Google Analyst Day 2007 - Part 1

Uploaded by Google on 25.10.2007


MALE SPEAKER: Would you please welcome Google's chief
financial officer, George Reyes?

GEORGE REYES: Good morning everyone.
And thank you for taking time out of your very hectic
schedules to make time with us today.
So today we plan to showcase our coming talent in terms of
the product managers.
But really we're going to drill down on our strategy for
search, ads, and apps.
I hope you'll find the day very informative.
Jonathan Rosenberg will kick off our session with a
discussion of our search and apps vision.
Then Omid Kordestani will discuss our ads
and enterprise business.
He'll be joined by members of both the sales team and the
product teams to talk about our recent
progress in this space.
We will also give you some insight into how we think
about these businesses and what kind of investments we're
making in them.
We'll break for lunch roughly around noon.
And we will adjourn at 1 o'clock.
After lunch you'll hear two people who are truly Internet
And that's Allen Eustace and Vint Cerf.
Eric Schmidt will then share some of his
thoughts and insights.
At the end of the day, Eric will be joined by me and
Sergei for a Q&A session.
Of course as in past years, we will have product
demos to offer you.
And we will also give you Google tune-ups in case you
want to refresh your devices.
So also just as a reminder, there's free Google access
throughout the property here without a required password.
And if you're having any problems, we have one
technician back there that can assist you with all of your
computing woes.
So with that I'd like to bring up Jonathan Rosenberg.

