, we've all been here, we've been here for the
last five, six, seven years, waiting for management and for the board to figure itself out, figure
what direction we should go in, and I mean, is it go time,
is it time we can just run and do some of the stuff
we've always wanted to do?" And I was like "By all
means, go!" (laughs) "Run, don't let me get
in your way, you guys have got some ideas about what to do, more power to you." And so people would come
and bring that perspective and so, you just would
immediately end up in dialogue with them trying to
understand what it was, that if you don't feel like
you're really empowered to get things done right now,
what's getting in your way? Right? And so everyone had
their different perspectives on the different things that
were getting in the way. And one of the things we did
is we rolled out a program we called PB&J, for Process,
Bureaucracy and Jams. Where we basically said, if there's things that are getting in
your way you could go up on this moderator tool and write it down and then other people would
come and vote it up or down as to whether or not that
was really a problem. And it was daunting because I remember someone came up to me and said, "You know, I don't even know
where you're going to start, there are thousands of
things at this company that need to be fixed." And that's a really daunting
thing to have said to you in the first few weeks of a job. But then the interesting thing was, we set up this PB&J and it
really helped us prioritize big things, little
things, things that were getting in the way, things
where people were just like "I don't understand." I'll give you some of the funny examples that we cleaned up and then
there were bigger things. There were parking gates
on the parking lot, and nobody knew why. (laughs) And like, you had to have your key, your badge didn't actually
open the gym door, unless you had taken an
hour-long orientation. And I was like "Really? 'Cause
like any hotel in the world will just let you go and
stand on a treadmill, no orientation needed, I
know we're all engineers, and we're a little clumsy,
but this seems a bit extreme." And so we just started trying
to get some of this stuff where everyone would just be
like, "I don't understand why, why doesn't the gym door work on my badge? I don't understand why. Why
are these turnstiles here?" And so we did a lot of little
symbolic things like that, just where people would report it, and as soon as it got
50 votes, we would go, see if it was a reasonable
thing to change, and then we would just change it. And so we did a lot of
small things like that. But then we also tried
to address things like our product shipping process. How could we launch products better? Because people would say things like, "Hey, it's really
hard to launch something. It's really hard to navigate the process, of how do I take my code and actually get it reliably into production?" And so we would trace that
through different HR processes around hiring people, getting promotions, compensation, different things like that. And so we had PB&J
across the whole company, and what was amazing
is we actually changed, as a result of different things that ended up on that moderator,
a thousand things in the first year of PB&J. Because what happened was it basically started to just direct,
it was an effort run by a person on my team, Patricia, but she basically got
people in each department who were really passionate
about just trying to fix all these little
things along the way. It really empowered everyone to say, "Hey, there's all these things wrong, we can fix them, and actually we can help, and because of all the votes up and down we can actually prioritize the things that are bugging the most people and really getting in the
way of the most people." - One of the things that's
interesting about that and how it relates to other things we've talked about in the class is it's taking a community-based approach using inside the organization,
being able to get a lot more scale out of it by taking that community-based approach. One of the things you
had to do with Yahoo, I think you've referenced
it a couple times, is there had been a lot of turmoil, the company had been around a long time, you really had to come
in and change the tone and change the nature
of what was going on. And PB&J was part of
it, but there were also probably some bigger things that you did to tell people it is go time. Can you talk about the
challenge that you faced in basically turning
such a large battleship and getting it oriented
and set in a new direction? - Sure, I would say I
think that there's sort of a view on culture, the
view that I have anyway, that was actually given to
me by one of my early APMs, Brett Taylor, 'cause
Brett had worked at Google and ultimately later worked at Facebook. And in talking with Brett I said, "What is it like to work
at Facebook after Google? How is it the same, how is it different?" And he said "You know I
can't really articulate it, but both companies have
really strong cultures." And he was like "You know,
one of things you learn is that really strong
companies almost always have really strong cultures." And he's like "When you're at Google, you know you're at Google,
even if you took down all the logos, and changed all the colors, you would still know that you're there, and the same thing is true for Facebook." And I would say the same
thing is true for Yahoo, and you'll find this is true in a lot of really strongly cultured companies. And when you sort of
grok that, you do end up having respect for the fact that culture is something
that's hard to change. And so I think about it a lot just from the medical background,
for me, thinking about it in the form of genetic engineering. And so at a very remedial
level, what you can say is in genes you can get genes
to hyper-express, right, you can get them to over
express, you can turn them off, but it's actually reasonably
hard to inject new mutant genes or new mutant DNA into a DNA strand. And culture really is
the DNA of a company, and so I really felt
that when I came to Yahoo it was very important to
me that we not try and change Yahoo into something else. That it is a great company,
there's great people there, we have a great set of
properties and assets, and we wanted to say, how we do make it the best version of itself, how do we take what's great about the company, their enthusiasm, their sense of fun, even things like our reporting,
the content we create, a lot of these different pieces, how do we get that to hyper-express and really become super
productive and efficient? How do we get some of the things
that are getting in our way as reported by PB&J and people just chatting about what's getting in the way, how do we turn those things off? But it was less about trying
to inject totally new things, and that's why programs
like PB&J worked so well, is because they really came
from the culture of the company and the values of the company. - [Chris] It helped people
express what was already there. - Yeah.
