Douwe Osinga's Blog: May 2011

Friday, May 13, 2011

If Google is the anti-Microsoft and Facebook is the anti-Google...

Karl Marx' Dialectical Materialism might have gone somewhat out of fashion, but in this blog post I'd like to make the case that it is more or less applicable to the evolution of tech companies. Marx took from Hegel that the internal tensions in an idea lead to an opposite idea, but applied it to the material world. So the ancient society gives way to its opposite, feudalism, which turns into its opposite, capitalism, which will turn into its opposite, socialism.

IBM
Big Blue was the first king of the then called computer industry. From the first commercial computers in the fifties all the way through their wildly successful series 360, IBM successful beyond compare. IBM dominated the industry to such an extent that they called the other computer companies the 7 dwarfs. IBM was a consensus oriented, management driven company who saw hardware as its true product. It wasn't until the PC revolution arrived in full that a new contender for the crown stepped up.

Microsoft
Microsoft came into its own as IBMs sidekick in that revolution, providing the Operating System for the IBM PC, but by providing the MS-DOS also to the clone builders, Microsoft was at the same time responsible for destroying IBM's supremacy in the PC world. Microsoft was the exact opposite of IBM: a market driven software company that would do whatever to win. IBM's natural base was the mainframe, Microsoft's mantra a PC on every desk. And no later than OS/2's humiliating failure to gain traction against Windows 95, it was clear that tech had a new king.

Google
And what a king it was. Microsofts dominance seemed so complete and their policies so aggressive that in May 1998 the US Department of Justice and 20 US States filed a suit for abuse of monopoly, almost exactly 8 months after google.com started serving traffic.
Google again was the anti-Microsoft. Engineering driven, rather than market driven, with "don't be evil" as its slogan. Google delivered its products as ad supported free web services, rather than selling installable software for PCs. By the summer of 2005, it was clear that even though Google still had a long way to overtake Microsoft in profits let alone turnover, Google was the new center of gravity for technology.

Facebook and beyond
Many have suggested Facebook is the new king in waiting. Fittingly, Facebook started up in the valley just around the time when Google prepared for IPO. Facebook is an anti-Google company; where Google's engineers are computer scientisty who want to solve hard problems and think that serious software is written in serious languages like C++ and maybe Java, Facebook's hackers will do whatever makes things work and think PHP works just fine. Google is about the algorithm and doesn't really get social.

The genetics of companies
Then again, Google is also the anti-Yahoo, focussing on ranking vs hand optimized, trying to deliver one answer quickly and to get the user somewhere else as quickly as possible, rather than offering everything the user might need. And Twitter is the anti-Facebook.

Analysts often talk about the DNA of companies; if so then founders and early employees are the carriers of that DNA. They bring ideas of how a company should run from practices they admire or hate in existing companies, especially if they've worked for those companies. In the early days of the company the culture is formed by merging the parent genes and often flipping bits. This company DNA in turn is imprinted on new employees. When these employees leave to start new companies, the cycle is completed.

Triposo

In my new venture we build travel guides for mobile devices. Since we have two ex-googlers on board it shouldn't surprise anybody we derive some of our DNA from Mountain View:
  • We think information is something to be organized, not owned. So our travel guides are built automatically out of content that is available under an open license. Once we get to the point where our users can correct our data, we'll make every attempt to share those corrections back with the community.
  • We think that computers are better at ranking than humans even if they're editors. Like a friend remarked: Triposo tries to do to Lonely Planet what Google did to Yahoo.
  • Free lunch! Well, we haven't really established that, but I like it. In fact we had free lunches at Oberon in Amsterdam before Google was founded.
Some genes of course have flipped and make us an anti-Google of sorts:
  • Google's products are web services that might work on mobile devices. Our travel guides are mobile apps that work even better if there is an internet connection.
  • We fear overengineering as a thread to agility, while Google tends to think that real solutions come with many layers.
  • Google doesn't talk about products that haven't launched yet. We believe that talking to outsiders about a good idea can make it great.
It's going to be interesting to see how things will develop.

Sunday, May 8, 2011

Leaving Google - part 3

This is the third post in a three post series about me leaving Google. If you missed it, the first part is Why not to leave Google, the second part is about Why to leave Google Anyway. This one is about what I am going to after leaving.

Way before...
In the time many now call the first dot-com boom, me and my brother where traveling in Central Asia. Seeing how everybody was carrying a Lonely Planet, we were speculating that the travel-guide-as-a-book must have had its longest time. The web filled with stories from actual travelers surely would take over soon. So we came home and started world66.com, a wiki type travel guide. Our plan was to do this on the web for a while, but quickly move on to mobile devices, since there was a lot of talk how this was going to be the new thing. Fast 3G connections would replace dial up internet and WAP was going to bigger than HTML ever was. We even built a bunch of PalmOs travel guides and offered them for sale.

