2009 Founders' Letter
Sergey and I got over our fear of failure and finally founded Google in 1998. If we had known then what Google would become in 2009, we would have been totally flabbergasted. The scale and scope of our services, and the opportunities they offer users, are phenomenal, and we are very lucky to be a part of this business. Rather than try to run through an exhaustive list of everything we have done this year, I’m going to focus on a couple of issues—access to information and a new model of computing— that are of particular interest to me, and on which I have unique perspective.
I was lucky enough to grow up with computers, and so from an early age I learned that there’s always more potential at hand with technology, especially as I struggled to read programs off kludgey cassette tapes. While I’m astounded at the pace and progress we have made on many areas of the Internet and computing, I am also amazed how slow progress has been in other equally promising areas. Often what is required to make progress in technology is focus. For example, there is a hundred times more activity in clean energy today than there was just a few years ago simply because more people are now focused on this issue. What really motivates me is this dichotomy of slow progress in some areas and fast progress in others. This is a tremendous social and business opportunity. Who would have thought in 1998 that anyone could get for free a highresolution picture of their house from above, and even from the street? That is Google Earth, Maps, and Street View. Was it a foregone conclusion that we would have these kind of products now? No, it was not. This progress happened because focused teams of people made those ideas a reality. We could just as easily have hit 2010 and not have had these services available on the Internet at all.
Finding important technological areas where progress is currently slow, but could be made fast, is what Google is all about.
I’m excited about our opportunities to make a big difference in people’s lives through technology. We can build these great new products into great new businesses too. Google Translate is a recent example. You can now translate pretty well and instantly between any of 54 languages—that is about 2550 language pairs—and search the web and read results in languages you don’t speak! We even have Google Translate for your Android phone—so you speak in English and it translates into German out loud! This is all using software for speech recognition and translation that we have developed at Google. Users around the world have noticed the speed and quality of our translations, which is why Google Translate is growing like wildfire. We’re putting this technology into YouTube too, so you can watch videos even if you don’t speak the same language, or have difficulty hearing—automatically. Imagine anyone in the world being able to watch and understand any video no matter the language. YouTube is an extraordinary platform, and for me is like another kind of tube that I use every day, toothpaste. Apparently I am not alone: we have over 1,000,000,000 daily views on YouTube. YouTube has new features like full high-definition content at 1080p and tools to help you share videos with your friends. Over the last year, YouTube has also been making a lot more money for us and our partners, with content partner ad revenue more than tripling in 2009.
Recently we announced a new project to build 1 gigabit per second fiber-to-the-home broadband networks for one or more U.S. cities and towns between 50,000 and 500,000 people. This access is about 100x faster than most people have today. We asked communities to come back with ideas, and one mayor had an unusual response:
“I, William W. Bunten, mayor of the city of Topeka, Kansas, urge the citizens of Topeka to recognize and support the continuing efforts to bring Google’s ‘Fiber for Communities’ experiment to our city, and do hereby proclaim that for the month of March 2010, the city of Topeka will be known as Google, Kansas.”
From such quirky tributes and detailed applications, we have seen a lot of interest in Google Fiber. Our goal with this project is to show what’s possible by driving technological development of home Internet connections at a faster rate. If we succeed, it will benefit users everywhere, as well as our own services, which can debut amazing new capabilities using higher speed connections.
Access to Information
Search and Ads
Roughly 70% of our resources are allocated to core search and advertising, and we have been doing a tremendous amount of work on both. Creating the perfect search engine remains our ultimate goal, but we’re still a long way from doing that, which is why we are not resting on our laurels. I have really enjoyed our new “show options” link that appears at the top of the results. If you click this you get a whole bunch of options, including time, geography, prices, images from the pages, more or less shopping, and even thumbnails of the pages. This has really improved my searches when I’m looking for something a little harder to find. We have made our snippets “richer” in all sorts of ways. We also improved personalized search, helping you get results more tailored to you, and have done a lot of work on getting real-time results to you in seconds. A lot of our focus goes towards improving core relevance—making sure you get exactly what you want when you type a query. Typically we are running hundreds of experiments at a time to improve relevance. And we made additional improvements around comprehensiveness, making sure we search everything in the world.
