Googler 13

Monday, February 14, 2011

Porn for Cookies (and other early Google cultural nuggets)

Case studies and consultants have written ad nauseum on Google culture, how to replicate it and transform your corporation! Ummm...no. Corporate culture is an interesting thing. It is a social anomaly reflecting the microcosm specific to the unique individuals participating (especially the top management/founders- since it tends to flow top-down) and the time and place in which it has sprung up. It is not developed or imparted via an employee handbook. It is organic- a living, evolving reality shared by only those who are part of it. We used to talk about “drinking the cool-aid” (please pardon any insensitivities around the phrase). We all talked, acted in accord with and truly bought into the idealism of our shared mission. Google was a rare gem. At a time when other companies were bleeding venture capital cash: hiring at a breakneck pace (sometimes just to fill desks and make it appear that work was being done- this is true- my wife was hired as an intern during the .com boom by a start-up to sit and make personal calls so that it looked like “big things were happening” when the company would walk potential investors through the office); and spending hundreds of thousands of dollars (sometimes millions) on lavish parties to get cache in the internet social and funding communities, Google was frugal. I remember early on when Larry [Page] bought a $3,000 green laser on the company credit card that he wanted to take to the Burning Man Festival to project live search queries onto the ground. We revolted since it seemed a poor use of funds. I think he ended up returning it.

The colorful and fun offices we lived in were also simply a natural outcropping of early Googlers’ interests. Ray had a lava lamp, then Chris joined and brought his from home....then there were lava lamps around the office. We had some large exercise balls around to sit on, one popped and Sergey [Brin] told our receptionist to order some more...she bought alot more! and then these colorful balls were just all around. One was huge (maybe 8’ diameter). For fun or stress relief we would take a running jump and bounce off the thing and usually spring and crash into the wall :-). For other companies, I can’t imagine that just getting a checklist of things Google had in the office and sending someone to Costco would enhance the experience of their employees. There is no back-story for them. It isn’t grounded in company lore or reflective of the employees. Manager: “Here. Sit on this green ball instead of your chair” Employee: “Why?” Manager: “They did this at Google”.

One of my fondest memories for the first few years of Google’s life was my friend Craig (3rd employee, 1st non-founder) making homemade bread in the kitchen and then walking through the office carrying it and some butter loudly declaring “bread. bread. bread.” and people would come out of the woodwork to grab a slice and say “hi”. You can’t script this stuff. Its endearing because its real.

I used to pay very close attention to our traffic (I used to forecast it for the company) and would often send out mass emails when a milestone was hit. One day I sent the following message to the whole company: “Congratulations, today is your day. Google has hit our first 100 queries per second day!” Can you imagine where else you could get away with broadcasting this to the entire company? If you disagreed with something the company did...that could be broadcast and a start a discussion as well. Employees with all our quirks had a voice and the freedom to use it. It couldn’t have been any other way.

We were a true Silicon Valley tech start-up. Amusing and adsurd things happened there on a daily basis.

[this a screenshot of an actual instant message conversation between Larry and Sergey’s original assistant and me, circa 2000]

The comradary and openness of the company to employees was another special thing. For many years, any employee could access all information about the business and performance of the company. You could go to our intranet and see our traffic and revenue every day from each source. Every Friday we would get together as a company and divulge every detail of what was going on with no fear that it would leak out. Trust bound us. We were always free to ask anything. It was always amusing for Omid (my boss) to stand on the table in our cafe and talk about our business progress. Until Eric Schmidt was brought on board, every week someone would cheekily ask Larry “when are we getting a [real] CEO”. It was all in good fun...and when Eric finally arrived, he joined an already profitable company.

We were treated very well internally. Amazing chef and sous-chef’s Charlie and Jim had free reign to buy organic & local foods to nourish us and keep us healthy. We were fed (everything free) things like Copper-river Salmon when it was in season. We had masseuses on premises. You could go onto their calendar and schedule to go for free whenever you wanted during the work day (when this changed to a $20/hr co-pay there was an uprising, but the charge stuck). I used to go Tuesdays after playing roller-hockey in the parking lot. As for roller-hockey, one of our early employees was a hockey fanatic, so she got professional jerseys made up for us.