- [Chris] In other words. - And I did see a question over here - [Chris] Yes. - I don't know if you still have it. - [Voiceover] I was more
zooming on what were the roadblockers for shipping
or deploying new products? I was very curious about that. - OK, roadblocks to shipping
and deploying new products, changes to the process
for releasing products. - Yeah, so the question
is, so for roadblockers, I would say there's a
lot of different things that we needed to look at. One, we had a lot of teams
that were under resourced, and/or, lacked sufficient
tenure of somebody having said "Hey, I shipped the last version," or "I know exactly how to get this done." And so one of the things we've done is we've implemented a tech council, which is about 10-12 most
senior engineers and architects at Yahoo to really talk
about how do we wanna do continuous integration, how do we wanna do continuous deployment, how do we wanna do regression testing, how
can we make shipping all that code really smooth? Because we didn't have a lot of processes for that in the beginning. And so we tried to institute
those types of processes but we also did things like,
people didn't even know necessarily who was shipping what when, so we instituted something
called a launch calendar, which we had similarly done at Google. But basically when you're
launching a project you go up on this tool and just say "Hey, I'm planning on launching
this change on this day." And that's actually super helpful for all the other
departments in the company, because that means PR, marketing, legal, everybody knows ultimately
what's shipping, and customer service, if
they start to get emails about a new feature being
seen or not working that well or working really well,
they know what it's about. So we had a centralized bulletin board for those launches to help
people coordinate better. There were just a lot
of different things that we needed to do to ultimately, basically make that
process much more fluid. And there was also just an element of wanting to move quickly,
so we started saying "Decisions in a week.
A good decision today is better than a perfect
decision tomorrow." And we want to just be
able to move quickly in terms of what we deploy. And that really overall
just sped up the cadence, which was important. - As you're trying to make
these changes at Yahoo, obviously part of that
is you're working with an executive team, you're
working with senior leaders. Talk about how you used
your management team to help make changes, how
did you work with them, how would you pick that up
from Google and other sources, where do you look to for management ideas? - I had really liked the
way that Eric ran his staff at Google, we would have a
big staff meeting on Monday, and then we would do
various strategy reviews on Tuesdays and Wednesdays, end up having a lot of one-on-ones on
Thursdays and Fridays, and then cap off the whole week with a big company meeting on Friday afternoon. So I haven't done all of those things, but we do have a staff meeting on Monday where we all get to talk about what did we do the previous week, what do we have going on that week, to really make sure that all
the different departments inside of Yahoo are cross
functionally working together. So if you need something from
another member of E-staff, that Monday meeting is a
great way to kick off the week to say, "Hey, this is what
we're shipping this week," or "This is what we're
trying to get done this week, and I'm really gonna
need some help on deals," or "I'm really gonna need some
help on marketing," or legal, or "I'm really gonna need to make sure that this data center is up and stable for this day and that we don't have any scheduled maintenance," things like that. So we'll have that on Monday
to coordinate everything, then we'll do deeper
dives, sometimes strategic, sometimes technical,
sometimes process-oriented, on Tuesdays and Wednesdays,
and for the first two years, now I do fewer one-on-ones, but
I do do them as we need them I did one-on-ones really religiously with members of the staff, I
tried to do them every week, just to ultimately
understand from each leader what was working for them and what wasn't. And then we added a meeting
on Friday afternoons which was an all-company meeting, where we're really
transparent and we talk about all the new people who've
joined, we talk about people who are passing milestones of being at the company
for 5, 10, 15 years. And it's amazing how many people-- you know there's this phrase
at Yahoo about leading purple, and it really is amazing, almost
every week we have someone who passed the 15-year mark, and the new people who come
to the company are like "Wow," because Silicon Valley it's
very unusual to have someone to have that kind of tenure. But it was funny because,
after they asked me about the big strategic dog and pony show, and I said well I'm not
planning on doing that any time in the first few months, and then a couple days
later I came and said, "OK, well, when do I get
to talk to the employees?" And the person I spoke to said,
"At the quarterly all-hands" And I had heard about
the quarterly all-hands, it was to go over the
quarterly earnings call that the CFO had delivered
the first day that I arrived. And I said, "No no, I get
it that's Tim's meeting, he's gonna go over the
previous quarter's revenue and (inaudible) and all of that,
but like, when do I get to, ya know, talk about
products and take questions from the company, and they were like, "The quarterly all-hands." And I was like, "Oh, I see," because one of the things
people would say to me most often when I was
spending all that time down in the cafeteria,
people would come up to me and they'd be like "It's
amazing that you're down here, because nobody ever talks to us." (laughs) People would say that all the time, and I just couldn't figure out, and I was like, "Well look, communication in a company as big as
Yahoo is really complicated, I'm gonna try and
communicate as best I can," but then you know you can
always communicate more, etc. and I didn't really get it and I was like, "Oh, is that why all the
employees keep saying that no one talks to them?" Because no one ever talks to them, there's only four meetings a year where the executives go
out and actually talk and engage with the employees. And so now we do that on a weekly basis, it's really become a big beloved tradition at Yahoo to use the moderator tool. And we do deep dives on new
products that are coming out and talk about some of the
current events of the day that are affecting the company. And it's overall been really
a great communication tool, but also something to
really bring us together. Because it means we celebrate
what went well together, we talk sometimes about what
hasn't gone that well together, and overall it's been a great
community-building tool, as well as really helping to build transparency in the company. Because I think one of the things that was really important to me was because the company had been
