Evan Williams cofounded Pyra Labs in 1999. Originally,
Pyra intended to build a web-based project
management tool. Williams developed Blogger to
manage his personal weblog, and it quickly became
an important mechanism for sharing ideas internally
at Pyra.
Once launched publicly, Blogger grew rapidly,
and Pyra Labs decided to focus on it full-time. But
Blogger.com did not generate a lot of revenue at first,
and as the Bubble deflated in 2001, Pyra seemed near
death. Williams remained as the only employee and
managed to bring the company back from the brink. By 2003, Blogger had one
million registered users. That attracted the attention of Google, who made Pyra
their first acquisition. Williams left Google in 2004 to cofound a podcasting
company called Odeo
Livingston: Tell me about how you started Pyra Labs.
Williams: I have always been pretty entrepreneurial, and I had started a couple
of other companies. In late '98 when I decided to start Pyra, I had been doing
Internet stuff for about 5 years. I actually started a company in Nebraska.
I had never even really worked anywhere. I was just totally self-taught technically,
but I started a company and kind of ran it into the ground over 3 years
or so, and it was a very educational, painful experience. But I knew I was going
to do that again. I just always knew I was going to start my own thing.
I went to college, and I dropped out because I didn't need to have a
degree--because I wasn't going to try to get a job with anyone. I came to
California after playing with the Internet for a few years because Nebraska
wasn't the place to be, very clearly.
I moved to California to take a job with O'Reilly, which ended up being very
fortunate as you'll find out later. I worked there for a few months, though I
111
Evan Williams
Cofounder, Pyra Labs
(Blogger.com)
8
CHAPTER
knew I didn't want to work for anyone. I taught myself web development--this
was in the middle of the boom and there was lots of work to be had as a contractor.
I knew I was going to start another company. I just wasn't ready yet. So
I was a web developer on contract for about a year and a half, and worked in
various companies like Intel and HP. Finally I got to the point where I said,
"OK, I am going to start another company." This was very much in the middle
of the boom.
I had visions of raising money and building something cool, but originally
the idea for Pyra was around web-based project management, or collaboration,
which was an area I had been interested in for a long time. The idea for Pyra
was the personal and project information management system: to build projects
for clients around their intranets and help them organize their work and personal
information. It is a web application where you would put your stuff, things
you are thinking about, things you had to do, things you wanted to share with
other people. There is not exactly a corollary to it today, but it is along the same
lines as Basecamp or Ta-da List (but more complicated). There are a lot of
products that are about organizing your work and stuff. That was what I saw as
the big idea, and I had specific ideas about how that could be done better than
it had ever been done before.
Around the time I was thinking about starting the company, I was talking to
a friend of mine, Meg Hourihan. She got excited about the idea and said, "Hey,
let me start it with you." She had been a management consultant and was really
smart, so I said OK. I had been contracting, so I had a little bit of money, so I
could coast for a little while, but we didn't know anybody. We weren't hooked
into the startup scene.
Everyone was getting funded, but it is still completely just a network. You
have to know the right people. Whether it's good times or bad, you have to
know people and you have to talk their language, and we were just from a different
place and not hooked into that at all. So we just said, "OK, here is the
product we are going to build," and we started building it. We actually kept contracting
on the side--I had a contract with HP. That's how we paid the bills--
we turned my personal contract into the company's contract and we did a little
work on that and we did a little work on our project, and that is how we started.
Livingston: What was the point where you really said, "I know this is going to
work and I am going to do this full-time"?
Williams: Well, for me it was always the point of no return. Meg actually kept
her other job, but only for a couple of months. We were pretty hardcore. So we
formed the company and said, "OK, we are building this thing." We hoped to
raise money. We just didn't know how yet. We focused on building the product
first.
Livingston: So, you built the product and then did you have to raise money or
did you keep relying on consulting fees?
Williams: Well, we kind of tried. We started talking to the few people we knew,
but we just didn't have any inroads for that. We wrote a business plan, I think.
112 Founders at Work
The first year was entirely self-funded. It was just doing this work mostly for
HP. HP basically funded Pyra for the first year, unbeknownst to them, because
at the time you could charge a decent amount of money for doing pretty simple
web application development. If one of us was working on that full-time, it
would pay for three of us (not that we were paying ourselves much). We started
working on things in November '98. We technically started the company in
January. Meg started full-time in February, and we hired our first employee,
Paul Bausch, in May. Then we got an office down here in SOMA.