Search ads are our main source of revenue and of course an important focus. We view our search ads as information for users, just like search results. With Universal Search, we now provide results in many different formats, such as videos, maps and news, and we needed to do the same for advertisements. For example, you can now see product ads with prices and pictures of the items, similar to the shopping search results you can get in organic search. It is interesting to note these retail ads can be cost-per-acquisition, which means the advertiser pays only if someone buys something. This is wonderful for the advertiser, who doesn’t have to take any risk at all. Advertisers can easily put in all their inventory without worry, rather than just a subset of the most important items. We also get to build great new prediction systems that do the hard work of estimating what bid yields the best results for advertisers, based on the cost-per-acquisition goal they set. There are new ad formats specifically for local businesses, comparison ads for financial products, and sitelinks for navigational queries. I’m really excited about the benefits new ad formats can have both for our advertisers and our users. We also have done some significant work to reduce what we call “scammy” advertising to make ads safer and more relevant for users. In addition, we made many improvements to our core advertising systems behind the scenes. There is a lot of technology used to make the advertising work and estimate clickthroughs of ads and so on. Improvements to these systems have very measurable and meaningful effects on advertiser and user happiness.
On display ads, we have really benefited from a successful integration with DoubleClick. We launched new analytics and media planning in DART for Advertisers (DFA), and have made big strides in the Google Content Network—the extensive collection of partner sites on which we run ads from our network. In 2009, we sold display advertising on that network, which includes YouTube, to 94 of the Ad Age top 100 advertisers. I’m also very excited about interest-based advertising, which helps deliver ads tailored to people’s interests. Users can adjust their preferences to generate more relevant ads, or opt out altogether (which very few people choose to do). A tool called Display Ad Builder helps you build display ads in seconds so that even the smallest advertisers can use display. Through our acquisition of Teracent you can automatically create thousands of potential permutations of display ads and automatically optimize each ad that is displayed. The DoubleClick Ad Exchange helps make the display industry more open, transparent, fair, and effective for everyone from ad networks to agency holding companies to large publishers. Over 50 U.S. ad networks have already signed up for the new Ad Exchange. There are a ton of improvements we are focused on making in all of these areas, and I am excited about our very substantial progress to date.
Sergey and I like to use as many of our products as possible, and we have both signed up for AdWords so we can get closer to the real experience customers face every day. Whenever we spend money on advertising, we like to know if we are actually getting our money’s worth. Turns out other people want to know as well! Google Analytics lets you measure in great detail the return on your investment, and everything else going on in your website too. You can directly and automatically use this information to improve your advertising. Getting many more advertising customers to take advantage of this system is a priority. The data Analytics provides, and the analysis it makes possible, is quite a contrast to traditional advertising where it can be very hard to know exactly how well any particular ad worked. This is because the Internet enables much more measurement, and we are trying to accelerate that trend.
Unfortunately no one I know has figured out how to be in multiple places at the same time, so location is important to everyone. As I mentioned earlier, I’m amazed at the geographic products our teams have built. You can get a pretty accurate 3D view of nearly anywhere in the world. Amazing. In the last year we have released our own comprehensive source maps of streets and addresses for Mexico and the U.S.—and users have been working on building and correcting over 60 countries. Street View has exploded around the world with more than twice the countries covered and has unbelievable, higher resolution images in many places. Our Street View images of Whistler at the Olympics had nearly as many views as there are Canadians! We also made many improvements to how we handle local businesses.
I love that you can now search for something in Google Maps and then see all the little dots on the map, no matter how many there are, or how much you move the map around. With all the progress we’ve made with geo products, I can now be found in my one-time location of Happy, Texas!
On a much more serious and sad note, after the tragic earthquakes in Haiti and Chile we were able to gather updated high resolution imagery very quickly to help the relief efforts in both countries. In Google Earth, you can view images of places over time by enabling “historical imagery”. I did this for Haiti and found it brought home the devastation of the earthquake because I would see exactly which buildings had been damaged. It was almost as if I was there.
I was amazed to see on Google Books a fully accessible archive of some priceless magazines, including Popular Science—going back 137 years! It has all the ads and everything, though they didn’t seem to have many ads back in the April 1872 edition. It is truly a dream fulfilled for me that we now have 12 million books scanned and available for searching at books.google.com. That is already bigger than almost any university library, and we’re not done yet. We negotiated a settlement agreement with publishers and authors to sell the full text of many of these books, so they can earn money from their work, much of which is out of print. It’s currently awaiting court approval, in the wake of much controversy and much support.
At the basic level, there is tremendous knowledge available in books and libraries that hasn’t made it onto the Internet. We now have relationships with over 30,000 publishers—an enormous number of partners. Together, we’re working toward a system where everyone has increased access to these valuable texts. I am very excited about the possibilities to help expand human knowledge, create new revenue streams for content creators, and improve the quality of search for every Google user.