in such a state of chaos, was to try and demystify,
really demystify management and our decision, and demystify the board and their decisions, and by having that kind of transparency,
'cause, if we have a board meeting we actually show the board slides that Friday at FYI. Which is like FYI, for your information or for Yahoo's information. And so we really try and show as much of the decision-making
process, what we're thinking about the decisions that we're making as we possibly can, at those meetings. Yeah. - [Voiceover] Regarding the
PB&J, if there was something that was upvoted, but you couldn't change, because it had business
needs, if that happened, how did you address that when it did? - Yeah, so PB&J, if we
had something that was upvoted but it was something
that we couldn't change due to business needs, we had a response at least in the tool. And so someone would actually
go up and write a response, and say, "You know we've looked at this, we've considered it, here are
the reasons we can't do it." And sometimes people would
re-raise it again later, or raise it in a different forum, and sometimes that next
attempt would work. But, we really try to
change as much as we can, and you really want to be open to questioning your assumptions and changing the overall structures
that are getting in the way. But, yes, there are
sometimes when you can't, but we did feel that one of the things is, we just had various
thresholds, so we were like, "Wait, if 50 people in the
company have voted something up, it deserves at least a response." If that many people chose
to come up to the moderator, we've fluctuated being about
a 10,000 to 12,000 person company, if there's that many people who are taking their
time to come up and say, "This really bugs me,
it's getting in my way," the least you can do is
read it, really consider it, hopefully change it, but
if you can't change it, at least explain so that those
people can come back up later and say, "Ok, that does make sense to me, I understand why that's difficult for us to change right now." - Chris. - So you were saying about
the strategy sessions on Tuesdays and Wednesdays and stuff, what does that actually mean, or what does a good strategy session look like? - So in terms of the strategy sessions, or deep dives that we have
on Tuesday or Wednesday, those tend to be just more free form time. And so sometimes we'll
look at things like, we'll do a deep product review, something new is
shipping, we'll go through the entire product experience, what the metrics for
success are going to be, how we're gonna measure ourselves, what the marketing and
rollout plan looks like, what any legal concerns look like, and we'll do a really
thorough launch review. Then you'll have another
deep dive where it's our People team wanting to talk about what benefits we're gonna
offer for the next year and really going through
the different benefits. So the structures of those
meetings tend to change quite a bit, sometimes they're an hour, sometimes they're an hour and a half, depending on what the topic
is, but it's basically an opportunity for members
of the executive staff to say, "Hey, I have this
topic, I need a decision on something, I'm gonna
bring you the data, into this meeting, any members of E-staff are welcome to attend that meeting, so we usually have really good attendance and participation across
the executive staff in those forums, and then ultimately coming to the decisions that they need made in those meetings, yeah. - A couple of questions,
let's go to this side. - [Voiceover] At one point you banned all remote working at
Yahoo, and you mentioned you didn't want to change
the culture of Yahoo, was that a major shift in culture? - So the question was around
the work-from-home ban that was instituted, and given
that I didn't want to change the culture, how is that
not a cultural change? Which I think is a good question. One of the things that happened, and it's sort of funny,
because, by the way, I have nothing against working from home, my brother works from home,
I have lots of friends who work from home, I've become like the anti-work from home poster girl, it's really not fair.
- [Chris] (laughs) I spent like a good six to
twelve months of my life where every time I left-- I realized I became
kind of a Yahoo hermit, because every time I left Yahoo, people wanted to ask me
about working from home. (laughs) And so, what happened was, it was really just a result of listening. Because I would say I've had
close to three dozen people over the course of my first few months, come up to me, and they would
grab me in the cafeteria, or grab me after FYI, and they'd be like, "I love all the changes
you're making in the company, I love how fast you've
gotten this all moving and the vision that you're painting, and my team is really inspired and we're working really
hard, and once a week we have to stop and call
this person who works on our project, that we never hear
from, and catch them up. And I kept hearing that
narrative again and again, of just like, wait, here we all are, we're all working really
hard, we're high performers, and we've got this person who's kind of letting the team down,
or not that available, and it was actually less interestingly, as I plowed into it, it was
actually less interestingly about the people who
formally worked from home, because I think sometimes when you have a very formal setup for working from home, I actually remember I had a friend who worked at Google for a bunch of years and then moved back to his home in Sweden, and he literally had a
two-bedroom apartment, and one was his bedroom, and
the other was his office, and he would wake up in the morning and get all ready for
work, and then walk across his living room, and spend
the whole day in his office, and at the end of the
night he would walk back and eat dinner in his kitchen, but he had to be very structured about it. And I think people who work like that, actually when they've got a
great set up for themselves at home and they're very
structured about it, it works well. But I think we all, one
of the things you'll learn when you enter the workforce,
is like, there's those days when you're like "OK, I've
gotta wait for the cable guy," or "I've gotta wait for this delivery, I'll just plunk down at the kitchen table and try and get something done." And you feel like you're productive, because there's no
interruptions, etc. etc., but the truth is you're probably, at least in my case, I'm
like 50-75% as productive in that state as I am
when I'm at the office. And so we actually, I
found, had less of a problem with working from home formally,
where there were people who worked from home five days
a week and had a great setup, but we had lots of people
who would just be like, "Oh it's raining really hard,