Livingston: So is that when you focused on developing Blogger.com?
Williams: No. We had personal websites and we were web geeks, but those
things were separate. At the time, blogs (or weblogs as everyone called them
back then) were just beginning to be talked about as a distinct thing. There are
those who argue that the first website was a weblog. It didn't really matter,
because early '99 is when people started saying, "OK, I have a weblog." And the
form and community were just sort of developing. Paul and I already had personal
websites for a few years. They weren't blogs; they were just kind of typical
homepages--experiments with web technologies. But we were reading folks
like Dave Winer.
Paul turned his site, onfocus.com, into a blog before I did. Being web app
developers, I think we both wrote our own scripts to do it--basically the same
functionality as Blogger. It seemed like not a big deal at the time, but it did
change my relationship with my website--even with the Web.
Livingston: It was easy?
Williams: It was easy, and that was a key thing for me because I wasn't lacking
the knowledge about how to publish to the Web... For a long time, people
understood Blogger as "It makes it easy to have a website." But a lot of things
before that made it easy to have a website. GeoCities made it easy to have a
website, but they didn't make it easy to publish anything on an ongoing basis.
So, for me, the idea that I could have a thought and I could type in a form and
it would be on my website in a matter of seconds completely transformed the
experience. It was one of those things that, by automating the process, completely
morphed what it was I was doing. If I could have a thought and then put
it on my site, then obviously I am going to potentially do that much more and it
is a stream for communication of a whole different type.
So that was a little bit of an insight. To me it was, "Heck, that's handy." But
it was not dissimilar to what other people were doing with weblogs. They were
either doing it by hand or maybe they wrote their own little script to do it. But
it's the little thing that clicked in my mind: "This is that little tweak that makes
it kind of maybe a big deal." Not that the future lit up in my head and I said,
"We are doing that." It was just sort of a hint, more in retrospect than at the
time.
We took the script I wrote to publish my site, and we made an internal site
where we could do the same thing. So, even when it was only Meg and I, we
had this little internal blog we called "Stuff," and we just put stuff in there.
Evan Williams 113
It was a blog, but it was just, "Here's a thing from a competitor or a potentially
useful page or just information for each other." It was a place where we collected
everything and, as we grew, it was the center of Pyra. It was where things
happened.
So this whole time we are building our real collaboration tool, with all kinds
of structure and big ideas trying to be implemented into it, but we were really
using Stuff a lot. And then Paul wrote a little addition to Stuff so that certain
things we posted to our internal blog we could put on our external company
blog.
We were one of the first companies to have a blog on their site--not that
many people were reading. But it was neat. We were publishing news, random
things we liked, whatever.
This must have been around March of '99, so all of this happened fairly
quickly. That's when I got the idea for Blogger--I know because I registered
the domain then. I totally pictured what it was because it was based on what I
was already doing and then the way we were publishing our own blog to an
external site. I said, "Let's turn that into a product." I have always been a product
guy and am just always thinking about products and thought this would be a
cool little idea.
While it did seem fairly easy to build, it was a dilemma, because one of the
big lessons from my first company was to focus. After my first company died, I
did an inventory of the projects I had worked on in the last year. There were
something like 30 projects that I had started on and not finished. My total
weakness was not focusing on things. So I had this idea and I loved it, but very
clearly we were only three people and we had to contract to pay the bills and we
couldn't start another product. We had this big thing we were trying to do. So it
just kind of sat in the back of my head, but it wouldn't go away. It kept bugging
me. Of course, what made me still think about it was that we were using it for
our own purposes and we were building this collaboration tool, but we were
doing this kind of collaboration with Stuff. We actually said several times,
"Maybe Stuff should just be our product?" And we agreed, "That's too simple,
that is too trivial." And also we didn't have the resources for two products. So
that went on for a long time, and it was in July, I guess, when we finally launched
Pyra, the app, and that actually got a pretty good, if limited, reception.
People started using Pyra and it was in evolving it that I came up with the
justification for why to do Blogger. That was based on the idea that we were
trying to solve a really, really big problem, which is organizing people's information
of all types. We said, "That is too big a problem to start with, so we should
focus it." We decided to focus it on people who were building websites, as a
place for them to collaborate. Then we thought up this architecture where
there would be little mini-apps, and Blogger would be one of those. So, with
that justification (Meg was actually on vacation for a week), Paul and I built
Blogger and launched it while she was gone. Which was a terrible thing to do,
but ultimately a good thing to do--but not a cool thing to do.
