In the summer of 2007, as I was finishing up my MBA, I was faced with a choice of where to work.  I had this job at Google, and was also considering spinning up a startup on my own.  I had spent the last several years studying entrepreneurship and wanted to put that training into action.  One thing that really pushed me to Google was some advice from a very wise friend who was also looking for a new gig.  He said something like this.

“So many people have adjustable rate mortgages, it’s ridiculous.  They’re going to reset and people won’t be able to afford to pay them any more.  I think a lot of people are going to default on their mortgages and lose their houses.  I bet that will drag the entire economy down.  That’s why I’m looking for a job in a nice big stable company.”

He went to Amazon.  Having tried to start a small venture in a recession previously, I decided that Google would be a great place to hide out.  I also figured (correctly!) that I could learn a lot about how an extremely well-run company works from the inside.

Bottoms-Up Management

I learned how bottoms-up management works.  The idea of giving as much control as possible to the individual contributors who are working on the problems themselves.  Google exemplifies the saying “let 1,000 flowers bloom” with teams seemingly running in every direction at once.  It’s worked very well for them, and I can see a lot of value in this structure.  But I also see its limitations.

Having spent a while working on building infrastructure projects at Google, I experienced first-hand the limitations of bottoms-up management.  Individuals are empowered to make their own decisions, and they optimize for what’s best for themselves, which means their careers and their projects.  It’s easy to optimize towards a local-minimum which is great for your project, but not necessarily ideal for the company as a whole.  Similarly, top-down directives are extremely difficult to enforce.  Executives can make decrees about how things should work, and it just doesn’t happen.  I remember walking out of a review with Larry, Sergey and Eric and hearing one of the Directors respond “well that’s Sergey’s opinion.”

I spent a bunch of time trying to encourage teams to standardize parts of their backend infrastructure on shared components.  Shared technology is rarely the easiest or fastest thing for an individual project, so let me tell you the road was challenging.  I’ll refrain from pointing out the holes in Google’s products that I was trying to plug which are still open.  (To be clear, I’m referring to missing features and UX annoyances, not security holes.)  But I also learned some important lessons in how to design such infrastructure so that it is a win-win-win for all the teams involved.

Alignment with Customers

The best thing about working at Google was the clear alignment with customers.  Because Google is so strong in web search, they can afford to invest heavily in primary demand stimulation.  Like Campbell’s paying for an ad campaign that says “Soup is good food”, it makes financial sense for Google to work very hard just to get people to use the web more.  It goes beyond the motto “focus on the user” — Google really does want whatever the user wants, at least 99% of the time.

This makes huge classes of problem solving so much easier, because there’s no cognitive dissonance trying to remember the difference between a great user experience and a sound business decision.  This particular aspect of the business was an extremely refreshing change coming from Real Networks.  I remember working hard on re-designing the Real Player installer experience and tackling that specific problem head on.  The company had become addicted to the pocket change they could scrape out of their unsuspecting customers by surreptitiously taking over their computers when the Player was installed.  None of that BS at Google.

Looking forwards

Google faces some serious challenges in the years ahead.  For many of the reasons I pointed out in 2007, Facebook continues to be a serious threat to Google’s primary business.  Android faces an uphill battle against Apple’s dominance in the all-important mobile space.  Amazon seems to be squarely in control of the hosted cloud computing space.  And Microsoft seems to be pulling their head out of the sand for office productivity software.  None of these problems will be easy for Google to solve.  But I’m confident the amazing folks I worked with there will put in very strong showings.

As for me, I’m going to work in social media.  I’m very excited about my new role which I’ll write more about soon.

Economically, companies should be valued at the present value of their free cash flows.  That is to say, project forward all the possible ways the company might behave, and take a probability-weighted average (expectation value) of the total dividends the company would pay in each of these scenarios.  Discount these cash flows by an appropriate discount rate and you’ll get a fair market value for the company.  This is called fundamental analysis.

