The Good Guy Behind "Don't Be Evil" and Google Mail

"Don't Be Evil" - Paul Buchheit outside a church, a stone's throw from where he grew up in Webster, NY. Photo by Rajesh Barnabas - December 2007

This was a big week for Paul Buchheit. His new company, FriendFeed, no longer in private beta, launched Tuesday. In addition, they announced $5 million in venture funding. The Wall Street Journal, Techcrunch, Techconfidential, other major news outlets and top blogs picked up the story. My interest, though, began three years ago.

Over a decade had passed since I had seen or heard about Paul. He did not attend the 10-year reunion at Webster High School, but had one of his old buddies take pictures of everyone. I found out then, that at least one of our 400 fellow classmates of 1994 had made a small fortune on the Internet.

Not many from Rochester have an entry in Wikipedia, but Paul has. It notes his contributions at Google, for inventing Gmail and AdSense, and for suggesting Google's now-famous company motto, "Don't be Evil."

Now this 2007 holiday season, Paul came home to visit family, and I found him for this interview. Location: Spot Coffee, Rochester NY.

Thanks a lot for coming out.

Not a problem, just on vacation so nothing much going on. This vacation turned out to be really good because of the Museum. (During this interview, Paul's wife April and 2-year old daughter Camilla were visiting Strong: Museum of Play.)

So you've been living in California for quite a while, but I don't detect a Cali accent?

Yeah, not so much. I think everyone in urban areas talks roughly the same. We all talk TV English.

You used to get 500 emails a day at Google?

Google is a very email intense place. So, on top of all the mailing lists, there is a fair amount of automated email. Some system goes down and it reports statistics of what is going on. So, there is a pretty huge flood of emails that goes on.

So is your life less hectic since exiting Google and starting your own business?

It's different because I have to worry about a broader set of issues. I was just emailing our landlord because our lease ends in August and he wants to know if we are renewing, and I have to think about our next location.

What does your new company structure look like? Is it the traditional triangular shape?

We're actually trying to build a less hierarchical kind of organization. We've all had experiences at big companies and also at start-ups and they both have their advantages and disadvantages. It is yet to be proven, but I believe it is actually possible to capture the best of both worlds: try to get the resources of a big company, but [also] the agility and ownership of a smaller company. And so, we're really trying to make a place where really creative productive people can actually get a lot done and have ownership of what they do.

I read that to recruit workers, Google put some type of logarithm or equation on a billboard and whoever solved it got a job interview?

I don't think they did anything that direct, but they did a lot of promotions like that to hook people interested in problem solving. So, I think the thing you're thinking of is they actually put a billboard on one of the big highways out there—101—that had some equation on it. So, it looks kind of strange, because you don't often see just this black and white billboard with some equation there. But somehow, when you had solved the equation, it yielded an URL, and when you went to that URL, it was a jobs page. So, it was kind of a hook for someone who has that internal need to solve this puzzle, and then they are going to go to this URL and think, "Hey, this might be a good place to work."

Hiring was definitely one of the things that Google did a good job of, and I think it is one of the reasons they succeeded. They were very aggressive but also very careful to hire really the best people. And I mean, obviously any group that large you will have some people that, it turns out, aren't that great. But given the numbers, it is actually amazing, there are just some really phenomenal people there.

What do you do during your spare time? Do you allow yourself to get away from the computer? You mentioned that you don't usually bring your laptop to a coffee shop?

I don't go to coffee shops that often, but if I did, I might bring my laptop. For leisure, I don't know if there is any consistent theme. I usually spend time with my daughter and wife, or my brother.

What is your wife's occupation?

She used to be a teacher. She now takes care of my daughter. My daughter was actually born 100 days early, while we were on vacation. That is right around the limit of how early they can be born and survive. She was in the hospital for 3 months. So, at that time we both stopped working. That happened when I was still working for Google. I went back after 7 months, worked at Google for another 6 months and then decided I was finished there. I took off about a year and a half before starting FriendFeed this fall (2007).

Did you always have in mind that you were going to start something like FriendFeed or…when did you conceptualize the idea?

I've been interested in start-ups for a long time, but you just kind of wait for the right time…

Why is the right time now?

In college I was interested in start-ups but I knew really nothing about how I would go about doing that. All I knew was that a lot of them were located in California. I just took a job with Intel in Santa Clara, which is roughly the right location. So, it was that one step forward kind of thing.