World66 did ok, but the mobile bit never happened. We sold the company to Internet Brands. My brother went off to run Oberon and I traveled the world with Google. But this itch that better travel guides were needed remained.

Enters the iPhone
Things started to change of course once Apple came out with their phone. Then Android came. Then the iPad. And now the tablet clones are coming. These devices aren't built per se for the next generation travel guides, but they'll do quite nicely indeed. So about a year ago, we started to experiment. We started by taking 7 different datasets available under free licenses like the Creative Commons and mixing them up. Wikitravel and World66 have nice general coverage of the world when it comes to travel, but they generally don't know where things are exactly. Open Street Maps has that information, but lacks the depth that Wikipedia can provide. TouristEye has a fresh look on new places. Dmoz and ChefMoz have been around for a while but still have a large set of businesses and restaurants.
Throw in some crafty Python scripts and let it stew for a while and a travel guide for the world rolls out. From this, with a bit of hand work we cut city guides for major destinations and launched them for both Android and iOs. Because of the Open Licenses we can keep all the data on the device so they work even where there is no Internet connection. Check them out and see where we are.

But that's not it

The picture below on the left is a travel guide from 1870 for Switzerland. The picture on the right is a Lonely Planet guide for the iPhone. Nothing much has changed. Murray's Handbook advises you that servants can be brought for free on a train and the Lonely Planet might have something on where free Wifi is to be had, but essentially it is still listings of things to see interspersed with general observations.







In the next months, we'll be rolling out a travel guide for mobile devices that actually takes advantage of being on a device and isn't just a straight forward port of a book. We have some great ideas that we think will make life on the road a little better. Some no doubt are things you have wondered about yourself:
  • If my phone knows it'll be sunny today, but rainy tomorrow, why can't it tell me to go to the park now, and leave the museum for the next day? When I look at the museum page on Monday, why can't it list the museums that are closed on Monday last?
  • If I tell my guide what I like in Barcelona, shouldn't it be able to tell me what I would probably like in Prague?
  • I like a coffee around 11AM. My guide should be able to learn that and when I take it out of my pocket around that time show me a little message saying "there's a great coffee place just 200 meters from here".
You get the idea.

So Triposo

We're a team of four. Me, the already mentioned brother (Richard), another brother of mine (Vincent) and Jon Tirsen, another Ex-Googler. You can check out our website, though since we're a mobile first company, you'll probably learn more from our apps. Drop me a line at [email protected] if you want to share ideas, express enthusiasm or collaborate. I'll be traveling quite a bit in the near future, so I might be in a town near you soon.

Saturday, May 7, 2011

Leaving Google - part 2

This is the second post in a three post series about me leaving Google. If you missed it, the first part was Why not to leave Google. This post is about good reasons to leave Google anyway. The final instalment will be about what I am doing next.

Google thinks Big

If you pitch an idea or a project to Larry and Sergey, their feedback is quite easy to anticipate. They'll tell you you have to solve the problem in a more generic way. I tried to sell them on data communities, a place where like minded people could collaborate on structured data around topics they're interested in. The feedback was predictable: why restrict yourself to communities? And why to structured data? Come up with something that solves everything!

Thinking big sounds great, but most big ideas start small and go from there. Google itself started from the notion that it would be interesting to look at back links for pages. Twitter started out as hardly more than a group SMS product that also works online. Facebook explicitly restricted themselves at first to one university.

Wave is a case in point. Wave started with some fairly easy to understand ideas about online collaboration and communication. But in order to make it more general and universal, more ideas were added until the entire thing could only be explained in a 90 minute mind blowing demo that left people speechless but a little later wondering what the hell this was for.

At triposo.com, we use Wave extensively for taking notes and developing ideas and it works great (yes, Wave isn't dead yet). But as a product pitch, that on its own wouldn't have made it in Google. The problem with this for an individual engineer like me is that you can't work with a small team on a medium sized idea, get users and expand from there anymore. You either have to pitch something as the third coming of Steve or your idea will be relegated to being a feature of something else.

Google has a Plan

In building 43 in Mountain View, there used to be a white board with the Google Master Plan, a huge flow chart that started with hiring smart engineers and led ultimately to user happiness by way of steps like flying robots and GoogleOS. All in good fun of course, but I guess many people did suppose there was such a plan. There wasn't. Any resemblance to later Google execution was purely coincidental (though it did talk about a GoogleOS).

Google's mode of operation used to be best characterised as strictly opportunistic. There were certain principles and leading ideas, but any project that met those and where Google thought it could do better than what was out there, would be taken on.

No longer. Google now has strategies. Once you offer an online spreadsheet and an online word processor, strategy demands that you also offer an online solution for presentations, even if it isn't actually much better. And you start seeing presentations with product road maps and competitive landscapes and unique selling points.

No doubt this approach suits a bigger company better. But the engineer in me wants to go back to that whiteboard; hire smart people that exploit new opportunities that become available as technology develops to build new products and services, which in turn leads to user happiness. Having a plan easily gets in the way of doing the right thing there.