Now anybody who’s tried their hand at such financial calculations will know there’s a lot of judgement calls involved.  Small differences in numbers like discount rate or growth rates have huge effects on the results, and these numbers are hard to judge.  So it’s definitely possible to come up with a believable (by some) model of future cash flows that will value any currently successful company at whatever huge valuation you want.  But that doesn’t make it correct.  Is Facebook worth $300 per user?  It’s not possible for me to click on a $10 CPM ad every day for 100 years, but maybe they can add more users to grow into that?  Maybe?  It sure seems high.  I think there’s something else going on.

For context, think back to March of 2005 when Yahoo bought Flickr.  IMHO that made Google feel bad because Picassa wasn’t doing so well.  I think they saw this as a big missed opportunity to help organize the world’s photos.  I think this was big on their minds when they paid too much for YouTube.  And Google is still very far from monetizing this investment.  But they now control the dominant way that videos are communicated on the net.  This has to help them feel good about getting closer to their corporate mission of organizing the world’s information.  Since it’s not clear right now how they’re going to achieve that goal for photos.

Now consider Facebook.  Left and right, Facebook’s internal applications are surpassing total usage of th best dedicated net applications.  Their invitation app gets many times more usage than evite, and I believe their photos app is actually well beyond flickr in terms of usage too.  I don’t know where they stand for videos right now.  But it’s clear that they are a force to be reckoned with.  As I’ve written before, their application platform is potentially game-changing because it’s very attractive for information service developers and democratizes the process of product development in a novel and powerful way. 

For all these reasons, I think Facebook has the potential to dislodge Google as king of the hill.  No, Facebook isn’t going to become the dominant search engine, or even the dominant deliverer of internet advertising.  But I think Facebook could become the dominant way the humans communicate with each other using computers.  This could be the leverage they need to claim the crown of innovative thought leader on the internet.  If I were running Google, I’d be concerned about this possibility.  If I were running Microsoft, I’d be excited to get a piece of this.  Any piece.  Because even a tiny piece (like <2%) means that Google can’t take control of Facebook.  And yesterday, Microsoft got their foot in that door.  So, the game is on.  It’s gonna be fun.

Original Web 1.0

Universal access to massive volumes of data.  Being able to search
through masses of data and find what you want.  Connecting people to
huge databases really well.  Key examples:

• Online telephone books
• Web search
• Huge e-commerce sites

But in all of these applications, the data set is static.  User activity will not change the data for anybody else.

Web 2.0.1: Democratizing use of the data

The users of these data make the data better.  They can collectively
organize the data.  (i.e. tags)  They can help filter good data from
junk.  (i.e. voting)  Or they can help you find the data that are most
interesting to you.  (i.e. collaborative filtering).  In other words,
you can interact with the data.

Any "Web 2.0 company" worth their salt has an API that federates out their raw data.  This enables other sites to use the data in new and novel ways.  But the primary problem with this paradigm is that anything built using these API’s is done from the ground up.  Using the gmail POP interface, it’s possible to build a better UI for gmail.  But to do so you need to first build an entire AJAX mail client — no small feat.  Better would be the ability to add features into the gmail UI itself.  But this is really the standard in web 2.0 API’s today.

Web 2.0.2: Democratizing the feature set

The next big trend will be enabling users to make more compelling ways to interact with
the data.  Users can change not just the data, but how other users see
and use the data.  Sometimes this means API’s with UI hooks.  Or other ways to enable new functionality into an existing site.  This kind of platform enables Independent Software Vendors to improve upon the UI’s that the original sites created.

Facebook is doing this by allowing ISV’s to add new communications features to their site.  Google Maps is doing this through maplets that allow developers to create new ways to interact with mapping data from within the fabulous Maps UI.  Right now these are the only two examples of web 2.0.2 platforms that I’m aware of.

Building this kind of API is very challenging.  There are several very different ways to go about doing it.  Here are a couple of ideas:

• Provide a server-server API that includes content generated from the ISV’s servers into the main experience.  Facebook style.
• Allow developers to author XML files that define new algorithms that are interpreted on the primary host’s servers.  Yahoo pipes is a service in this style, but they’re not doing anything to enhance an existing service so it doesn’t really meet my 2.0.2 criteria.
• Allow developers to author javascript plugins to run on the client machine.  Greasemonkey is essentially doing this.  This strategy has the best shot for a lot of applications in the long term, IMHO.  But it comes with some serious problems right now.