MALE SPEAKER: Thank you, George.
JONATHAN ROSENBERG: So I always like to remind
investors to start their analysis of Google by looking
at what we said in the past, and in particular to the
single definitive source document the codified how we
run the business.
And that of course is the founder's letter.
I think it's important to see that we've maintained the long
run vision, the product principles, the 70 2010
guidelines, and the mission which we set
forth in that document.
So we used the past two Analyst Days to give you
additional deeper insights into our strategy and plans.
What I'd like to do is formally welcome all of you to
this our third installment.
At the first Analyst Day back in 2005.
Eric, I'll remind you, talked about his vision.
At the time, he described what we now all dub today as cloud
computing, and how he imagined that we would build out the
end user value proposition that is actually in fact the
current basis for our business model.
Sergei talked about the culture, and how we're going
to take the culture to multiple offices
all over the world.
Larry reinforced the need for end user quality in everything
that we did.
I pointed out that access to information would be a strong
catalyst for growth in online commerce.
And Omid set forth the plans for the broader ecosystem and
all the partnerships that we would need to develop in order
to achieve our strategy.
Then you all came back for Strategy Day 2006.
And Eric talked about strategic acquisitions.
And he talked about the areas of focus at that time, which
were search and ads quality, partners, systems and
infrastructure, and of course doing all of this with an eye
towards a truly scalable architecture.
I talked about product innovation.
And I talked about the ample head room for growth we
foresaw in the monetization engine.
Omid ran a panel discussion of executives.
They talked about international growth, our
sales strategy, which of course by then we were
implementing aggressively worldwide and at scale.
In both of those previous Analyst Days, we talked about
the immense scope of our opportunity.
Our view here, of course, has not changed.
If anything, the opportunities created by the growth of the
Internet over the last 18 months are even greater than
we thought they were back then.
Just last week on our earnings conference call, Eric said we
are now on the cusp of a world where everyone can create,
share, collaborate, and find their content in the cloud
anytime and anywhere.
This whole new era of cloud computing has people around
the world generating tremendous
amounts of new content.
And with it, the amount of information on the web, of
course, is growing incredibly fast. YouTube for example
today has eight hours of video uploaded every single minute.
So what does all this mean?
Fortunately for Google, this is not an example Maslow'a
Observation that when you have a hammer everything
looks like a nail.
With all this information out there, it turns out search is
even more important than people may
have originally thought.
What do you know?
Search fulfills a basic human need for knowledge.
And getting all this new content authored and then
indexed in a way that's very powerful is basically going to
be a fundamental recursive loop.
And it's going to continue to drive search traffic for many,
many years to come.
It also means that this era of PC based applications and data
is transitioning to one of apps and content
living in the cloud.
Now users can focus on what they actually want to do
online as opposed to which application they need to use.
We've followed a path of development across the
enterprise and apps efforts that may actually seem random
to some of you all from the outside.
But if you think about it internally, we're meeting
these applications and these acquisitions
into a cohesive platform.
We've developed a common infrastructure for storage,
for annotations, for content, for access control, and for
real-time communication.
In advertising, you've all heard this story before.
It continues to be very, very robust. We've got an ad
network that's large.
It produces excellent conversions.
It targets ads judiciously.
And thankfully, as all of you know, it is of course
extremely lucrative.
It's delivering ads that are highly relevant.
They're useful to users.
And we're doing all of this even as we scale massively
both domestically and abroad.
So these are the basics around which Eric coined the Google
holy trinity of search, ads, and apps earlier this year.
And what's more, at the time he mandated that we deliver
against this strategy in 40 languages across all of our
product lines throughout the world.
These are still the pillars we're going to use today to
communicate our strategy to you.
So a couple of weeks ago, we ran a conference here called
Zeitgeist. We had a host of amazing
people exchanging ideas.
Al Gore was here.
Bill Clinton was here.
Fred Smith was here.
John Chambers was here.
Chambers told us how he is turning Cisco's organization
completely upside down.
He referred to it as blowing up his command and control
style and replacing it with a flat organization that
empowers everyone to act.
What he told us was that Cisco has gone from a company where
10 people were empowered to make important decisions to a
place where 300 people could make important decisions.
And he said that he was hoping to move that to 6000 people in
the near future.
Well like Cisco, here at Google we push a lot of our
product decisions down to the experts.
The product people and the engineers at the front line
who live and breathe these products every day.
When I first came here, I showed up with what people
described as the typically Dilbertian senior management
big company view that it was my job to quickly dictate
decisions by imposing top down strategy slides.
So I started walking around the
Google halls and dictating.
It was a disaster.
The Google patient quickly rejected the
Jonathan donor tissue.
And I learned very quickly that it was my job as a
manager here to aggregate viewpoints, and otherwise stay
out of the way of the legions of brilliant and smart people
who we hire and in whom all of you invest.
So Chambers' talk at Zeitgeist reminded me of that lesson.
And it got me to thinking.
Instead of having executives tell you about all of our
products, it'd be far more valuable insightful to have
the real experts on these products, the ones who
actually do the work, and the ones who make up the headcount
in that little line in your spreadsheet that relates to
employee growth that I know you're all concerned about.
So I've not produced traditional boring MBA
strategy slides.
But I hope you will all kindly stipulate, having worked with
me for the last six years, that if I really wanted to
bore you with damn good slides I could.
Rather, what I'm going to do is open the kimono a little
bit and give you a look inside the innovation machinery led
by the people who actually do the work.
What you'll see is a format that's very similar to an
internal Google product meeting.
Rather than going through mind numbing pie charts and endless
bullet points, we put our trust in the experts leading
the projects.
And we asked them to tell us their vision for the product.
And the bigger the vision, the better.
How will the world be a different and better place
once your vision is implemented?
Larry and Sergei always believed that strategically,
Google was about solving big problems for large numbers of
people, not small problems, and not even big problems for
small numbers of people.
So we asked how will large numbers of people being much
better off?
The product managers come in, and they
tell us their insights.
What do we know about the market that
other people don't?
How are we going to approach the product in a unique way
that's substantially better than anybody else?
What does our internal log data show about the products?
Are they growing?
Because if they're growing, that's the single best tell
that we have that they're going to be successful.
At Google we constantly feed the winners
and starve the losers.
This is actually the opposite of what you see in typical
large GM oriented companies with big divisions.
There what you often see is general managers desperately
trying to shore up the unsuccessful dogs by giving
them more resources.
Here we do the opposite.
Finally at Google, we say show them.
As you know, we ship early, and we iterate often.
Because we know that our users are better at discovering uses
for our products than we are.
These principles also apply to us internally.
We don't have people just talk about what they're doing.
We have them come in and show the management team a demo.
Show me a demo.
That's what the founders say.
Most recently, we've been insisting that these demos
show a more coherent experience.
Sergei has adopted a new mantra that he calls features
and not products.
And what he means by that is showing incremental
functionality in the context of a larger Google experience.
So who are these experts, the people in whom we're going to
put all of this trust?
Well one of the things that I'm proudest of in my tenure
here at Google is what we call our Associate Program.
It started as a bet between Marissa and me.
Which she said is that she could grow good experienced
product managers faster than I could hire
them from the outside.
So over five years ago, we started hiring folks
right out of school.
And we immediately put them to work on important things.
These are highly technical computer scientists with very
limited business experience.