To get yourself in the vicinity…

Yeah, so, I was just taking baby steps and was at Intel for about a year, but that was not my kind of place. It was a very big company, just not for me, so I left. And that was when I joined Google, which at the time was just a little start-up of about 20 people. We were located in a little office in Palo Alto. It wasn't necessarily something that I expected to succeed. I just thought I would be getting my feet wet with start-ups and learn about starting companies. As it turned out, obviously that job lasted longer than I expected. I just thought maybe they would go out of business after a few years and I would move on. But seeing as how they did so well, I ended up staying for about 7 years. In the interim period after I left Google and before FreindFeed, I did quite a bit of investing and advising of startups. So again, it kind of broadens perspective, and I learned about how different people are doing things and what all the issues and implications are.

In Jessica Livingston's book Founders at Work — Stories of Startups Early Days she attributes your success to essentially two things: One, they work on what they find interesting, not just on what they are told to or supposed to. Two, they pursue ideas not generally expected to work. Can you give me an example of this? Is it something where you stay up late at night racking your brain for a way past this next programming knot or…

No, no, a lot of it comes down to just a funny idea. Probably the nicest example of both of her points was the content targeted ad system at Google called AdSense.

You see these little ads on the side of the page at Google about cars. The very first version of that was actually a side project while I was working on Gmail because we had no way of making money for Gmail. So, people kept asking, "Well, how are you going to make money?" And it was kind of an irritating question because I just wanted to build an email product. So, there was this idea floating around that maybe we could target ads to the content of the messages so we could show people relevant ads. But it was generally considered to be a bad idea, very difficult, not technically tractable, so no one pursued it. But on the side, I figured out a way to build a content targeting system really quickly. I just built it one night and released it on our internal prototype of Gmail.

At first people were really upset. (Laughing) They were like, "What are you doing? You were supposed to be building an email thing and here you are building an ad system." We agreed that this was a bad idea. But beyond that, a number of people and, most importantly, Larry and Sergey [founders of Google] noticed that the ads were surprisingly good and relevant.

It is better than getting some stupid ads…

Right, pretty quickly people realized, yes, we really can target ads, because look, if this is something I had written in a couple hours and it was pretty good, then maybe this really is something good. So very quickly a project was formed to turn that into a real product. They released it within 6 months, and that was the end of my involvement with it.

So in the end, that was an example of not something we really planned to do, and not something I was supposed to be working on. But at this point it makes billions of dollars of year, those ads, so it turned out to be a good thing.

But wasn't there some privacy issue controversies around people's email being scanned for content?

I think it was something where the concept was really unfamiliar. The first reaction is, it does sound a bit weird. But if you actually think about it, there really are no privacy violation issues, because it is exactly the same as a spam filter technology. But we discovered, that when people hear about it, it sounds bad. But when you actually experience it, you just see it and it makes sense. You think, "Oh, this isn't so bad. It's just that whenever I am looking at a page, it shows me some related ads."

Not a lot of people would leave Google, especially when you did. The company is doing great. Was there any issue of not getting credit?

No, no.

Is that part of the business, I mean, when you work as a team and then some people get credit and some don't?

That's always a problem and that's another thing we really would like to address with our new company: to try to give credit to people who are actually doing the work, not someone who stood up and took credit for it. But no, that really wasn't a big factor in my leaving Google.

I just really enjoy creating new things and releasing new products, and the reality of a big company is that over time there are just more and more roadblocks between you and just releasing something. On FriendFeed we might come up with a funny idea and just do it and release it and that can all happen within a matter of minutes, literally. We say, "Hey, this would be cool. I'll write the code and release it and it's live."

With big companies, you're like, "Hey, this would be a great idea," and then there are approvals you have to get, and then this big long pipeline to launch, and it would take months and months. And then you think it's not even worth it. You don't even bother. That extra cost of getting anything done just causes you to not even bother sometimes, especially with things that are obviously not that important but might be kind of fun ideas or interesting. It is really an issue of opportunity costs, because you don't really know about all the great things that didn't happen, you only know about the things that did happen. So, when you create processes that slow things down, you never really realize how high the costs of those processes are.