and traffic's bad this morning so I'm not gonna come into the office." And they would just mail
their team in the morning, and be like "Working from home today, don't want to lose an hour in my commute," but the point is, that
actually you're gonna now be interfacing with seven
other people on your team all of whom either have to keep
stopping to write you emails or call you to talk to you,
and it's just not as efficient as if you were sitting
at the next desk over, or at the next cube over. And so that was really what
we needed to put a stop to, it was interesting because we
granted a lot of exceptions to the work-from-home ban,
if you wanna call it that, where we kept letting
people work from home because they did have really
formal and good setups, but we did want to
basically send the message that this was Yahoo's moment,
and it wasn't the right thing for all companies, but it
was the right thing for us right then, to have people in the office so we could collaborate better, have less of these friction points, and it was more about
sending the message to people who were kind of casually or
occasionally working from home and not being that productive about it, that that's probably not
the best thing to do, at least at this moment
in Yahoo's history. - And you were changing the default, it wasn't that you suddenly said nobody can work from home, and if you work from home you're fired. It was, the default is, we
expect you to be in the office. The other thing I
learned from this is that even the CEO of Yahoo still
has to wait for the cable guy, which is a very sad
statement on the world. (laughter) We have other questions, let's go all the way to the back there. - [Voiceover] On the point of
culture and hiring and people, one thing that people are
shifting to understand is that when you go out with your company one thing that it's
really really nice to have more than anything else is
the people you hire, right? And so when you go into a company as CEO, you did not have the
opportunity to be that intentional and deliberate (inaudible) And so there's all these stories
like this Larry and Sergey (inaudible) 15 times or
whatever, so what do you do in that position when
you're taking over a ship and the ship comes with
the package, right? So, how do you go about that? - Well you do your best to meet everyone-- - Quick quick-- - Oh sorry, I'll repeat
the question, or you can. OK so I would say, the question is, I mean one of the most important
things you can do is hire, and when you're starting
a company from scratch you get to hire who
you want and you get to shape it in a particular way. When you're coming into
an already formed company, it feels like it's already a
package that's well established how do you handle that? And, a couple different things, one, we did change the
overall hiring practice, because there's different
ways you can hire, one of the things that was happening, and I would just hear it again and again in the narratives from different employees who would talk to me is, managers could hire who they wanted, it was not a committee or a
wisdom of crowds based approach. Also on the annual compensation elements, there were basically
no performance reviews, and you could give raises and bonuses to who you wanted to give
raises and bonuses to. Which basically meant there
was kind of a lot of cronyism, for lack of a better word,
because you had people hiring their friends and
then giving those friends promotions, raises, bonuses, etc. And so we had to really think about how did we want to approach performance, like maybe we should
have performance reviews. Like, when the company's
performance isn't that great, one of the things you may say is wait, the company is just comprised
of a lot of individuals, so maybe we should actually
do our own performance reviews and keep track, so we set
up a goal system for across the company, we set up a
performance review system for employees to give people feedback. And I also set up a hiring review process, which was, my view was
I still want managers to be able to hire
whoever they wanna hire, but if it's really a good hire, you should be able to justify it. Does that make sense? It doesn't mean that you
should hire someone different than you otherwise would have hired, it just means that if somebody asked you, "Hey, how come you're hiring person A?" You should have a really good reason, as opposed to, because I want to. And so that hiring process, in
my view didn't really change how many people we hired,
or even who we hired, but it did change the
level of scrutiny on it, and the level of thought
that managers would give to why are we hiring
the people we're hiring, what are they gonna do, and ultimately, what will they help us
achieve when they got there? And the other thing is,
companies are living organisms that change over time, you
know people come and go, especially in tech, there's
a lot of transitions that people make, and so
from that perspective, you'll know over time that you will get to ultimately shape the culture. Today at Yahoo actually we've hired, we have a company of about 10,700 people, and about 6,000 of them,
maybe just more than that, are newer than I am. Right, so the company is
already more than half people who have been hired
in through this process or into this vision, into this direction that we're currently taking the company. It's sort of interesting, 'cause
you can almost think of it, I remember reading
somewhere that every cell in the human body, I
don't know if it's true, but every cell in the
human body turns over every seven years, so every seven years you're an entirely new person. (laughs) And you know, that's kind
of true for companies too, there'll always be some
people who are the same, but by and large there
is that kind of turnover, and if you approach
that in a principled way and really try and manage
it, it can be something that can either be very
good for the company or very bad for the company. And so we've tried to make
that a really positive process of renewal, of getting people
who are really motivated to come and work on
returning Yahoo to greatness. - One or two more scale
questions and we'll also return to the audience and alternate. One of the things that
you've done at Yahoo is you've used M&A, you've
acquired a number of companies, small companies, large acquisitions like the Tumblr acquisition. Could you talk a little bit
about what's worked well about that and maybe if there's things that haven't worked well, and
again, all backwards-looking, obviously you're running a public company, we don't need to know about
anything forward-looking. - We've done a lot of
acquisitions at Yahoo now, several dozen, and I would
say that one of the things that we've done is we've kind of classified them in three different groups. We have talent acquisitions,
we have building blocks, and then we have strategic acquisitions. And they roughly map to size and scale, but we did each kind of