Its functionality was really dead simple at first, but it did what we needed it
to do and we already had the script. We thought it would take a couple days--it
114 Founders at Work
turned out to take a week, but we just launched it, while Meg was gone. She
was pissed, of course, rightfully. We launched a whole product, and she's the
cofounder of the company. But we talked her into thinking that it made sense.
"It will be this little thing that won't take any effort. We just push it out the door
and it will attract people to our real thing and we can go back to our real thing."
Livingston: Did it catch on quickly?
Williams: It caught on a lot more than we expected. It was really designed to
appeal to web geeks. It wasn't a mass consumer product. It was, "If you're a
web geek like us, you might find this interesting." It's good to appeal to the
alpha geeks sometimes. I thought it would be pretty cool if 1,000 people used
Blogger. It didn't explode at first because it was fairly technical. You had to have
a website and you had to know what FTP was. You had to know a bunch of stuff,
but things that you would know if you were a web geek. We put it out there and
people started using it and the existing weblogs started pointing to it. Like
Peter Merholz (he is credited with coining the term blog), who pointed to it. It
started getting traction and a lot of people who were like the "cool kids" were
using our product, and we were really excited.
We launched it in August and we had a dilemma on our hands right away, of
course, because we now had a product that people were using, but it wasn't the
"real" product.
The problem was, we didn't see a business in Blogger. This was during the
boom, but we weren't one of these companies that was just, "Let's get eyeballs."
We talked a lot about the stupidity of a lot of the dot-coms and raising too much
money. We were very product driven. We wanted to create cool stuff, and we
wanted it to have a sustainable business. We wanted to probably sell the company
to somebody eventually, but we didn't see any business model with
Blogger. Also, we hadn't raised money, so making money was pretty important.
The other product served a business need and was something we thought
people would pay for. We thought Blogger was this free little thing that would
get people to pay for the real thing. So we very clearly had a dilemma on our
hands: we could focus on the stupid little Blogger app that people were using,
or we could work on our real product. We tried to split our time amongst those
two things and contracted to pay the bills. We were three people, so that was a
little bit difficult. We had endless debates about what to do about that. I think
we ended up doing another rev on Blogger in November that made it much
better, and then people really started using it.
Livingston: Did you start to make money?
Williams: No, not until much later. But we did get wired in, so to speak.
Blogger was how people found out who we were, within a community that was
at first San Francisco-based web design geeks but bled into a lot of different
communities, like Silicon Valley and a lot of leading Internet thinkers. They
were attracted to publishing blogs, and this was a thing you used to do that.
So it got us known a little bit, which was very helpful. For example, we met
Jerry Michalski, who emailed out of the blue and then became an advisor. Jerry
knows everybody and was tremendously helpful.
Evan Williams 115
People were using our other app, too, a little bit, but it wasn't very mature
because it was much more complicated.
In early 2000, we started actually raising money, and O'Reilly invested in us.
They were some of the only people I knew. I guess I left an OK impression on
O'Reilly. I only worked there for seven months as an employee and then
another couple as a contractor doing a completely different job, but left a good
enough impression that I was able to go back there and say, "Hey, look at this
thing." They were aware of Blogger, but we were still doing Pyra, too, and they
agreed to invest.
Livingston: So, you had Blogger out there but you weren't totally focused on it.
Were you worried that competitors, since it was a simple thing, would try to
copy it?
Williams: There were a couple other products out there, but they weren't very
substantial. No one was really paying attention to it. It's hard to fathom now, but
blogs took a really long time to be taken seriously. But, yeah, we felt we needed
to make it a lot better and spend a lot more time on it, and we didn't have the
resources to do that. But at the same time, we didn't think there was a business
there, so we weren't that concerned about it. All the time, of course, we are
debating whether or not there was something there and debating why it was
appealing to people. I thought about it a lot and I came to the conclusion why it
was appealing and the impact it had, and I started to get some insight about its
potential.
I started leaning more and more toward Blogger by late '99. I think Meg
and Paul were pretty much pro-Blogger and I was the one who was still on the
fence. Pyra was my baby and I had all these ideas I wanted to see realized. I felt
the need to focus, but it was also like, "This is the cool thing that's taking off."