With that AdSense idea I mentioned earlier, that probably would have never happened at another company. Why did we create product at Google? Why didn't it happen at Yahoo or Microsoft? Probably because we had lighter weight processes. It was just easier for people to do things.

It still took six months…

But that was them just taking a rough prototype I had built and turning it into a product. That was very good. Six months is a very reasonable time frame. I am sure I wasn't the only person thinking about this. There were probably people at all the other Internet companies who thought of these ideas. There was probably an engineer at Yahoo, but they had no way of actually trying it out, so they never created the product. So it is really important to create opportunities for people to test and prove ideas.

What is the risk of moving too fast? You are in the beta phase for FriendFeed? Isn't there a risk in pushing your product out there too soon?

For FriendFeed we haven't pushed really hard to get users, because we don't think the product is quite ready. We have what we think is just a good amount of users to test out a lot of concepts to understand where the product is working and where it isn't. So, it does require a certain amount of judgment about just how many people you are going to expose your new ideas to. And also being careful you don't do anything that causes harm. With Gmail we were very careful storing people's data. We designed the system to recover from different kinds of failures.

This is one of those late night questions that people throw around. The storage centers for Google are located in many places and internationally now, right? I am asking this, just in case another US city gets bombed.

Yes, the data systems are all around the world. One of the unusual things about the Gmail system is that you could literally destroy any one city in the world and there wouldn't be any data loss because everything is redundant. So, you could survive just that sort of issue.

Take me back to 2001 to the company values meeting. I read that you and a colleague of yours—Amit Patel—came up with this phrase…

"Don't Be Evil."

And the leader of the meeting kept on dropping it off the list. Can you remember some of the other phrases that were considered above yours?

They were all these forgettable standard boilerplate company values like, "We strive to honor all our commitments," something that sounds really nice but that doesn't really have much punch. There were probably about ten of them that were these nice platitudes but nothing that stood out.

What is the background to coming up with that theme? You mention that your competitors were…

Yeah, part of the inspiration was what was on our mind at the time. Well, there were a couple of things on our mind. Most of the search engines were actually selling search results at the time. At Google we had the advertising but were always very careful to mark the advertising as being advertising, just like if there is an ad in the newspaper, it always says, paid or sponsored by. You don't mix the part that is editorial and the part that is paid. So, we had the same concept that if something is paid for, you need to tell the users that this is a paid advertisement. The other search engines didn't do that. They said, "Hey, this would be a great way to make money if we just sold off the search results." And so, you could actually pay money to get included in the search results for the other search engines. We thought they were kind of exploiting the users, exploiting their trust because they think, those are just what comes back naturally, when in fact, it's whoever paid to get there. We just had a vision that this was not the right way to go.

And then of course, back then there was a lot of talk about Microsoft, and some of their decisions were clearly not in the best interest of the users of the world. Obviously you want to do what's best for yourself but at the same time it should be good for the world. There are a lot of businesses that only exist for their own benefit. It doesn't matter that they are effectively stealing from other people. The spammers, for example, they're very parasitic. And so, we always wanted everything we did to not only be beneficial for us but also for our users.

Did you play a role in the huge increase in storage that Gmail became known for?

Yeah, one of obvious things right from the start with really all the other email systems is that at the time when we launched, Yahoo had 4 megabytes, Hotmail had 2 megabytes. So, we released with 1000 megabytes. We didn't want to be just a little bit better. We wanted to be orders of magnitude better. What I was aiming for, from my perspective, was that I should be able to just keep everything, because it's a waste of time to try to decide what I want to save or throw away. And then, I might throw out something I actually needed. So, I wanted to make it easier for people to just keep everything because that's the way I like to work.

What is interesting is it wasn't just the free Web mail that had this issue. The same thing went on with a lot of corporations where people had a quota. When I worked at Intel, I had a 30 megabyte quota and it was always inconvenient to go back and delete emails.

As it is, you don't have a paper trail anymore…

Yeah, to some extent they were trying to prevent a trail for legal reasons, but also it was because the systems they were running were not very good. So, there have been some second order effects of Gmail, because a lot of these organizations now try to give someone a 50 megabyte quota. And you look like a fool if you try to do that because you say, "Hey, wait a minute, Gmail is giving me 6 gigabytes, how can you tell me…" It's just not acceptable anymore. And so throughout, a variety of organizations are forced to give more to the users, which is an interesting side effect.