When they get here, we give them actual responsibility.
Because we believe that fresh perspectives and real insight
trump experience and previous accomplishments.
We also know that the associate product managers
can't rest on their laurels like the
complacent higher ups can.
Because after all, they have no laurels to rest on.
Hear that guys?
So they tend to be ambitious.
They try things that other people think can't be done,
because they don't know any better.
There's really no way that my staff or I could keep up with
what's going on this revolution
that we call the Internet.
After all, if you want to understand social networking,
which is certainly a phenomena that's here to stay, it's
probably a good idea to have people working
on it that are dating.
For the official historical record, I am not dating.
So I'm going to let these managers who actually do all
the work come and show you what they're up to.
The demos that I'm covering are in the area
of search and apps.
And after our group, we're going to have Omid come up.
And his team is going to talk about ads and monetization.
So first we have Jack Menzel and Universal Search.
JACK MENZEL: All right.
Thanks, Jonathan.
So as Jonathan said, my name is Jack Menzel.
And I am a product manager here at Google.
I'm also a recent graduate from Brown University with a
degree in computer science.
And I'm here to talk to you guys about the vision of
Universal Search.
And so when Johnson told me do this presentation, I was
really struggling as to how I can convey to you simply and
concisely what is the essence of Universal Search.
What is the vision?
And I just couldn't do it until I broke down and I
really just embraced my inner geek.
And I invoked something which is very close to my heart,
which is Lord of the Rings.
And so I wrote you all this poem.
I hope you it.
Here we go.
I'm going to read you this poem.
Universal Search Universal Search.
One search to index them all.
One search to find them.
One search to bring them all, and in results combine them.
And while this is really fun, to harken back
to Lord of the Rings.
And it's always great to refer to this kind of really nerdy
stuff in my background.
In all seriousness, this poem embodies the three point major
points of Universal Search.
And those are number one, comprehensiveness.
If it's on the Internet, we need to fight it regardless of
whether or not it's a video.
It's a blog.
It's a news result.
Whatever it is, if it's not in our index, we can't give it to
you as a result.
Second major point is relevance and ranking.
Just because we went out and we found all these awesome
images and videos, and collected them all, and we
indexed them all, it doesn't mean we're just going to dump
them on you every time you do a query.
What we do is we build this fabulous infrastructure that
actually allows us to rank and index every single one of them
on every single query.
The scalable infrastructure is really amazing.
So no matter how long tail your query is, no matter how
obscure, we're going to run through all that data.
And we're going to give you a set of results which are
really relevant your query.
And then finally presentation.
Once we actually have these results, we give them to you.
We actually want to make them actually scannable and
efficient for you to find the information that you're
looking for.
And so that means a text snippet may be great for a
text web page.
But it may not be so great for a video.
So we're experimenting in that field.
So that's really the vision of Universal Search.
But what really gets me going are the demos.
So let's hop on over here to this machine
on the right here.
And let's bring up some actual queries.
So here we go.
So here is a popular query these days.
Here's a query for Hillary Clinton.
Hillary Clinton is a great example for a
Universal Search query.
Because she actually has a whole lot of really good
content of different types on the Internet.
And so let's just look through here and see what we can find.
Well if you just browse through this result page, you
see that we have the Hillary web pages.
We have a Wikipedia article.
We've learned a lot about her platform and all about her.
We now go into the news results.
Because Hillary is a very newsy query.
Then down at the bottom, you'll notice we've actually
blended in a video result, which is interesting because
while not as relevant obviously as the number one
result, which is Hillary's web page, we do think that is
relevant at this point.
And it really does make our results more
interesting and diverse.
Because hearing Hillary speak is an interesting telling
thing about Hillary.
Then finally we have news archive results.
Because she's been in the news for a long time.
And then we also have searches related to Hillary Clinton.
So you can actually explore the space around that area.
So that's a great example of how we've taken great care to
do a good job of ranking and relevance.
There's a lot of different types of content in here.
But for reading through this, you don't get the impression
we're just dumping it on you.
We give the impression that we've really taken time to
focus on just giving you a relevant list of results.
For the next demo, I actually have to share with you guys an
essential crisis that I recently
have been going through.
So I was on Facebook the other day.
I was hanging out.
And an invite popped up.
And it's from my friend, Ryan.
And it says Jack, I need you to join in the epic battle
between pirates and ninjas.
And we want you to be a ninja.
I was like whoa.
That sounds intense.
I'm enthusiastic but a little apprehensive.
So like any good researcher, I immediately bring up a Google
search on ninjas.
Because I need to learn what does it mean
to be a modern ninja?
And so I start digging through the web pages.
And I find real ultimate power-- and ninja's have a
great sense of humor actually--
from Wikipedia that they started in 15th
century feudal Japan.
But I'm not really getting an impression of what it means to
be a ninja these days until I get to this
inline video result.
And I find these kind of crazy guys in a gym.
And they're doing ninja flips.
And I'm like whoa.
This is awesome.
Like this is really what it means to be an urban ninja.
Oh they've got weapons.
This is great.
So what this really shows is how with the new kind of UIs
that we are introducing with Universal, we're able to
convey things that a text snippet would
not be able to convey.
It'd be pretty hard for me to write a two line text snippet
to say whoa, these flips--
get how cool these flips are.
So let me finish this off with rounded out.
Because we saw the weapons there, and that really struck
home with me.
I think I don't need those to be a modern ninja.
So I did a query for ninja weapons here.
And I see some pretty good relevant
results about ninja weapons.
Down at the bottom we blended in some image results.
But finally this is the one that I think is most
And this is actually an entire book result.
And we blended this in here.
And this really speaks to the
comprehensiveness goal of Universal.

