The engineering staff at Google threw a big party for Silicon Valley nerds last Thursday night, complete with band and Cinco de Mayo-themed food and drink. The last time I visited was so long ago that Segways were still cool (Google still has a few but today they gather dust in a corner). Google has grown up to employ over 3000 people and occupies a campus built for Silicon Graphics (SGI; kids: this was a Unix workstation company that bloomed in the late 1980s and faded as Sun grew). The center is built around a volleyball court and an endless pool, complete with lifeguard until 9 pm. The company provides all of the fun things that profitable companies can provide, e.g., haircuts, massages, day care for kids, free meals, etc.
Larry Page, one of the founders, gave an inspiring talk about what a great time this is to be an engineer. He recalled how at one point Google had five employees and two million customers. Outside of Internet applications it is tough to imagine where that would be possible. Page also talked about the enjoyment of launching something, getting feedback from users, and refining the service on the fly. The Google speakers made a persuasive case that there is no better place to be a programmer. No startup company is going to have a 5000-machine cluster available to launch a new service or a guaranteed first day audience of 100 million people. Financially it might also make much more sense to work at Google as opposed to a startup. For teams of engineers who create a lot of value for Google the company is able to hand out $millions or tens of $millions in bonuses, to be shared among a group of 5-10 programmers. That is admittedly a small percentage of the new advertising reveue that Google earns from a new service but it is in absolute terms more than someone is likely to make creating the same service at a startup, where hardly anyone is likely to find out about it and use it.
One of the anecdotes that Page related was about an experienced Silicon Valley executive who told him, several years ago, “in the long run, every company is led by either marketing or sales; you just have to choose which it is going to be in Google’s case.” This prophecy does indeed seem to be true for the big tech companies. Microsoft never does anything because an engineer thinks it is fun or cool; they wait for the marketing department to notice a new product from a competitor and then go to work. Oracle seems to be led by their sales organization. They add features if customers are telling the sales people “this is what I need to make it worth buying the next release.” Google remains an engineering-led company. They launch Google Maps with satellite imagery because they can.
As I wandered through the party and through the offices I kept noticing more and more familiar faces and the names of former students whom I remembered as among the smartest and nicest. They will, of course, need all of those smart people if they are to deliver on their long-term goals. Doing search right will eventually require machine understanding of natural language, i.e., full artificial intelligence.
So, would you demand more than a dollar a year to work at Google?
I don’t know if it’s the “last” best place, but it’s certainly among the newest. It does seem like a cool place to work. Beyond the things you mentioned (day care for kids, massages, etc; all things offered by MS), I also like the fact that they give you something like 20% of your time to experiment with new stuff, something that I don’t think even MS does. It’s a great idea. It’s suprising (not) that it took so long for someone to figure out that company growth comes from inspired, smart, *and* employee experimentation. And by putting in a policy of letting engineers bring “hobby” projects in-house, they are better able capture innovation that might be otherwise spunoff into their own companies.
Hmmm, perhaps it’s high time I sent in my application to Google. 🙂
It is not surprising that mature companies such as Oracle are led by marketing and sales. They are in maintenance mode in which the organization structure, decision-making process, and leadership temperament all have been selected to innovate on sustaining technology — giving customers the new features they want in Oracle’s database. Not surprisingly, the people most likely to know what customers want are in sales and marketing. Google, on the other hand, is still young enough that anything cool but not guaranteed to make money is still given consideration. Google is perfectly set up to work on disruptive technologies. An engineer at Oracle who tries to bring up free satellite maps as a “business idea” will probably get, on his next performance review, some comment such as, “Engineer Smith needs to focus more bringing world-class engineering to better serve our existing customers instead of coming up with technically interesting but irrelevant ideas.”
The ideas on disruptive technologies are not new. Professor Clayton M. Christensen has been discussing disruptive technologies for a long time.