I couldn't decide.
The money was actually raised around both. There wasn't a very specific
plan. We had this thing that had buzz and then we had this thing that had all of
this potential. So it was like, "Here's some money, go do whatever." We ended
up not really getting the money until April or May of 2000, which was around
the crash, but (around here anyway) it wasn't like everything was over all of a
sudden. People had faith.
We were still able to get money without a lot to go on. We raised a half of a
million dollars from O'Reilly, Advance.net (Conde Nast's parent company),
Jerry, Meg's parents, John Borthwick from AOL, and Jerry's father-in-law. A
half a million dollars was a ton of money to us at the time. We ramped up to
seven people and shortly thereafter decided we were going to focus on Blogger
and developed it.
Livingston: Do you remember why you finally decided, "OK, we will do this"?
Williams: I had come to the conclusion that blogging was going to dramatically
impact the Web. After I thought about it a lot and saw what people were doing,
I decided that this made tons of sense. The conclusion I came to then was kind
of the one I stuck to, which was that this is going to impact the Web because it
116 Founders at Work
is a native form for this medium, just like all new mediums start out imitating
what came before them and then they kind of find out what they are good for.
We even looked at Blogger and, technically, it was trivial (at least until it
came to scaling it). It wasn't based on any new technology. But that made sense
to me because it was not that the technology was new, it was that we had figured
out this medium, at least one of the native forms of what the Web was good for.
It was about freshness and about frequency, and it was about the democratization
of media and giving power to everybody and the universal desire for personal
expression and the attraction to a real, compelling personal voice. And
hyperlinks. And all of those things were just inevitable forces that were going to
terrifically impact the Web and media in general.
It was kind of the first time that I had started really seriously thinking about
media, and then at the same time Pyra had all these big ideas that were going to
take a really long time to build, and this was much more fun. So I said, "Well,
we can figure out a business. We can charge for pro accounts and we can license
it to companies and we can just make up the obvious businesses around it (even
though they weren't necessarily that strong)."
Livingston: Was it easy to make up businesses around it to make money off
of it?
Williams: It's easy to make up things to write down about how we are going to
make money off of it.
Livingston: Well, how did you make money off of it?
Williams: Well, that didn't come for quite a while later. So, we had raised
money at this point and we decided to focus on Blogger. I wrote the business
plan for Blogger after we raised the money and said, "Here is what we are going
to do." We hired some people. We were seven people in the middle of 2000,
just focusing on all types of things. We redesigned Blogger, with the help of
Derek Powazek, who created the famous orange "B," which was great. It just
kept growing; there were probably hundreds of new users a day.
Livingston: But you weren't charging them?
Williams: No, we weren't charging any money anywhere. And we had all of
these features planned. We had most of the features planned that later became
standard in the blogging world--and some that haven't yet. We were totally
focused on building the product and community around it once we had raised
the money, because this was still, "you get enough eyeballs, you have buzz,
you'll be fine." And the extent of the crash didn't dawn on us that quickly.
I don't think it dawned on a lot of people. We just wanted to build momentum
with this $500,000 and then raise more money later in the year.
At the time it was pretty much the belief that, if you have buzz and you have
users and you have good seed investors, you can raise more money. We said,
"We'll make money, but this is down the road so we don't need to focus on that.
We are going to focus on building more features and getting more users."
We just went on that path, and in the fall we knew that we were running out
of money and started trying to have some conversations with folks. I also wasn't
Evan Williams 117
sophisticated at all about how you do that. We felt connected at that time, but it
wasn't necessarily to the money crowd, and we probably wouldn't have been
able to talk the talk of VCs anyway. So the money wasn't coming. We scrambled
around. We decided we could launch some for-pay stuff.
Other companies at the time were going into enterprises, since companies
had money. At the time it was like, "Consumers aren't spending money. Go to
companies--they're the ones with money." So many companies at the time took
their consumer Internet thing and made it an enterprise Internet thing and
then died anyway. We debated that a lot. We had a good story about why this
was really useful inside companies, and we had a friend at Cisco who wanted to
use it and we got it installed inside of Cisco. It was just a pilot and we started
saying on our site, "We have enterprise Blogger." But there was a lot of pressure
internally, a lot of debate about just doing enterprise, which I was pretty
adamant that we didn't want to do because, whether or not it would make
money, I thought it was pointless. At this time I was very much excited about
the idea of democratizing media and that's what mattered. It mattered more
than the company, really. When you are in that mode, it's hard to say that the
company doesn't matter, since everyone's heart and soul is in it, not to mention
their livelihood.