A New Kind of Computing
Google Chrome, Google Chrome OS, and Android are all very exciting to me. What we are aiming to do is to redefine the nature of commercial computing by making it modern, simple, and open source. Sergey and I (and Google) grew up with Linux and we have all benefited greatly from that open model. We believe that it is a great way to run a healthy and vibrant high tech ecosystem. In fact it is how the Internet came to be.
All of these products are open source because we believe that is the best way to improve the ecosystem. An open model not only inspires innovation among developers, but also helps generally improve the quality of the software through peer review and public scrutiny of the code. And both are good for users. Google has released over 12 million lines of code across over 350 open source projects, and we host over 220,000 open source projects on the Google Code site. We have had tremendous response from the developer community with more and more developers participating in our ecosystem— an important business goal for us./
I think Google Chrome is a beautiful, fast, and simple browser. I just read a review where it handily beat all others in speed and won the overall award. It is an amazing product, and usage is growing quickly, with over 40 million active users despite the fact that the product is just eighteen months old. We have worked hard to improve the security model so you can browse with less worry of your computer being compromised. We have all sorts of technological magic to make the web into a much more robust platform, so you can run powerful software as easy as viewing a web page. Chrome is so small and fast to install you can get it on your computer faster than you can make your morning coffee. Make your life better and install it now at google.com/chrome.
I love Chrome!
Google Chrome OS
One day several years ago in one of our meetings everyone had a laptop out and was working (this is unfortunately typical behavior, and I feel partially responsible because I demanded power for laptops in all our conference tables). By doing a survey of the room I noticed that only a few people were running anything besides a web browser on their laptop. This seemed rather surprising as you have this big complex OS but it was only running one program, the browser. We decided it would be a good idea to rethink what you are running on your computer from the inside out. If we spend our lives in the browser, and the cloud, why not have the whole computer organized around that? It turns out if you think this way, you can really change a lot about computers. They get simpler, easier, and faster. Google Chrome OS boots from a cold machine in seconds you can count on one hand. This is great and is about the same time it takes most laptops to wake from a suspend (a much more complicated battery-consuming and error-prone process). I should note that Chrome OS is not out yet, and in mentioning it we have violated our own policy of not talking about things before we launch. We knew we wanted to develop Chrome OS in concert with the open source community and of course that had to be in the open. Therefore we had to pre-announce Chrome OS. One reason we don’t like to pre-announce is that we don’t like to pretend we know how long things take to become great products. So we don’t really know exactly when you’ll get a super-shiny polished Chrome OS netbook in your hands. I’m still planning on being young when it happens.
It is amazing to me that everyone doesn’t yet have a smartphone running Android. Doesn’t everyone want an open, Internet-enabled computer in their pocket that is as good as a laptop from a couple of years ago? The reality is that the costs are still a bit high for everyone to switch today, especially with carrier costs and contracts, but that is changing really quickly. My Google Nexus One phone has no trouble playing music through Bluetooth over my car stereo, interrupting to read street names and display a map from Google Maps. I should note that driving directions that prompt you, just like a real navigation system, are free on the new Android phones. Get your car dock ready and you will have an amazing experience with updated traffic and even a photo from Street View of your destination. I can’t even count all the partners we have in our Open Handset Alliance (sounds like Star Wars, doesn’t it?)—turns out there are now 65. We have over 20,000 applications in our market, my favorite is an app called FaceIt that displays a Dracula face you can put in front of your mouth that moves when you talk. Android is another product only in its baby stage, and yet we have already seen significant uptake. These types of projects take a lot of foresight to develop. We acquired Android in 2005, so it spent quite a while in gestation before launching. We also have over 60 carriers in 49 countries and 19 languages. Android has changed my life and I can’t wait for what it does next.
Our employees, or Googlers, as we call ourselves, now number about 20,000. This seems like a big number. But given the importance of the web, we think there are not yet enough people working in earnest on the many exciting opportunities in technology. Our challenge as we expand is to keep everyone organized and motivated. This keeps Sergey, Eric, and me quite busy, and I’m sure it will keep us and the rest of the team engaged for a long time to come.
Google has grown very quickly in the last eleven years. While we’ve undoubtedly had a lot of good luck, we have also worked really hard on search and advertising for more than a decade. That focus has paid off, both for our users and our business. Google is now a much larger company, and with size comes scrutiny and a certain amount of skepticism. We get that. But we also know that while new technology is often disruptive, it can help solve many of the problems we face in the world. We’re excited about the possibilities before us at Google and plan to work hard to make those possibilities real.