I don’t think it’s accurate to say that Google is an engineering-led company. I think that they are precisely a marketing led company now. Take Google Maps as an example. Various companies have been putting US map data and route finding algorithms on the web for many years. Microsoft has had the Terra Server with aerial photographs of the entire US on-line for years as well (in many areas which much higher quality than those on Google Maps). Google bought Keyhole, another company that had previously strung together satellite maps. The innovative ideas were to put the vector maps and satellite maps together (which is of questionable utility) and to improve (marginally) the interface to searching and navigating the maps. Only a very wealthy company could do this (due to the high price for buying a company with all those satellite maps and licensing the vector map data) and only a company that’s a media darling and a marketing dynamo could bank on the coat-tails effect to make their service that’s only a marginal improvement over existing services, immediately so popular. I think that there are lots of great things about Google, and they certainly do have an interesting corporate culture. But their marketing savvy, and the sycophantic attitude of the popular press, are key components to why some of their new services are very popular.
Samuel,
Take Google maps for example…
Google is implementing a project that up to this point was only attempted by government agencies.
Why haven’t other prvate sector companies pursued this at length? For all of the reasons that you list, eventually the service just doesn’t meet the bottom line.
I was extremely lucky to get entry into IT through Geographic Information Systems. A well funded government agency was willing to invest millions of dollars, primarily to improve the safety of the users of public highway systems. In the eight years that I worked with the technology I was able to expand into imagery, 3-d topo, trip routing, dynamic segmentation (display of live tabular data) and cartographic publishing.
No private company could have justified the investment, but as we progressed we moved from integrating data, to departments, to agencies… and towards the goal of ‘Societal GIS’. I remember getting blown away by a Jack Dangermond speech in the mid nineties. He laid out the Societal benefits of an integrated geographic information system.
I did my bit to work within the constraints of government before I bailed to the private sector (just dry database work, no tasty mapping).
I cannot express my elation that a company as dynamic and well funded as Google has gotten the gis bug.
Read these articles, maybe you will to:
http://www.gsd.harvard.edu/users/pbcote/enr-100/outline.htm
http://www.gisdevelopment.net/technology/gis/ma03022.htm
hey phil: why dont you work at google — not for the $, for fun. you may find it quite enjoyable. best place i have ever seen.
I think it’s a terrible idea to have the smartest people concentrated in a few companies. Just in case Google becomes a target of terrorism, I suggest that Googlers dig themselves an underground shelter. In the true spirit of Dr. Strangelove, they would also designate 10 … for every … to bring into the shelter.
Christian and “dc”: Why don’t I apply for a job at Google? My former students are all there and they know everything that I know! So why should I have to go into work every day week after week? The whole point of teaching the next generation is so that I can have some time to relax and enjoy. If my old students were to get stuck on a problem and needed my advice of course I make myself available to them. Ditto for any of my old classmates who are at Google. I still solve a lot of SQL puzzles for my working friends.
Gary,
I think GIS is wonderful, but I don’t think that it’s true that Google is implementing a project that up to this point was only attempted by government agencies. Microsoft’s TerraServer has been on-line since June 1998, and I’ve had vector-based maps on my computer since 1991, when I bought DeLorme’s Street Atlas USA CD-ROM. MapQuest has been on the web since 1996. GoogleMaps does not seem to be a technical achievement—if there’s any real technical innovation that would be in Keyhole, the company that Google bought which put aerial photography and vector-based maps together.
Thanks very much for those two interesting articles!
–Sam Lipoff
Samuel,
The true value of GIS comes from the integration of different data sources.
This information may be geographically based like the data that terra server and delorme sell (which are generated by government agencies), tabular like business addresses, us census or crime statistics (still government generated), or…
Perhaps drawn from widely held resources available over the internet, rectified to a common geographic standard and displayed interactively…
That is where Google comes in, and… would this be EVIL!? Definately not, its just the google way.
Sigh… I have to go complete some docs on a new sop for a change control process to comply with new reporting standards…
The clickable, draggable part of Google maps is a good idea, but as was pointed out, it’s not a Google invention. Google maps is integrated with databases of locations of addresses and businesses, but it’s basically the same kind of thing that Yahoo! maps offers, also. So I don’t see any real innovation coming from Google. And it goes without saying that my opinion is in no way biased by the fact that I was chewed up and spit out by Google’s interview process.
Rubber Bracelet
Sky Diving Movie