Livingston: But you're changing the world?
Williams: Yeah. I didn't think we could do enterprise and still do the consumer
site well, even though we had talked about it from the beginning. I sort of realized
later on that, if we do enterprise, we are going to have to focus on enterprise
and the consumer stuff is going to suck and that doesn't sound fun. Also,
we probably won't be good at enterprise, because we don't know how to sell and
service companies. So, we had lots of arguments about that. Then I said, "What
we need to do is charge money from the consumers, just to have a Blogger Pro,"
which was always in the plan, and everyone said, "We can't make money doing
that. No one pays for stuff on the Web."
In late 2000, we built a version with many more features but never felt that
we had it to the point where we could feel comfortable charging money for it.
So, we talked to a couple of companies about merging--private companies who
had funding. We had a couple of serious conversations and came close to doing
a deal. Actually one deal was with Moreover, which did headline aggregation
before there was much RSS out there. It was started by Nick Denton, who is
the guy who runs Gawker Media. Nick and his cofounder, David Galbraith,
were fans of Blogger and wanted to buy us. It wasn't a particularly attractive
offer, but we were on the brink and thought, "Maybe we can get in there and
everyone would have jobs."
Everyone in the company wanted to do the deal but me. But I had conceded,
because I wasn't going to be the asshole who denied the chance everyone
had to still be employed. We were out of money. Fortunately Moreover's
board wouldn't approve the deal. It was some ridiculously lowball offer--it was
something like a million dollars worth of their stock. But they're a private company.
So it was basically like, "We'd give everyone jobs."
118 Founders at Work
Livingston: And your financial future is contingent on them getting acquired or
going public.
Williams: Right. After two years of pouring our hearts and souls into this, it
seemed like a crappy option. Fortunately it didn't work out, though. There was
another company in New York that was a startup but that had some funding we
talked about merging with. And the group that was funding them actually gave
us a bridge investment. They gave us $50,000 while we tried to figure it out.
But that didn't seem like a good deal either. They wanted to do it, but we
decided not to. (That company went away fairly quickly.)
Livingston: Then you ran into Dan Bricklin.
Williams: Yes. So this is December or January of 2001 and the second potential
acquisition hadn't worked out, so basically we got to the point where we sat
everyone down, and I said, "OK, everyone is laid off as of today, including me."
We had warned them a few weeks earlier that we didn't have money in the bank
to meet payroll. Obviously when you are in that state, tensions rise a lot and
morale wasn't good and relationships with my cofounder weren't good. I said, "I
am going to stick around because I took a half of a million dollars of other
people's money. And we have all of these users." (The service was still running.)
This whole time, this service is growing. In terms of users, we were getting
more and more successful. Which also caused other problems in that we
needed more hardware and we had all of these scaling problems. In January,
right around the time that the rest of the company was being laid off, we did
what we called the Server Fund Drive. We posted it on our website and it said,
"Hey, we know Blogger is really slow. It's because we need more hardware. We
don't have the money to buy it. So give us money and we will buy more hardware
and we'll make Blogger faster." Surprisingly, it worked really well. We had
a lot of goodwill and people liked us and we had a good brand within our users
because we were very personable and used the blog and we were just honest.
We just said, "We can't buy hardware, but we have plans and we are not going
to go away if we can get past this hump, so send us some money." So people
sent us money.
Livingston: What was the biggest check you got?
Williams: We used PayPal, and I think we got bigger amounts from fewer
people than we expected. We had several thousand users. 100 people or so gave
larger amounts, and I am not sure what the absolute biggest was. We suggested
$10, $20. Several people gave us $100, and then a company, CMP, which published
Web Techniques magazine, offered to buy us a server outright, up to
$4,000 worth. So between the users and them, we had around $17,000 to spend
on servers, which is more than we had ever spent on servers, so it was a
bonanza. It worked better than we had ever expected.
We told people we were only going to buy hardware, so I wasn't going to use
that to pay people. I just spent that money on hardware, but it got the site back
up and running well and meanwhile we laid everybody off. Meg and I weren't
getting along well at all and she decided to leave and everybody else decided to
leave too.