acquisition in a different way. So, talent acquisitions, one
of the things that happened is, and I know because when the
board was recruiting me, it was really important to me. One of the things that happened is, we knew we had to hire new
people into the company, but everyone I talked to would be like, "I would love to come into Yahoo, I would love to work on these projects, but I don't want to come by myself." And I understood that,
because I didn't want to come by myself either, right? I was like, wait, if you're
gonna hire me in here, I'm gonna have to be able
to hire some of the people that I know the company needs to hire, and if I meet new people who
have the right skill sets, etc. And so one of the interesting things about talent acquisitions is it
works really well for us, because we could bring
in really terrific people and they would say, "OK,
but I don't want to come by myself," and we'd actually
be able to kind of group hire, four or five people in
small sets into the company. And the nice thing about
it is because they were already working as a team, they could hit the ground running really quickly. So one of the things we saw, for example, was we had about 30 people
who worked on mobile in a 12,000 person company,
14,000 person company, the day that I joined that
was one of my interesting cafeteria conversations that we had, where I had a real heart
attack in that moment. Because I had just sold
the board on let's remake Yahoo for mobile, and in
the mobile generation, and then I met this guy
Tony, and I was like, "What do you work on?" and he was like, "I'm a mobile engineer." and I was like, "We have a mobile team? That's
great, how many are there?" And he was like "30," and I
was like "Oh my gosh!" (laughs) And so we need that 30 to be like 500 people really really fast. And one of the ways we
did that was we would go and do talent acquisitions
where the team had built a beautiful application,
but it really wasn't getting the size or scale that they
wanted it to, but they were still great at building
mobile applications, and they knew how to work
together really well as a team. So we acquired a lot of teams and put them right into the mobile sphere,
and it really helped us reinvent our app strategy
and get new apps out quickly, because these teams could move so quickly. The other thing is Yahoo now
as a 20-year-old company, has a lot of technology that's quite old. And so it's not that unusual for you to go and dig in to a project's tech stack, and find that, wait, some of
this code is 10 years old, 12 years old, 15 years
old, hasn't been updated, some of the people who know how it works or could easily explain how it works aren't even there anymore. And so some of the
strategic building blocks that we brought in would
be basically companies where we not only wanted the people, but we wanted a key piece of technology. So for example, there is a
company we bought called Xobni, which is inbox backward,
and they had been working on reinventing mail, but one of the things they got really good at was contacts, they were constantly
parsing all your emails so when you got something that said, "Hey, my phone number's changed," or they saw in the footer that the person had a new title or a new email address in the footers or headers of the email, they would ultimately immediately
update it dynamically. And so we were like, this is great, because if we can just
take this technology and just replace the contact address book in Yahoo mail with this new,
more modern, easier to maintain piece of technology, it's
a very clean swap out. Not only do we get great
people, but we get technology that's much more scalable, more modern and easier to maintain. And then we also wanted to be able to make some large strategic acquisitions, Tumblr, obviously at about $1.1 billion, BrightRoll which is a
video ad network we bought, and Flurry, are the three
biggest acquisitions we've done. And there, our goal is
can we buy a company that does something that
relates to what Yahoo does, but pushes us in a new direction. Like Tumblr pushed us into social. Yahoo's always been very
strong at display advertising which for many people
is banner ads, but now banner ads don't work as
well as they once did, they're still very
effective, but they sometimes don't work as well, and
the new mode is much more storytelling through
video, which is why you see so much more video advertising. We said, wait, if we want
to keep kind of moving in the vein that we're in,
being a really strong offering for brand advertisers for
them to get their message out, display 2.0 is video, so how
can we buy a company that really moves us strongly into video and BrightRoll had the largest video ad network in North America. And so we were really excited to do that. And Flurry really reinforced, they were one of the largest mobile
analytics platforms, and they also are looking
at not only how people analyze their apps and
optimize their apps, but also how to monetize their apps. And so there was a
great opportunity there. But we've come up with this term MaVeNS, mobile, video, native, and social, it being sort of an acroynm for that. And we really want the
strategic acquisitions to push us forward hopefully
in at least one of those areas, if not multiple of
those areas, in terms of how they push us forward,
how they push the business. - Got it, OK. Other questions? Maybe from this side, front row. - [Voiceover] Yea you talked
about the strong culture in Google and you obviously
spent over a decade in that company so how do you make sure that you didn't try to clone that culture and how did you make sure
that you helped shape a culture that was unique to Yahoo? - [Chris] Got it. - Sure, so the question is,
since I had been at Google for so long, how did I make
sure that when I came to Yahoo I didn't try and clone the culture. And in truth I'm sure there
are things that I brought over, you know consciously or subconsciously, but there's a couple of things. One is, I loved Google
and I loved the company and I loved my time there and the people I was able to work with, but you know, you always have some of
your own perspectives. Where you're like, I
liked this part of it, but I would've done it this way, or I liked this part of
it, and I would've done it this way and I think that
would've been better. And so there certainly
were some ideas that I had where I was excited to
take them, modify them, to incorporate an insight that
I had and see how they worked. And so I did some of that,
but the other thing is you always want to work somewhere where you're really
excited about the people and the culture, I mean
that's what kept me at Google for as long as I was there. Because one of the things that happens is, if you're working somewhere
for the right reasons, you believe in the mission of the company, you believe in the people of the company, the last thing you wanna
do as a new entrant into it, even as the