(Steve Schimmel & Sergey Brin playing roller-hockey in the Google parking lot, 2000)

Our culture was aided by initiatives of the individual employees. I started the Google Wine Club in 2000 and an always growing group of us would meet every few weeks at night at the office and talk and drink wine. I would always pick a theme and prepare a lecture and people would bring in bottles relating to it. Employees from every department would come. One rule was that they weren’t allowed to talk about work. This led to fascinating topics and side conversations about amazing things people had done, their hobbies, interests and dreams for the future. These events paid huge (if largely invisible) dividends for the company. Sales folks and engineers, legal and HR folks who otherwise would not have known each other when passing each other in the halls became friends, which led to exchanging questions and knowledge during the work day. They were connected in a way that would not have happened otherwise. They now knew each other socially...which is a very different bond. All of these things made for a cohesive existence and united us as one culture.

And that brings us finally to “Porn for Cookies”. Working for a start-up, especially one that is fanatical about excellence can lead to all sorts of odd requests. There was a time when we had a computer screen on our receptionist's desk for our office visitors to view live search queries being performed by users in real-time. However, as the universe of our users “normalized”- i.e. the proportion of average people searching grew versus academics doing “research”, more and more queries became “adult” in nature. I can say that before we finally took it off the front desk you could stand there for a few minutes and get a extremely entertaining show! This highlighted a problem...more and more porn was being added to the web every day. We did not want to show pornographic related search results when they were not intended. As a “business development renaissance man” I had myriad one-off and odd tasks. One of the more interesting ones was to get a random sample of search queries from the logs group and do a statistically significant categorization of what people were searching for. I pulled about 400 queries and started emulating the search to see what came up, follow those links and fit them into a category. Umm...wow! After ten minutes on the task, I had to close the blinds to my office and lock the door. Yes...in fact, people were searching for porn. The problem then became how to make sure pornographic results didn’t make it into innocuous search results. After looking at the external solutions for porn filtering, Matt, one of our early engineers decided that they weren’t good enough, so he was going to build his own solution. The job of aggregating a list of pornographic words was too daunting for one person, so his wife had an idea. She baked cookies to incentivize Googlers to help her husband look for porn. His request for assistance went out with a company-wide email: “Porn for Cookies”. His wife proved to be a great baker and Matt got the help he needed :-)

Google culture. It is uniquely Google's..authentically fashioned from the first employees. Take the fundamental lessons but don’t try to imitate it. Let it evolve and flourish. As it says in Google’s “10 things we know to be true”. “You can be serious without a suit”...at least that was true for us. Come up with your own “10 things” list and make some cool-aid.

Saturday, February 5, 2011

Your Start-up Sucks (post includes a Google business plan slide & pre-adwords ad result)


Ok. Your start-up probably doesn’t really suck, but someone who doesn’t know any better might look at your “plan” or “product” and think so.

Having a vision, building a working prototype, getting others to believe and invest in you...these are the first steps to starting and growing your own technology company. You might think you have to have it all figured out from the beginning, but you don’t. Intel famously had a one page business plan. Google had the “beta” moniker on it for what seemed like an eternity. Get it out there and begin the perpetual improvement cycle that will define your future.

Google is one of the the most successful, innovative and valuable technology companies ever created...but it had humble beginnings, just like everybody else.

When I was a junior in college, I did a semester abroad in Florence Italy to study the Renaissance. I greatly admire the disruptive and game changing philosophies that came from that unique time and place. An interesting tidbit about the people living at that time is that they knew they were living in a renaissance (a period of rebirth and illumination). As such they began to think of themselves and what they did as being important and they started to keep personal journals and collecting memorabilia. It was pretty clear to me at the time that 1999 Silicon Valley and Google specifically was a special time and place. I’ve built a fine archive of early Google memorabilia- my own Smithsonian Google exhibit of sorts. I’m going to share a tiny sample of it here.

Google was started by two Stanford Computer Science Ph.D drop outs. Most of the early employees were just out of college and had never had another job. By the time Omid Kordestani and his original business team (including me) joined from Netscape in early ’99, the company had already written a business plan and was shopping it around. Google got its $25 million in funding with the following as its Business Model slide:

(this is a scan of the actual 1999 business plan page + my watermark)

As you can see...we did NOT have it figured out.