Doing this correctly would allow ISV’s to add new features to Gmail.  Think about it — if I wanted to change the way gmail messages were displayed, or how addressing happened, or whatever it was, this kind of platform could provide hooks for making gmail better in a way that a POP interface never would.  And even though a POP-style interface theoretically could do this, there would never be momentum because having a high-level base to build upon means that there are network effects from the extensions.  (Rails achieves a similar advantage over other web frameworks — just having a standard, any standard, means people will build upon that standard rather than argue over which library to use and extend none of them.)

3rd-party javascript

The big problem with this approach is security.  There is none.  You need to completely trust the ISV before you should allow their code to run in the context of your site.  The kind of editorial review required to do this today would completely kill the democratic goal of such a platform.

The world needs a security sandbox to run third-party javascript code inside.  This way primary site hosts could allow ISV’s to run their code on client machines safely.  Here are a few examples of places this kind of tool could be used.

ISV’s could…

• Add UI features to Gmail
• Create alternate ways to share and discuss images on Flickr
• Define new mathematical formulas to run client-side on a web spreadsheet
• Create new playlist selection / shuffling algorithms for Rhapsody

… and much more.  Even better, individual users (not developers) could pick which UI extensions they wanted to use.  Any site which provides such an API has democratized the feature development process in a very important way.  Not only does it provide a distributed mechanism to figure out which features are best, but it allows users to self-segment as to which features work for them.  Without such a mechanism the entire service must have the same features for everybody, which means product designers must play a political game where they’ll never make everybody happy.  Right now I think really only Facebook has solved this problem.

Building a security sandbox is an area that Microsoft could probably do best and fastest.  They are good at code API’s and layered security models,a nd they have a perfect place to do it with Silverlight and the CLR.  They’re trying to position Silverlight as a faster way to run DHTML, which is something else the world desperately needs right now.  But I just can’t imagine them doing anything this innovative or generally valuable.  Doesn’t sell more Office.  Doesn’t sell more Windows.  They don’t even really have many services that could use third party extensions, and they’ve lost touch with the ISV’s who might build such extensions too.  Google Gears could conceivably add such an extension.  There’s precedent there considering the javascript threading extensions they provide.

This will be a difficult problem to solve, I have no doubt.  But I hope somebody with the resources to leverage a solution takes it on, because I think it would really make the world a better place.

• Build your app without social networking data
• Start from scratch with your own social network
• Integrate your app with Facebook

The third choice is so simple, it is the obvious best choice for most new information services.  As I see it, this is the fundamental power of the Facebook platform and why they’re going to go very, very far.

What I’ve been thinking about building in my copious spare time is a web-based feed-reader that assumes you over-subscribe to feeds.  That is, it expects you to "subscribe" to more feeds than you can fully consume.  These days many of the most popular feeds on the web meet this criterion even if that’s all you subscribe too.  I don’t have time to follow any one of TechCrunch, Scobleizer, Engadget, or even Radar in their entirety — I generally don’t even get to skim all their headlines.  But I know people in my social network do, and when they do it would be a small extra effort for them to help me identify the posts that are worth me reading.

This could be done by explicitly recommending articles to friends, or by tagging, or rating, or any of a number of well-understood-yet-often-poorly-implemented mechanisms.  Additionally, I could subscribe to a meta-feed coming out of a single-friend or a set of people in the social network graph that could expand several levels.  And of course there would be meta-feeds covering the aggregate opinions of all users.  The result would be that I could "express mild interest" in a feed by "subscribing to it" and the system would help me figure out which of the voluminous posts were actually worth reading.  Or if other users tagged posts, I could find good posts on a particular topic.  It would encompass a lot of the utility of digg, techmeme and link blogs all at once.  Another step in the process of democratizing information consumption.