It's amazing.
Because you can actually read the entire book here.
Whereas before, for me to actually find this book, I
might have had to go out to a library.
And then I would have had to hope that the library actually
had a high population of ninjas.
And then they actually have the book.
But we've stocked it here.
And it's amazing.
Because we've actually brought something online that was not
online before.
And that really is a great example of how Universal
Searches be more comprehensive.
So that's Universal Search.
In my future, it looks like I'm going to have to go work
really hard at this ninja training stuff.
Because I want to be prepared.
The future of Universal Search.
Well with this awesome scalable infrastructure that
we have, we're really just scratching the surface of
what's possible.
So what do you expect to see from Universal in the future.
Well you're going to see more result types, so more
different types of content.
You're going to continue to see it be ranked with the
utmost relevance.
And then finally you're going to see it be presented to you
in a way such that you can still really be presented with
a scannable list. And you do a query, scan down it, find what
you want as fast as possible.
So that is Universal Search.
Thank you, Jack.
Next we have Jessica.

JESSICA: Thanks, Jonathan.
So I started at Google about three 1/2 years ago.
And I had just graduated from Stanford where I studied
mostly math, and a little bit of computer science.
And I'll never forget my first day at Google.
Because Marissa sat me down, and she said you're going to
be in charge of the Google homepage.
And I was like I thought we were supposed to have
And I think of the homepage as a big logo, a text box, and
like two buttons.
I was like my dog can manage the Google homepage.
Are you serious?
And little did I know how much work that would
actually turn out to be.
So we realized that people at the time love the classic
Google homepage.
And so we don't change it much.
But we realize that some people want a lot more.
And that is why we developed the product iGoogle.
Now when I explain what iGoogle is to people, they
always ask me is this a portal?
And I think it's a fair question.
I think it's a very fair question.
Is this really a portal?
I think portal has come to be a very dirty word on the web.
And I think it's fair to ask why.
The word portal has come to mean not just information on
the page, which is fundamentally a good thing.
But it has come to mean a walled garden, a site that
tries to lock users into their own content, into their own
products, and doesn't necessarily give
you the best solution.
It gives you their solution.
Portals don't give you the best games.
They give you their games, their news, their email.
we wanted to take a different approach.
So iGoogle currently indexes over 200,000 feeds and over
20,000 gadgets.
It's a completely un-editorial, algorithmically
run product.
We don't editorialize iGoogle.
And we don't have any philosophy about what users
can and cannot put on the page.
If you want to put Yahoo mail on your Google homepage,
that's fine.
I think that this one single thing, this philosophy, has
been the secret to our success.
IGoogle has seen tremendous growth over
the past few years.
Since we launched the product two years ago, it's
consistently been one of the fastest
growing products at Google.
And that growth trend continues today.
So that being said, let me let me show a
little bit of iGoogle.
So it's not really good enough to have 200,000 feeds and
20,000 gadgets unless you can give the users the right feeds
and the right gadgets at the right time.
So really this is about search, not about
Let's go to iGoogle.
So here's our new set up that just launched this morning.
And it's entirely algorithmically search based.
So what am I interested in?
Well I'm interested in news.
I'm interested in games.
Let's select art, technology, and entertainment.
And I want a beach theme on my page.
So I click see my page.
Here's my page.
I have an entire corpus of tabs.
People think that we choose the stuff by hand.
We really don't.
It's all entirely run by algorithms. So let me put in
my location So say I'm in New York.
And I save that.
Notice that the theme on my page actually changed.
So the themes on the page are dynamically set to match your
local time conditions.
And actually the theme's timed right down to your local
sunrise and sunset time.
So let's look at my content.
So I have news.
I have games.
I have art.
And again, people think that we choose this stuff
But we don't.
So this isn't just limited to categories that we choose.
You can actually put in any arbitrary category here.
So let me put something I'm
interested in which is astronomy.
Notice that says I'm feeling lucky automatically adds stuff
based on the tab name.
So this is where search kicks in.
Now I have an astronomy tab where we have the best feeds
and the best gadgets.
How do we generate this?
Well what we do is we look at our users.
And we look at which users have a tab called astronomy.
And we look at which gadgets and feed are most popular on
those users' pages.
So in this way, we're leveraging the single biggest
advantage we have, which is millions of users organizing
and creating and customizing personalized homepages.
Whether users know it or not, they're making iGoogle better
just by using it, by tagging content, by
telling us what's important.
Let's try another example.
So travel.
This one's interesting because it comes up with some more
commercial results.
So you'll notice that there is a gadget here from Priceline,
which is simply a booking tool.
People think Priceline paid to be here.
They didn't.
Users actually find this useful, and will add it to
their homepage.
And so in this way, a lot of companies are creating
gadgets, and getting presence, and connecting with users on
their homepage before they even connect with a search.
And so you'll hear a lot more about gadgets as a form of
advertising in a later section today.
Let's try a local search.
So lets try San Francisco.