Evan Williams 119
Livingston: So you had a major difference of opinion?
Williams: Yeah. I think a lot of that came back to the enterprise thing, which
she and some other people felt strongly was our best chance of making money.
If I was the guy in charge and we were dying, it's reasonable to conclude it's my
fault. And certainly there were other things I could have done. So everybody
left but me. (A lot of them needed to leave since we couldn't pay them anymore.)
Everybody left, and the next day, I was the only one who came in the
office.
Livingston: How did you feel that morning?
Williams: That was a really bad time. Actually the day that everyone told me
they were leaving... I told everyone they were laid off and said, "Work with
me if you can." And at the time, everyone had already missed one paycheck,
and they'd had it. These are, of course, my friends, and we were hanging out all
the time and we socialized together, so it's much more than just the employees.
I think that same night I broke up with my girlfriend of 6 months.
Livingston: Sounds pretty grim.
Williams: Yeah, it was just the craziest bad time. The good news about all that
was Blogger was still running and, with no employees, we didn't have expenses.
So we went from having $50,000 a month worth of payroll, to a couple of thousand
for our server infrastructure and our rent. It is probably closer to ten,
between five and ten, but a manageable number, not paying me anything. I
took some money every once in a while to pay rent, and I had long since put all
my money in and credit cards and everything else, but that was actually a much
more reasonable place to be because we didn't have to make $50,000 a month
to pay people. We had to make a few thousand dollars a month.
So then other ideas started being much more feasible, and I was in some
other conversations. Now that we were known, opportunities came up. One of
the first opportunities was a little company called KnowNow, who wanted us to
build something, and later actually two of the people who worked at Pyra,
including Meg, worked for this company, and I did a little deal with them to
build something that was never launched. They killed the project, but it got me
$35,000, which was like months of burn rate at that point.
Shortly after that, in February, I ran into Dan Bricklin. Dan wrote me after
reading my blog. We were pretty public in terms of our communication, so I
posted when everybody left, and I wrote this whole story on my blog that was
pretty widely read, "Here's what happened: everyone's gone. It's just me." I got
a huge outpouring of support from that, and one of the messages was from
Dan Bricklin, who said that he thought what we did was great and he wanted to
help. We ended up meeting at an O'Reilly conference, which was in February
2001. We met and he basically agreed really quickly. He assessed the situation--what
I needed to keep going. (We had a lot of back bills at this point; we
needed to pay our hosting bill to keep the lights on.)
There were some confusing stories about what that deal was. Dan had a
company called Trellix that he later sold, which was a web publishing platform.
120 Founders at Work
Trellix licensed Blogger in order to add blogging to their feature set. They did it
in such a way--Dan drove it in such a way--that if it was a traditional license
(months of due diligence and really figuring out if we wanted this), it wouldn't
have helped and he knew that. So, he was like, "a) there's a legitimate business
reason to do this, but b) we are going to push this through so it is really good for
you." It wasn't a lot of money--it was around $40,000--but with a contract later
on that ended up helping as well. But it was what we needed at the time.
Livingston: So you were back in business?
Williams: Sort of back in business, but both of those deals didn't get me ahead.
They bought me a few months, but between just keeping the service running
and fulfilling on those deals, I didn't have any other time. I wasn't really making
progress, because it was just me. First of all, I had to keep the service running,
which was a really big deal in itself--we had several thousand users and I had to
teach myself Linux system administration and Java, so I could just keep the
servers up and fix bugs here and there. Things would break, and I'd go in and
fix them on the live site and figure it out as I went. That was very time consuming.
The technology wasn't rock solid by any means, and it kept growing and
growing and I didn't want to shut it off. Between that and fulfilling on these
deals, which were mostly giving stuff to other people, I wasn't building in the
real things that were going to make a business. That was a lot of just day-to-day,
by-the-skin-of-my-teeth stuff for several months. Still I had the idea to build a
paid version of Blogger, but that was going to take a lot more development and
work to launch that.
Then there is another part going on around this time that I can't talk too
much about. Suffice it to say my former teammates didn't all go away happy,
and I spent almost as much money on my lawyer in 2001 as I paid myself.
The other thing was that all those people left, and then I was being badmouthed
within this community of people we knew. The story apparently was
that I fired all my friends and I didn't pay them and took over the company. It
was really ugly, and of course we had all these mutual friends and there were
parties we were at. I basically went underground and did nothing but try to
keep Blogger going.