Google has a Way

There is a strong and specific culture at Google. A culture that entails that you take nothing for granted. That the fact that the world has been doing something in a certain way is no argument not to do it in a totally different way. And this has led to many innovations - from the search first homepage, to the NoSql movement, powering webmail by Ajax and many improvements when it comes to running datacenters energy efficient. And these are only the things that have been published!

But that's the thing. Being secretive is part of that culture. Google hardly ever talks about things that haven't launched yet. This allows Google internally to be really open. There are very few internal secrets and that works because the rule is that you don't talk about anything to outsiders. Living as an engineer in this peculiar world with its own technology stack and ideas about technology is very exciting especially in the beginning. There is so much to learn, so many ideas to explore.

Google has done its bit to keep the web open and to fight anybody trying to turn it into a so called walled garden. But being inside is a bit like being in a walled garden. You can't discuss interesting developments inside with people on the outside of course, but it is even hard to partake in discussions outside. And so after a few years what you learn and what you talk about seems less relevant for the real world. Over time this starts to outweigh the advantages of being on the inside.

So then...

The first two points make it harder to build exciting stuff in a smaller team than it used to be. The last point makes it less interesting to stick around. So then when the next idea comes along, it's time to jump.

In the final episode in this trio I'll explain what this next thing is!

Monday, May 2, 2011

Leaving Google

Part 1: Why not to leave Google

I am leaving Google and thought it would be useful to put the answer to the question I often get as a reaction to this “what’s wrong with you?” in a triplet of blog posts. That and to promote my next endeavour. This is the first part, why not leave Google.

As a hugely successful company in an rapidly changing industry it is no wonder that Google is the continuing target of media attention. In the 7 years that I have been employed there, the way Google has been depicted has rather changed a lot though. In the first years, Google was this company of whiz kids who just tried things because they seemed like a good idea and surprised the world and themselves by creating wonderful products. A couple of years later, business analysts started writing about the Google Way and how by taking the long view and doing the smart thing Google succeeded where short term profit driven approaches failed.

Lately however, the meme of Google being the new Microsoft, being the new evil empire seems to be gaining traction. Of course if “don’t be evil” is your informal motto, you can be expected to be held against that standard, too. But it annoyed me to no end how the press was generally ignoring Heinlein's Razor, assuming malice where at most stupidity could be found.

So before answering the questions why I am leaving Google and what am I going to do next, I wanted to first rectify some common misconceptions about Google, implicitly answering the question why not to leave Google.

Google has gone Evil
Search for [google evil] and mostly you’ll find posts about how Google has gone evil. No doubt Google has become rather big and involved in many products and markets. This does lead to complications and difficult decisions. Either way you go on a decision, somebody will call evil. Google being secretive doesn’t help - the idea here is that information can be freely shared inside of Google as long as nobody brings anything outisde - it means that mistakes easily seem like conspiracies.
Meanwhile, it is still the case that any plan at Google can be tackled by pointing out that it would in fact be evil. Even if said plan would make a lot of money.

All the good engineers are leaving for Facebook
Well, that’s clearly not the case; I am starting my own thing for example. :)
Facebook and Google do employ of course the same sort of people. A drive between their headquarters takes less than 15 minutes. Of course, you’re going to see employees switch companies. And Google employs 25 thousand people, so in absolute numbers, this can be a significant number. But like the man said, leaving Google for Facebook is like divorcing your wife for her sister (Luuk added, younger sister) and I do think more people leave Google for startups than for Facebook. And so it should be.

20% time is a myth
Back when the business analysts were talking about the Google Way of innovation, a lot was written about how Google engineers could spend 20% of their time on whatever they wanted to work. Recently, there are more articles about how this is just a myth. At Google, you’re expected to do work 5 days a week, just like anywhere else. You can’t just learn how to play the harp or whatever.
Well, yes, of course. The myth might have been that you basically get a day off a week to do whatever. It is still work. And you’re still held responsible for what you do. So yeah, you can do whatever you want to, but you have to actually want it and it is still for Google. Which means it isn’t for everybody. Also, it isn’t always the next big product. Hundreds of engineers at Google spend their 20% time on mundane things like promoting test driven coding, cleaning up old code, mentoring new engineers or just helping out on another project that needs helping.

You can’t get anything done at Google anymore
There is some truth to this one. As an individual engineer you are going to have less of an impact on an organization if there’s 25 thousand people working there than on an organization with 25. But there is of course a flip side to that argument. Having a huge impact on a small organization can still leave you pretty much invisible. And as Chrome and Android have shown, if a project at Google does succeed, it can shift the (tech) world.
As an individual engineer you can still get your pet project out of the door on Google Labs. The world or even most of Google might not notice, but that’s the same as in your 25 person company. Anybody check out sharedspaces.googlelabs.com?

And they pay well, let me work where I want and Googlers in general make for excellent co-workers. So why quit? I’ll leave the Google bashing for the next installment in this series.