This one I think is interesting because it gives
me SF Gate.
I have food and dining.
I have some good gadgets here.
I have Bay Area traffic gadget.
I have upcoming shows in San Francisco.
See what's coming up.
There's an exotic dance cruise.

If that doesn't work, it looks like Kid Rock and Hansen are
back to back at the Fillmore which is good.
And so I can get on my local Bay Area information right
here on one tab.
It's really easy to organize.
Finally, I'm going to show you guys something really cool the
just launched yesterday.
Sign in here.

So this was a tab developed by the Google Finance team that
leverages one of our new gadget protocols, which is
called inner gadget communication.
So the Finance team put together this tab.
Now this is basically Google Finance in a componentized
structure on my Google homepage.
And I currently have a query typed in.
So watch this.
Say that I change the query.

The gadgets sync.
They actually talk to each other and inner communicate.
Let me try again.
Try a different ticker.
Watch that.
So in this way, the gadgets actually talk to each other
and respond.
I can decide which ones of them I want to think, and
which ones of them I don't.
I think this is really, really
interesting, and really powerful.
Because with a product like iGoogle, someone can take
Google Finance.
They can componentize it.
They can sync it.
They can create a version of Google Finance that exactly
works for them.
And this is incredibly powerful.
Because the web in and of itself is becoming
It's not just about driving people to your site anymore.
It's about getting your content out there whether that
be RSS feeds, whether that be gadgets, whether that be
social apps, whether that be tabs of content.
And what's amazing about the componentized web is that the
sum of the parts is actually greater than the whole.
When you take something like Google Finance and
componentize it, there's a ton of different ways that this
content can be mashed up, can be created, can be
And that's really what we want to do with iGoogle.
We want people to have their content in the way they want
it, and really have their best from the web.
Thank you.

JONATHAN ROSENBERG: Next we have Albert with Trends.
My name is Albert Chang.
And I'm a recent grad of Penn in engineering.
And my product is Google Trends.
So for those of you who are unfamiliar with Trends, our
vision for the product is to make our search data
universally accessible and useful.
And we consider a product like Trends--
It really is about tapping into the wisdom of the crowds.
And you think about all the millions and millions of
people who use Google every day to make their searches.
And it's really about having them answer your
questions for you.
So I'll walk you though some examples that really
illustrate what I'm talking about.
Can I get demo right please?
So here's a screen shot of Trends.