CEO, is change it, right? And for me, you know, Yahoo had been just such a vaunted enterprise, right? I remember I was here at Stanford, I remember someone telling me, "Hey, there's this guy named Jerry," I think his username was
literally Jerry, J-E-R-R-Y, and you would go to
~Jerry and he's keeping a cool list of notes of fun websites he's finding on the web. (laughs) And so before Yahoo was even in HTML, I remember going and poking around his world readable
directory to try and find cool URLs, and so then
I watched it become, Jerry's guide to the world wide web, and then Dave and Jerry's
guide to the world wide web, and then ultimately Yahoo, and
then launching as a company. But I had seen so many parts of that, and then at Google, for
some of our first years, one of my first assignments was help us win the Yahoo contract, because we had a joint board member, he had gotten us some
ins and some meetings with Yahoo, and I remember going to Yahoo, and having to show off
some of the AI algorithms and things that I was working on, to try and show them why
Google search might be a better alternative than some
of the other search providers they were working with. And so I'd had this all in my history, and so I had really
looked up to the company for a long time, I had a lot
friends who'd worked there, had a lot of respect for
it, and you know for me I think I had enough enthusiasm for wanting to be a part of that company and culture, and understand it, that the last thing you wanna do is disrespect it or change it. - Two more questions that we prepared, and then we'll throw it
all open to the students. Because there are two questions
we always try to ask folks. The first question is, obviously
you're very busy as a CEO, but how do you invest in
yourself, what do you find the highest ROI, is it reading
books, do you have a coach, is it meditation, people
do different things, what do you do to invest in yourself? - Well I do all kinds of different things. So I mean I have a
three-year-old son who's amazing, I spend a lot of time with him, I love to ski, I love to travel, but one of my core philosophies is, and I saw this early at Google, it's sort of funny, because
a lot of people think that Google just happened. Right, they're like "It was amazing, it just grew like wildfire, no one's every seen anything like it." And I assure you the experience of being inside of it
was nothing like that. (laughs) OK? I mean it was like, can you
actually work 130 hour weeks? There's only 168 hours in
a week, and the answer is, if you were very strategic
about when you shower, eat, and sleep, yes you
can work 130 hours a week, but you won't have time to
drive home and back and forth, so you better move all
your clothes to the office. That was the experience of trying to grow Google during those ages. It's funny, I actually have a friend who has a co-working space in the city, and I want to the co-working space on a Saturday afternoon,
there's a bunch of different companies there,
there was no one there. On a Saturday afternoon, and I was like, "This does not bode well at all for any of the companies in this space." Because, Sundays were
really hopping at Google, but Saturdays it was very
rare that there weren't at least dozens of people in the office trying to get something done. Because we felt such incredible urgency, and there was just so much to get done, and just such an opportunity, that it was a lot of hard work. And as a result, people thought
a lot about burning out, and you know, can we keep going
at this pace and this rate? And what I saw, both in
myself and in other people, was that we were able
to work incredibly hard for incredibly long periods of time, but one, you have to be really passionate about what you're working on, and two, you can't have everything you want, but it's OK to carve out something that really matters to you. So, I have a good friend, Craig, who was the first employee
Larry and Sergey hired, Craig Silverstein, he
was their first employee. And Craig had this rule,
I remember 'cause I had to usually plan my code reviews around it, that on the last day of every month, no matter what was happening, he left the office no
later than 7:00. (laughs) And I remember because I would be like, "Oh no, it's the last day of the month, I'm not gonna be able to get a
10:00pm code review tonight." (laughs) And there were just
little things like that, and I noticed that the
people who were the happiest, the sanest, able to work for the longest, kind of would have that. And I started calling
this finding your rhythm. And a lot of people talk about balance, and when they think about balance, and all of you guys, you
know how hard college is, and how much you have to push yourself, but this notion of like
three square meals a day, eight hours of sleep a night, I was like, you know that's
what some people need, but it's not what everyone needs. And it's much more about what do you need in order to not feel resentful? And so I started asking
people who I was worried about burning out, like, what's your rhythm? And I got very interesting answers back. There was Nathan who had
recently graduated from Stanford, and I said "Look, I appreciate
how hard you've been working, but I really want you to think
about what matters to you, and when do you get resentful?" And he came back with this
very interesting answer, and he was like "Tuesday night dinner." And he was like, "Every
Tuesday night my Draw mates and I have a potluck at
one of our apartments." The old Stanford Draw
group would get together. And he's like, "And if I
miss Tuesday night dinner, or worse if I have to cancel it when it's supposed to be at my apartment, I'm just bummed the
whole rest of the week." And he's like "That's
when I would find myself on Wednesday and Thursday, I'd be like, well I worked this hard
and I didn't even get to go to Tuesday night dinner this
week, so I'm leaving at 5:00!" And you'd have that kind
of perspective of, wait, it's OK for you to have
the view of like, wait, you've given a lot of
yourself to this company, but you should be able to say wait, I can't have everything I
want, and all the leisure time that I want, but I can have the things that really matter to me. I had a mom of three,
Katie, who was working a lot with our Bangalore
office, and she kept doing conference calls every night
at like 1:00 in the morning until like 2:00, 3:00am, and
I was really worried about it, and I said "You know,
Katie, just explain to me what your rhythm is," and she said, "Marissa don't worry about
me, I don't mind the Bangalore conference calls in the
middle of the night at all, what bugs me is when I'm
late to the piano recital, or I miss the soccer game,
and that disappointed look on my child's face when I walk in late." And I was like, "OK,
then Katie we're going to make sure you're never late
to another kids' thing again." And it was amazing because