Its pretty obvious now that our Advertising program was to become the elusive and mysterious “Revenue Stream #3”, but even that had a bumpy start. As I stated in a previous post, Larry Page’s vision was that “advertisements can be as relevant as search results”. In the “reinvent the wheel” culture that we worked, this meant looking for any and all potential solutions to the problem of creating RELEVANT advertising. What is unknown to most of the world, is that before Adwords, Google created and went live with its first attempt at advertising in October 1999. That program lasted for less than one hour. Google Engineers had devised a theory that we could do a good job matching search queries to relevant books on amazon.com using a certain grouping technology. We signed up for Amazon’s affiliate program and started showing books with links to purchase. These were our ads. I happen to have printed and keep a few examples from that short lived experiment. Many of the results ranged from mediocre or funny to outright scandalous. Below is the most innocuous of my examples. I wish I could show you some of the racier ones, but my wife made me promise her that I wouldn’t :-)


(click to enlarge- the search query is: "ugly girls". the ad is: "Just for you : A Picture Book of Helen Keller")

Google is a great company. It got that way, not because it was perfect, but because it was filled with smart people who were willing to learn and adapt...and that’s the way to end up not sucking.

Tuesday, February 1, 2011

Steve’s basic deal methodology

“Triangulating in on Value”


As a self-taught negotiator, I had to come up with my own logic for how to value a deal. One thing that I learned very early (especially on the buy-side- when I was the one with the checkbook), was that the price a vendor first quoted is just a starting place. Often we would end up nowhere near that price. Most of the deals I did at Google (at least on a volume basis) were to license other people’s technology or data for use in our products or services. This included IP geocoding, language translation software, white/yellow page data, etc. The challenging thing from a pricing standpoint was that the way we would be using these things was custom to us. Cookie-cutter rate card pricing was not going to be applicable. The value was ambiguous and we considered it a trade secret exactly what we were doing with their product. Many vendors wanted to participate in the revenue we would receive from whatever service their product was integrated into. I always rejected this concept. As I explained it, its like a supplier of silverware trying to charge more to a restaurant that has $30 entrees versus one that has $10 entrees. My position was that I might pay more for the vendor’s premium offering, but not more for the same offering just because I could monetize well. In addition, I was relatively limited in the intangibles that I was able to offer. We wanted to keep it secret whose technology we were integrating, so we could not allow the vendor to announce that Google was using their products. This made things difficult for me, as vendors were very willing to negotiate down their prices for this ability. The most I was ever able to offer was that they could include us in a list of customers (without disclosing any details of our relationship) in their sales presentation materials (but never on their website).


One thing I learned was that people aren’t always good at understanding the value of their own products and therefore don’t always have good logic for their pricing. As a fundamental analyst, I was always ready to take on the debate. Ultimately, I always came up with the price I was willing to pay, which didn’t always have much to do with the quoted upfront price. Nobody works for free and I didn’t want to drive my vendor’s out of business...but I did want to pay what I believed was fair and equitable.


Here’s how my thinking worked: I always tried to do what I coined “triangulating in on the value”. When I considered how much to pay for something I attempted to discern:

  1. How much money did I expect us to make from our product/service that was integrating the vendor’s product. I needed to ensure that I what I was paying for outside technology was low enough to provide us with reasonable profit. If you pay too much, you can’t make money.
  2. How much profit is the vendor making (much of this was educated guesstimating). I wanted the vendor to make a profit, but not an exorbitant one. I did not want to squeeze the vendor so hard that they ultimately felt cheated and therefore provided poor ongoing customer support for us or difficult re-upping of the contract at the end of the term. "We've never done a deal at that price" is good. "We'll go bankrupt if we give you that price" is bad.
  3. What was the cost of substitute goods. I can’t say that I remember ever dealing with a vendor where they were the only one with the type of product we needed. That being said, there were specifics that made one vendor a better fit for us than the other. If the preferred vendor was substantially higher than another, I made them answer to why, as they were quite familiar with their own competitive landscape. There was never a very good reason, and invariably this question would make them immediately more flexible.
My goal was to land at a price within a range of where those things intersected.

In summary:

  1. Make sure what you pay for outside resources allows you to make a reasonable profit on your finished goods.
  2. Allow for your vendor to make a reasonable profit- but no more.
  3. Know the prices for substitute goods so that you have a sense for the market and negotiating ammunition.