So I want to share with you some examples that illustrate
the power of Trends.
And I'll start with a personal one.
So when my wife was pregnant, we were trying to figure out
what to name our baby son.
And we had settled on the name Conner.
But we couldn't figure out how to spell it.
So was if going to be C-O-N-N-O-R or C-O-N-N-E-R?
And so we sort of went back and forth, back and forth.
And we are asking friends.
We were asking relatives.
And we just could not come up with a good answers.
So would did we do?
Went to Trends.

So you compare Connor versus Conner here.
And you can see Connor with an O-R is clearly more popular
than Conner with an E-R here.

Let me introduce you to my son, Connor Chang,
But more seriously, there's a lot of really good business
applications that Trends has.
So one example is something that Hyatt
went through recently.
They were building a hotel in the Caribbean.
They couldn't decide whether to name it Hyatt Regency
Trinidad or Hyatt Regency Port of Spain.
And they could have hired a naming agency to spend a lot
time and money to really figure out what
to name this hotel.
But instead what did they do?
They went to Trends.
And so they plotted Trinidad versus Port of Spain.
Trinidad is in the blue up here.
And Port of Spain is in the red.
And relative to Trinidad searches, no one searches for
Port of Spain.
So as a result, you can go ahead and book your
reservations at the Hyatt Regency Trinidad.
It comes online in December of this year.
Another problem that companies face is figuring out what to
stock for their shelves for the holiday season.
And this is a perennial problem.
And if they could only figure out what products were going
to be hot for the coming year.
So let's look at how trends would have
helped for that problem.
We're looking at two products here, American girl dolls and
digital picture frames.
These are popular gift items over time.
I can see last year a huge spike in terms of searches for
digital picture frames.
So for those of you who cover retail, you know that digital
picture frames were the surprise hit for last season.
And if you're a retailer and looking at Trends, because
Trends is now updated daily, you would have known as early
as October last year that digital picture frames was
going to be a breakout hit for the season.
So think of how useful and powerful
this is for a retailer.
The next example I want to show you something that Nissan
is actually is going through right now.
So they recently launched the Nissan Rogue which is a new
SUV. You've probably seen the commercials already.
They're everywhere.
They have product placements on the show Heroes.
They have TV advertising, print advertising, and online
But how do they know that they're
really making an impact?
And how do they know that their campaign really worked?
Well of course, go to Trends.
And so you see in the red here.
This is the Honda Pilot.
This is one of the competitors that they're going after with
this product.
And you can see from the searches for Nissan Rogue, it
really spiked right around lunchtime.
You'd see that indeed their campaign is working.
And so if you're the marketing manager for this campaign, you
can go to your boss and ask for more money.
Because you know what you've done is working.
And you can put more resources behind it.
So just showing you three examples of how Trends is
really powerful for businesses.
But let's not forget that we're dealing with
a ton of data here.
And sometimes when you're dealing with this data , it's
really hard stick to parse up the interesting insights and
create a story from the data.
And so we acquired a product called Trendalyzer which helps
apply these visualization techniques to
macroeconomic data.

So here is a version of Trendalyzer.
Each of these circles represents a country.
Each of the colors represents a region.
And the size of the bubbles themselves represents
So we flip over to the chart here.
So what we're comparing here--
as you go further up in the chart, it's greater Internet
As you go further right, this is GNP per capita.
And so let's see what happened to this over time.