she would just tell me in the morning, "Hey at 4:00
today I've gotta leave." And it was always amazing, because at 3:45 Katie would stand up to leave,
and someone would be like "Oh Katie, we're almost
done, can you just stay for five more minutes?" And I'd be like, "Nope, Katie's gotta go." And I will say just in
myself, I've watched my rhythm change, because
in my 20s and early 30s, before I was married,
for me I really wanted to see the world and it was all about travel. And I would find, about
once every four months I wanted to go somewhere new, I wanted to be out of the office for at least a week, and it was actually very good for me, because I would have week
where I would be like, oh, I was gone for a whole week and everything kept running smoothly, and it was also good for my team, because they'd be like,
and she was gone for a week and nothing went really that wrong, right? And I found if I pushed myself
to like six, eight months, and I didn't go on a trip
I would start feeling really resentful, like
yeah, I really wanted to go to Iceland, or I really wanted to go here, and I'd have to keep canceling
and postponing the trip and it would kind of just get to me, and I'd be like OK, I need
to ultimately do that. And now for me it's a lot
more about family time and having the things
that really matter to you, and that will change
over time, which is why I tend to not be too structured about it. Because I think if you're
really structured about it, you can sometimes cling to
something where you're like, yes, this is part of my routine,
this is part of what I need to do to invest in myself,
but what you really need to stay rejuvenated and not get resentful has actually changed over time. - And one of the interesting
things that I think came out of that discussion
that you just had, was that you have to ask people, right? It's not enough, you can't count on people necessarily just coming to you and saying, "This is what I need," you had to actually go to them and enforce it. Final question, we often ask this. If you could go back in
time to that day on 2012, first day on the job, what
would you tell yourself to do differently, what
have you learned since then that you're like oh I wish I knew it then? - There's so many different things. It's funny 'cause there
definitely are things where I would say go faster. Right, there's definitely things here where I'm like wow, I can't
believe I've been here for three years and we haven't changed this or fixed this yet. So there's some areas where I would tell myself to go faster. And there's some areas where I would tell myself to go slower. - [Chris] What's one of those areas? - Well, in retrospect, I
don't like the word turnaround I prefer to think of it
as Yahoo's renaissance. But, we are in a mode of turning around, and one of the things
that happens as a new CEO is that people sort of
expect that some time in the first year or
two that you're going to basically bring in your own team and exchange that executive
staff to really be your own. And I did that in pretty much
about four to six months. And I would argue that
was probably too fast, and I did make a few hiring
mistakes in that time, some of which have been very public, and so you really have to live with those decisions in a very public way. And they were really big learnings for me, but I would say in some of those cases, it would have been good to have just gone a little bit slower. So in some cases I think for me if I could go back to myself in 2012 it would be a little bit
more about different pacings. Some things that should
have moved even faster, and some things that should
have moved more slowly. - Got it. Excellent. You
had a question, I know. - [Voiceover] Yeah I did. I'm actually asking this (inaudible) - That's OK. - [Voiceover] Things have changed. So, you have the perspective of someone who went straight to a small company right out of school, but
you've also seen an APM program (inaudible) for many
years, what do you think are the trade-offs you get from those different kinds of experiences,
and what kinds of skills do you think each builds
(inaudible) in a person? - Sure, so the question is, you know, being at a small company,
being at a large company, running the APM program, how do all these different experiences compare and contrast and what do you get from them? Is that a fair characterization? It's funny 'cause it was
actually when Reid asked me to come and talk at the
class it was a comment I reflected on, my mother,
and probably she says it to her children because we
in fact are her children, but my mother will say to
my brother and I, she's like "I loved every phase of my children, I loved babies, I loved toddlers, every phase was more fun than the last." She just had this real enthusiasm for it, where there are some people who'd be like, "Well I really loved this
age, and not that age." And I always just thought
that was very special that my mom said that,
but it was funny for me because I got married later in life, and have had children later in life. You know, for me, I
just had this experience of being at this company. And I realized that for
me, it really was that I loved every phase. I loved it when we were 20 people or less, I loved it when we were dozens of people, I loved it when we were
hundreds of people, I loved it when we were
thousands of people, I loved being an individual contributor, I loved being a person who
managed a single digit team, I loved managing tens,
hundreds, thousands. I loved every phase of it. That said, one of the things you'll see is and it's important to try all
those different experiences. Try small, try big, try running a program, try running a function,
because it really tells you what is it that you like. And I found that I like a
lot of different things, and I like a lot of variety. One of the things I've seen
in some of my colleagues is they'll be like, this is my sweet spot. Like, I love it. It was interesting there were
some business development people early on in our time
at Google who were amazing, and when the company got to be
about 500 people, they left. And I remember talking
to one of my colleagues, and I was like, "I'm just so
devastated that they left." And then as we chatted about
it, other perspectives came up, but they were really very
much jacks of all trades. And they loved it when Google was small, because they could work on inbound deals, outbound deals, licensing
deals, advertising deals, all these different types of deals. As we got to be a bigger company, and they needed to get more specialized, it wasn't as much fun for them. But the funny thing is,
my friend and I both said, "Hey, ya know, if we started
another company tomorrow, we would totally hire that guy to be our lead business development
guy the very first day," because that's his sweet spot. And so I think that when you
try a lot of different things, when you try things you know
you haven't tried before, or you don't feel ready