I’ve been talking with friends about building this in the context of a facebook app for reasonably obvious reasons.  I’d call it "the outside world" as a reference to the fact that college kids are generally so isolated from external news, and this would be a social way for those few who do read the traditional-news to share good things with their friends.  Facebook’s restrictions on apps processing social networking metadata would make somegood features difficult, but the advantages in marketing and lower barrier to entry probably outweigh that.  Now my idea is out there for the world, so I’m not getting a jump on anybody.  If anybody wants to take this idea and run with it, drop me a line and I’d be happy to help advise.  I might just do it anyway because the Facebook market and the Google Reader market are both healthy and the basics just aren’t that hard.

But it sounds like you’ll have stiff competition.  Quoting from Blogoscoped’s analysis of the video:

Google’s recent big social effort is called Mocha-Mocha (or
Mocka-Mocka?), and will become the infrastructure for all social stuff
across all of their applications. As a part of this, a new
feature called Activity Streams will be introduced or at least
implemented in Reader this quarter. This will be comparable to
Facebook’s News Feed (Minifeed?) feature, and integrate Gmail’s
addressbook and contact list.

Also there will be some other Gmail and Orkut integration, but this might just mean there will be links to Reader.

Hearing that Brad Fitzpatrick has joined Google and because it’s the kind of thing I do, I’ve been putting some thought into how Google could reasonably add social networking features to their services.  I’ve been talking to folks about how Facebook is currently Google’s biggest strategic threat because they’ve done such a good job integrating the social network into new feature development, and in doing so have democratized new feature development in a way the world has never before seen.  This need struck me as a good way to start integrating social networking features into Google. 

Orkut is and only ever will be a toy IMHO.  Let the Brazilians keep playing with it and don’t push it on the rest of us.  Between contacts and knowledge about whom we chat and e-mail with, gmail has vastly more meaningful set of social networking data.  As we’ve learned watching LinkedIn and okcupid and other social networks thrive side by side, it makes sense to have different social networks for different purposes.  Orkut is a toy network and should not be the basis of anything more meaningful.  Sorry, Orkut.

Amazon makes extensive use of so-called A/B testing to try out new UI’s and optimize the user flow.  This works very well for them because their end goal is very well defined: they want people to buy stuff.  They are facing a very hard optimization problem, but their objective function is clear and easy to measure.  So they can try out new UI’s for 1% of users, and if it does well according to this well-defined metric, roll it out to a broader audience.  This is essentially best practice for any modern successful online company.

Google has done a lot to democratize the internet — notably by democratizing search through PageRank which allows anybody to implicitly vote on the relative merit of a web page.  They have also democratized the way some product development choices are made through through their policy of encouraging developers to build whatever they want in 20% of their time.  The result is that everything you can possibly imagine is probably being worked on by at least one googler, and the ideas with merit gain momentum and get built into real services.  But before they get launched to the public they still must be approved by a central authority.  Sure Google does A/B testing like everybody else, which is great for UI tweaks and to verify that new services won’t crash when hit with massive traffic.  But it’s extremely difficult to do A/B testing on major changes to functionality.  For example, it’s hard to imagine testing a change to how g-mail delivers mail through this kind of test.  Moreover, depending on how the test goes, the change is either rolled out to the entire user base or not at all.

Facebook’s platform offers another alternative.  ISV’s have the opportunity to offer major new kinds of functionality to Facebook users in a very democratic way.  Users can try out the new features, and if they like it, they’ll tell their friends about it, and the feature will spread.  Some features which are only appropriate for a certain segment of a user base can naturally find that segment.  This mechanism doesn’t really lower the cost of adding new functionality compared to how Google does it — Google is always launching new features that you’d never know about without reading their dozens of product blogs.  But it democratizes the process of figuring out which of these new features are valuable enough for a mass audience.  To continue with the democracy analog, these decisions are still made by a communist-style central-planning committee in Google’s world, whereas Facebook users can vote with their keyboards on what features are worth using.  This will make the Facebook platform very competitive in the arena of user’s attention.