So in the '70s and '80s, not very many people were on the
Internet obviously.
But as soon as you hit 1990.
You see a huge surge in Internet growth.
And you see that across the board.
So really a massive change in terms of usage.
And you can look at it from different regions as well.
You can see how in Asia, they're doing pretty well in
terms of adoption and also in overall income, whereas a
region like Africa, they're really lagging behind
particularly on the income axis.
And so we can select particular
data points as well.
So let's look at India.
Let's look at China, in look at the US.
And then we'll look at it on a linear scale and see what
insights that provides us.
It will play back over time.
So I just expect here that richer countries are adopting
the Internet faster.
And I can see where they progress over time, and where
they are right now.
And so it's interesting to see places like China and India
who you commonly think of as these huge power houses in
terms of adoption of the Internet.
But if you look closer at the chart, they are really not too
far different from Fiji in terms of Internet penetration
or per capita GNP, or the Ukraine or Panama.
And so it reminds us that there's a huge runway of
growth ahead of them for these regions.
And so I've shown the visual agent power of a product like
And you think about the data that's in Trend.
You combine the two.
And you can really think about really compelling information
for whatever audience you have and whatever focus area you
have. And so I encourage you to try out both Trendalyzer
and Trends.
Thank you.
Thank you Albert.
Next we have Keith who's going to talk to you about Gmail.
I'm Keith Coleman.
I came to Google about three 1/2 years ago from Stanford
where I did my undergraduate master's in computer science.
And since coming here I've been working on a growing
group of products that includes Gmail, Google
Calendar, Google Docs, and Google Reader.
And today I'll be talking primarily about Gmail.
Now a lot of our focus with Gmail is on fundamentals,
making the service really fast, and giving users access
to their information from any device from
anywhere on the planet.
And we've been working on some big things here.
I have a few new changes to show you guys today.
Let's start with speed.
Mary's always telling us that slow products never win.
So we're fanatical about performance.
We have JavaScript timers in all of our apps that tell us
down to the millisecond how long each request takes.
And for what we don't believe those, we actually have stop
watches on our desk.
We sit there and time every request that we
think is too slow.
Now when we wrote Gmail back in 2004, we coded it in a
programming technique called Ajax.
This is back before the term Ajax even existed.
And we learned a lot in this three 1/2 years about how to
build good, fast, large Ajax apps.
And pretty soon we're going to be rolling out a completely
new JavaScript architecture for Gmail.
No one has ever seen this before.
But I'm going to show you guys a demo here today.
So if you can pull up the left laptop please.
This is the new Gmail JavaScript architecture.
You can see it looks exactly like Gmail does today.
but under the hood it's completely different.
Now as you guys use the web, you're used to clicking on
links, waiting maybe a second, maybe even several seconds for
a page to come back.
But we think that's too long to wait if you're reading a
lot of email.
So the new version of Gmail actually pre-fetches all of
the messages on the page.
So when you click on them, they load instantly.
Let's see here.
And then you can click Next, Next, Next, Next.
You can even hold this down and just watch all
the mails go by.
That's how fast we can make JavaScript applications today,
using the newest browser technology and the newest
computer platforms.
And this thing will really shine on MacBook Pros, on the
new version of Safari.
That's going to be released soon.
You'll really see more and more speed gains as you switch
to these newer platforms. Now speed isn't the only thing
that's different about this new JavaScript implementation.
As Jonathan pointed out, a lot of our focus is on building
reusable infrastructure.
So when you load the contact manager in this new version of
Gmail, you actually get the same contact manager that you
see in Google Docs and in Google Video.
And when you compose a mail, you'll be using the same rich
text editor that you see in Google Page
Creator and Google Groups.
The Gmail team didn't have to write any of this code.
The architecture just let us put it into the product and
reuse everything that they had already done.
So you get the same experience everywhere you are in Google.
Now one of our other philosophies at Google in
general, and in particular with Gmail is that we should
never hold users' data hostage.
In the case of mail, that means you should be able to
get your contacts and your mail from anywhere in the
world and use it any way you want you.
And in the case of mail, we launched back in 2004 free POP
access and auto forwarding for all of our users.
Pop access lets people download their mail onto
devices like Blackberries or into clients like Outlook.
And forwarding let's them forward their mail to other
services or other accounts.
It even makes it easy for people to switch away from
Gmail if that's what they want to do.
Now I see that a lot of you guys out there have
So you're used to this really nice feature where when you
read your mail once, it's marked read on the server.
So if you delete your mail on your Blackberry, it's deleted
on the server.
The problem with POP is that it doesn't do this.
You still have to read your mail twice.
You have to mark it as read on your device.
And then mark it as red on the server which is pretty
There is another protocol called IMAP
which solves this problem.
Now it provides such a good user experience that a lot of
web mail services, particularly those that rely
on advertising for revenue, don't give this
out to their users.
And they don't give it out for free.
But we think that's a trend worth breaking.
So earlier this morning we launched free IMAP
for all of our users.
And now when our users go into Gmail and go into their
settings, they'll see not just forwarding a
POP but also IMAP.
So they can get that same really killer Blackberry
experience on their iPhone, even on Outlook
if they want to.