to do, you learn a lot about yourself, in terms of
the things that you might like, or what some of your strengths are. And so for me I found that I
love all these different phases of companies, I love all these
different phases of scale, and for some people they
may say, "OK, I get it now, I don't like it when it's too small, I don't like it when it's too big, I like it when it's this size." And that's really important,
because that's when you do your best work is
when you're really happy and really enthralled,
not only with the mission but also with the phase
that the company's in. - One final question, back there. - [Voiceover] Hi, so I was wondering, when it comes to decision-making, no matter small or big, like (inaudible) CS, or running
Google, or running Yahoo, what's a drive for you internally, how do you know it's the
right decision to make? - Sure, so this goes back to
my decision to go to Google, but it's a criteria I've used since. So, it was 1999, I actually
think I had a class in this classroom at that time. It was the spring of 1999, it was a heady time in the Valley, I had about 14 job offers,
Google was my 14th offer, and I had a really hard time deciding. Because I had applied to all
kinds of different things, to this (inaudible) phases,
like I'd gone for some management consulting
roles, I'd gone for some teaching roles, I had
gone for some startups, I had gone for some big companies. And I was very good at
picking the creme de la creme of each one, if I work at a
big company it'll be this one, if I work at a startup it'll be this one, if I do the teaching job
it'll clearly be this one. But it was very hard for me to integrate across the different roles. And so, at about Spring Break, about a month before I made my decision, I said "OK, I'm gonna spend Spring Break thinking about the best
decisions I've ever made, radically different decisions,
and what they had in common. And then I'll see how those criteria apply to my job search,"
and so I put on the list of really good decisions:
coming to Stanford, changing my major to Symbolic Systems, spending one summer working at the Stanford Research
Institute up in Menlo Park, the Research Institute
that gave rise to Nuance and a lot of other really
great technologies, Siri, and then one summer
working at the Union Bank in Switzerland, in their research lab. So I kind of looked at
those four decisions, and said OK those are
really different decisions, but what if anything
do they have in common? And I realized that they
had two things in common: one, I had always worked
with the smartest people I could find, because I
think when you work with really smart people they
challenge you they make you think better, they make
you think differently, they make you justify your decisions more, it just ups your game. And I'd always done something I was a little not ready to do. I think if you're really
honest you'll remember being left here for the first day, like I remember that first
night sitting down in my bed and being like, why did I
think this was a good idea? I'm really far from home,
I don't know anyone, this is kind of daunting. I remember going up to, when
I changed to Symbolic Systems, I remember having to call my
parents and say, "OK, I think I'm not going to come become a doctor, I think I'm going to become
a Symbolic Systems major," and I was like, I don't
even know what that is, and I have no idea how to
describe it to my father and explain why this is a
good investment of his money. (laughs) And then I remember going to SRI, and I had literally
only taken CS106A and B, and I somehow thought I could program, even though all I knew
how to program in was C at the time, and so I ended up programming with these legends of
artificial intelligence who had spent 20-25 years
programming in Lisp, and I was like running
around the halls of SRI being like "How do I
get a global variable?" Even though I know you're
not supposed to use them. (laughs) And these old time
programmers being like, "Set env," and I was like "Ahhh!" And then moving to Switzerland, I got there and I didn't speak German, but I knew that I wanted to live and work in a different country,
and everyone in the lab spoke English, and so I
thought it was a good idea, and I got there and they had
set up an apartment for me and my landlady only spoke German. And so there was a lot of hand signals and a lot of directions
about where things were, most of which I could follow. And then she handed me a document that was like 60 pages
long, entirely in German, and asked me to sign it, so I had no idea what I was signing, so I signed it. And then I went to the grocery store, and it turns out in Europe
there's a very different system for buying produce,
where you're supposed to print your own sticker, as
opposed to having someone just wait at the end
of your shopping time. And so I ended up in this conversation I couldn't really have with
this grocery checkout clerk, who was very upset that my
grapes did not have a sticker. (laughs) And she yelled at
me for quite a long time and then finally just
stormed off in a huff and came back with the
grapes with a sticker and pointed at it, and
I was like, oh my gosh, I can't even buy produce here, how did I think this was a good idea? How did I think I could live and work here successfully for a summer? But in each of those
cases, as daunting as those first days felt, I found
that if I pushed through it, and was just like, OK, I don't
really feel like going back tomorrow, but I'm just gonna
get up and go back tomorrow, the most amazing things happened. I had such an amazing
time here at Stanford, and Symbolic Systems, the people
I got to take classes with, and the legacy now of Symbolic
Systems in the Valley, led really by Reid, is just such an amazing thing to be a part of. And I got to do my honors thesis
at SRI in conjunction with the teams there, which was just amazing. And the Union Bank in Switzerland is really what led me to Google. Because if I hadn't written that program following people around the web, that summer, Eric Roberts
never would have said, "Hey, there's these guys
who are working on research really close and similar to yours," And so, in those cases I think
that the two biggest criteria for me are: work with the
smartest people you can find, and do things that you don't
feel entirely ready to do. Because in that moment,
that's how you find out a lot about yourself, and
you can surprise yourself, and be good at things you
didn't think you'd be good at. And at worst you learn your
limitations and you learn things that you're not good at, or that you don't like that much. And so that really guided
my decision to go to Google, it guided a lot of my
career decisions at Google in terms of things that
I tried or wanted to do, and it guided me to Yahoo as well. - Excellent, Marissa, thank you so much for coming into class
today. (audience claps) Thank you everyone.