Livingston: There was a whole social component to cofounding a startup with a
friend.
Williams: Which I think is a theme for startups in general because people live
and breathe them and become friends, date and merge their lives together. And
then, if things go bad, it's bad in ways that are much more devastating than your
work going badly.
So that was pretty much 2001. The funny thing about Pyra is that every calendar
year was pretty distinct--'99 was the first year, we were self-funded; 2000
was the year we got money and ramped up; 2001 was the year that it was just
me and it sucked. But somehow by the end of 2001 I started rebuilding. We
cleared up the legal thing, and things were looking up.
Evan Williams 121
Then eventually I started launching some for-pay features of Blogger.
Things that people would actually pay for. So 2 years into it, Blogger itself was
starting to make money--not directly but through some little ways. Like the
blogs we hosted, we had advertising, which never made any money because it
was during the time when web advertising didn't make money. (After it made
money the first time and before it made money again.) I created a mechanism
to charge people to take their ads off and that actually made money. I said, "Pay
me $12 a year, and I'll take the ads off your blog." I started with this "product"
because it was probably the easiest thing I could build that I thought people
would pay for. And they did.
I did a couple other small things like that and got to a point where it was
paying the hosting bill. I had gotten rid of my office by then, and I had no place
to work at home. So I posted on my blog that I needed to rent a desk somewhere.
This company, Bigstep, offered me a free desk, which was nice.
Then I just started building more things. Working from the Bigstep office, I
designed and launched the Blogger API, which didn't make any money, but
became important later. I actually hired a contractor programmer and had
started working with Jason Shellen on business development stuff. So things
were looking up. And then 2002 was a completely different year altogether.
We finally launched Blogger Pro, the paid-for version of Blogger. The paidfor
version of Blogger did very well for us and we brought in some other
people. With Jason's help, we did a big deal in Brazil with this company that
wanted to license Blogger. So 2002 was a ramping-up year again. Everything
was on the uptick, and we had a completely different team. We were getting by
and the money was increasing and we were building new stuff and it was
looking good.
October 2002 was when Google came knocking. We had a small office
downtown--more of a conference room than an office. It was Adaptive Path's
first office, which we moved into after them. And we had brought on a tech
support guy and a sys admin. Then Google called us up. I forget how that
happened... I think that was O'Reilly again.
At this point I think it looked like Pyra came back from the dead. Blogging
in general had exploded all this time. We got a lot more competitors, but the
phenomenon just exploded. We were a less substantial part of blogging, but
blogging was a much bigger deal. So it drove our growth and it legitimized us as
being a major player in an increasingly big space.
So O'Reilly was talking and they said, "I guess Pyra's still alive." We had a
meeting up at O'Reilly around this time, and Tim [O'Reilly] and Mark
[Jacobsen] were trying to figure out how they could help us. One of the suggestions
was to introduce us to folks like Amazon and Google.
Soon after, according to the story I heard, Larry or Sergey were on a call
with Tim and Tim mentioned us, and Sergey had recently been at this conference
where everyone was talking about blogs, and was interested in blogs and
he said, "Yeah, we want to talk to them."
We're like, "Alright. Why?" It didn't even occur to us that they might want
to buy us because Google hadn't bought anybody at this point. And they were a
122 Founders at Work
search company. So we brainstormed all these ideas maybe we could do with
Google and we went down there, and it turned out that we were meeting with
their corporate development people--meaning the people who buy companies.
We started talking about ideas and within the first 5 minutes they said, "Yeah,
there're lots of ideas, but it's hard for someone like us to really partner with
someone so small as you; why don't you just come here and do all that stuff?" So
we were like, "Oh, that's interesting." (We tried to play it cool.)
I had had one or two other conversations with possible acquirers. One was
Lycos back in 2001, which would have been terrible--though it would have
made a lot of sense for us because they had Tripod and Angelfire (two of the
biggest publishing sites that there were). But they didn't have any money for
that area, so that didn't go anywhere.
What Google said was, "Would you consider being acquired?" And we said,
"Well, we've talked to people, but Google's never asked before." Like everyone
else, we thought very highly of Google, and we said, "Let's talk about it."
Four months later, we were sitting in Google.