A media analyst said, "There is no company on the face of the planet that scares as many businesses as Google." Do you think you guys came up at a time when there was this backlash in America against corporations getting a little too greedy and caring more about their shareholders than their customers? Whereas, Google seems to be a newer company in setting the precedent for caring for its customers.

I think, with the Internet there is more of a free-flow of information and awareness of what's going on, so there is a competitive benefit to being good to users. In the past you could take advantage of people and no one would ever find out. A lot of the value we created at Google was creating that positive brand that people trust. A lot of the things people do to make money in the short term ends up costing them in the long term because you lose that trust with people. And that's exactly what has happened to Google's competitors. People just don't trust them as much.

What are the parameters to your new project FriendFeed? What is your vision? Who is your target audience?

It's a tricky thing to figure out, when you're creating a new product, to know who it's exactly for. We are starting off with a small group, but the vision is to make this something that is valuable to everyone. But inevitably the people these products appeal to are early adopters who are more into technology and trying new things. But that's okay, because they will tell their friends and families and bring the second wave of users.

Friendfeed seems a neat service, but do you think that there is a percentage of Web users who would find it too much work to mentally filter things before they put them in a friend feed? For example, do I want my mom to know that I favorited the YouTube video by 2 Live Crew?

To be clear, all of things that are shared only happens if you explicitly share it. If you just watch that video, no one is going to find out. It's only if you click that button that says, "I want to share this with my friends."

How do you intend to hook someone who is not into these social networks like MySpace or Facebook? How does your product distinguish itself from those?

We're not a social network. We are not trying to compete with Myspace or Facebook. We're trying to appeal more to regular people. Honestly, none of us are big users of the social networks, because it just doesn't feel that useful. But I think there is a common desire among people of all ages just to share different things. My aunt sends me articles she finds. My mother cuts things out of the newspaper. So, we're much more about sharing interesting content and the socialization that goes on around that. What works is when people are just chatting about a funny story or whatever is in the news. You see this happen with all age groups. So, a lot of it just comes down to seeding topics for people to chat about.

How do you plan on competing against all these other sites that can also aggregate and then share personal feeds? For example, 30 Boxes, Jaiku [now being acquired by Google], Tumblr…

We have a different objective. We are putting these all on one page. Everything is in one place. Then if you want to pursue items, they are one click away. It will be less cumbersome than sending everything through email. Email gets heavy if you're trying to send messages or pictures to everyone.

When do you plan on rolling out FriendFeed to the general public?

Sometime early in 2008.

If you want to make inroads on the MySpace crowd, you should find a big-name rapper to promote his site.

Funny, you should mention that. I have been seeing MC Hammer at all the tech conventions. He's been promoting his site DanceJam that is supposed to be this YouTube-esque hub for dance videos.

Is it liberating to be working for yourself on your own projects?

Yes, but on the flip side there are a lot more responsibilities.

I am assuming you are pretty well off at this point. Does it become difficult to motivate yourself to go after the next million?

Well, actually it provides me room to be a little more long-term. If you are not starving, it affords you to take bigger risks.

What's the working environment like? I read that when you were at Google you used to arrive in the afternoon and work late into the night? Are you changing things for FriendFeed?

We have a fairly casual environment. We believe that people are more creative in enjoyable settings and when they are doing work that they are genuinely interested in doing.

How can Rochester nurture a creative culture and entrepreneurial environment like Silicon Valley?

That is a difficult question. If there was an easy answer, then it would've happened by now. People are very accepting of risk taking and failure in the Bay Area. That's what sets it apart. I joined Google thinking it would probably fail. As it turned out, I was wrong. Very few startups get big. Everywhere else, it seems like a shameful thing to fail at business, but in Silicon Valley it is seen as more of a trophy. Having tried and failed shows experience. I don't know if there is any formula.

Only portions of this interview, originally intended for publication in Rochester Magazine, made the 300 word cut. Read it in the March/April 2008 issue on newsstands today. Read the rest here. For Newsvine readers, their loss is our bounty.

Rajesh Barnabas and Paul Buchheit were classmates at Webster High School, class of 1994. John Lam contributed questions and edited for copy. He is a software developer and consultant in social computing and also News Editor for the Black Rock Beacon.

Back To Top