Mind you, it wasn't an easy decision. I struggled hard deciding if going to
Google was the right thing to do. We weren't desperate. We actually had a term
sheet on the table for $1 million in investment from Joi Ito's Neoteny (who
ended up investing in Six Apart). And after 4 years of pouring my heart into
Blogger, I saw a lot of risk in giving up control.
Eventually I decided Google was right. I really thought we could do huge
things at this point, and Google had done bigger things than most, so I wanted
to get in there and learn and get those resources.
Livingston: At what point did you most want to quit?
Williams: There were a lot of points in 2001 that I seriously considered quitting.
Everybody I knew just thought I was crazy. And I was getting negative feedback
on the Web; people who used to be my friends were posting negative things
about me. We'd gotten enough press at that point... the Industry Standard,
which was the bible of the dot-com era, had this annual list, the Net 21, titled
something like, "The 21 people who had made lemonade out of lemons," and I
was one of them. It was pretty cool, but the title for me was "The Idealist"
because I hadn't sold out. Like I had a chance to have riches and I didn't.
Someone took that and wrote a parody: "The Egoist." Because there was a
story--not really on the surface, but very clearly underneath the community
that I was previously a part of--that was a very negative story about what happened
in the last days of Pyra, because all those people left and they weren't
very happy (completely understandably).
For the most part, the old Pyra employees were cool with it later. But,
during 2001, these stories got out that I took over this company and kicked
everyone out and was just this terrible guy. That was the worst part.
And I was writing this service that was free and thousands of people used it,
and all I heard were the complaints when it wasn't working. So for many reasons
it was bad. I don't know how close I came to quitting. I don't think I was
terribly close, even though I should have been. I was always hallucinogenically
Evan Williams 123
optimistic. That's the only reason I kept going. Not because I thought I could
take this suckiness for a long time, but that it's going to be better tomorrow. I
had all these big ideas, and I could never stop thinking about the product and
the thing I was going to build next.
That always being around the corner in my mind is basically what allowed
me to go through all the bad stuff. As well as the fact that, at that point, it was
just pride. It was so public. If I would have stopped, that would have been very
public also.
Livingston: Were there any other really stressful moments?
Williams: That's an understatement. I can think of many. For example, when
the site got hacked on Christmas day. I was in Iowa, visiting my mom, and I
didn't find out until the next morning. Someone was able to run an update on
the database that changed thousands of users' passwords to the number 1
(which people started to realize when they couldn't log in and used the forgotten
password feature to get theirs via email).
Having your site hacked is stressful enough, but here I was in Iowa trying to
assess the damage over a dial-up connection and a tiny laptop. And I didn't have
a sys admin or anyone else working for me at the time. I ended up spending
most of the day in a Kinko's doing damage control. So much for enjoying the
holidays.
Livingston: What advice would you give someone?
Williams: I think one of the things that kills great things so often is compromise--letting
people talk you out of what your gut is telling you. Not that I
don't value people's input, but you have to have the strength to ignore it sometimes,
too. If you feel really strongly, there might be something to that, and if
you see something that other people don't see, it could be because it's that powerful
and different. If everyone agrees, it's probably because you're not doing
anything original.
I had the personality that never liked school and rejected the normal way of
doing things. Even when I was in school, I'd try to make up alternative solutions
to math problems. When I was at Google, they had this huge focus on academia.
Grades were super-important. Getting good grades at a good school is one
filter of brains, but it might also suggest you like following rules.
Another thing is that luck comes in many forms--and often looks bad at
first. I always look back on the deals that we didn't do and the things that didn't
work out, and realize what seemed like a bummer at the time was really lucky.
Like the early acquisition opportunities. These obviously would have been
really bad, as opposed to what happened later. Through that whole experience
that's one of the biggest things that I've taken away: if you have some plan and
it doesn't go that way, roll with it. There's no way to know if it's good or bad until
later, if ever.
Livingston: What was the most surprising thing?
124 Founders at Work
Williams: One thing that I used to be bad at was paying attention to how other
people are feeling. So when problems came up with some of my coworkers, it
totally surprised me. That stuff shouldn't surprise you, and it did.
I think I was also surprised by the success of something so simple. That's a
mantra for many people in the technology world--simplicity. But what we built
wasn't that amazing. It was the idea of putting a couple of things together and
being able to establish a lead by doing something really, really simple. How far
you can get on a simple idea is amazing. I have a tendency to add more and
more--the ideas always get too big to implement before they even get off the
ground. Simplicity is powerful.