As Jonathan mentioned, one or other mantras here is features
not products.
And Gmail is a really good place to make this happen.
Because people use email for a lot of things.
They use it for sharing vacation photos.
They use it for coordinating events.
They use it for collaborating on documents.
And the experience here isn't always as good as it could be.
I'm sure a lot of you guys have been on threads like this
where you'll have a series of emails with slightly different
versions of the same Microsoft Word attachment.
Then you'll have to at the end go through and merge all the
changes into this one final version.
We have a product here called Google Docs which makes this
work better.
You just have one document online.
Everyone can edit it in the same place.
So we ran this experiment where we put this open as a
Google Document link next to all attachments in Gmail.
People can click it.
And it just opens the document right in Google Docs, gives
you that one version that you can edit from everywhere.
And it turned out that users really like this.
So if you switch to the right laptop,
you'll see the reaction.
This is Google Docs traffic after we
launched this feature.
And we saw more than a doubling in the number of
users that were going to the product.
And while this is just sort of scratching the surface of what
we can do, it shows how promising this direction is,
that by integrating really useful functions from other
products into a large product that a lot of people are
using, we're actually able to give them a better experience
and also introduce them to these new concept that they
aren't otherwise seeing.
And this is the direction we are going to be heading in the
future, where rather than having to come to Google and
figure out which product you first want to use, you just
come to Google.
And based on your task, we automatically direct you to
the best product for the purpose.
Thank you.
Next we have Rajin who's going to talk to you about apps.

My name is Rajin Sheth.
And I am product manager for Google Apps, which are out
collaboration products for businesses and organizations.
I graduated from Stanford University, and came to Google
three 1/2 years ago from VMware.
And I actually remember the exact date that I decided that
I really wanted to come to Google.
And that was actually when I got my first Gmail account.
So I logged into Gmail for the first time.
And the experience was amazing.
And there were a couple of things that I noticed about it
that were especially critical.
So the first thing was that it was an incredible experience
for email that I have never seen before.
I had thought that email was a solved problem.
But Google had managed to innovate on email and really
build email from the ground up again for the connected and
information-heavy world.
And as a result of that, it was amazing for me to see that
I could actually use email in a very different way than I
was using into before.
The second big thing was that it was a heck of a lot better
than my corporate mail.
And all of a sudden I was able to see long threads in a much
more digestable way, find information in
a much easier way.
My days of foldering mail and deleting mail were over.
So I was realizing that the innovation in the consumer
space was happening much, much quicker than
the enterprise space.
And so I was inspired to come to Google to help bring that
train of innovation to enterprises and businesses.
I've been driven by Eric's strategy about having apps as
the third pillar of Google's strategy.
And I want to share with you a few of the insights that we
picked up along the way, and actually show you them with
some of the new features we've had with some of our products.
So the first trend is all about connectivity.
How can we make these applications work really well
using the fact that people are connected to each other, and
people are connected to the web while they're doing this.
So take spreadsheets for example.
It used to be that innovation in spreadsheets was about the
next great toolbar feature.
But we believe that we can create a potentially even more
powerful application by focusing on how people can
collaborate with each other, and how people can leverage
the world's information when they do this.
So here's an example here of a
pharmaceutical industry dashboard.
So what I've done here is put a very simple
dashboard in place.
And actually the only thing that I've entered in here is
the company name Bayer.
So one of the things that I can do here is that actually I
can do a Google search right from my
spreadsheet, and look up facts.
So for example, for companies I can look up the number of
employees that they have. And this is actually Google Search
pulling results in real time.
I can look up the stock symbol as well.
And so this is pulling in real time as well.
Using Google Finance, I can actually look
at the market cap.
And using Google News, I can look up the latest headlines
about Bayer.
So let's say I go here, and I want to actually go and add
the PE ratio.
So I can go here and add a column that's the PE ratio.
And then all I have to do here is look it up in Google
Finance, and say that I want to look up the
PE ratio for Bayer.

And it puts it right there in its place.
And you'll see actually these numbers will start
to change over time.
Because these are pulling in real time as well.
So another thing is I don't actually know the
pharmaceutical industry very well as most of you do.
So I don't know what are some of the other competitors in
this market that I should be tracking.
One of the other things that the Google Spreadsheets team
recently released is really auto fill on steroids.
So you think about auto fill as a traditional feature where
you take the number one and two.
And you can expand it to a list of three,
four, five, and six.
We actually auto complete against a variety of sets that
are out there on the Internet.
So for example here, if I click here, and then I drag,
it actually shows me the various other
companies that are there.
And it completes against a variety of other
sets that are there.
So then I can actually go through here and drag across
the various columns, and drag this down, and see real time
information about a variety of these companies, and bring in
more and more information about the
industry in real time.
The second insight I want to talk to you about is
So Google has certainly not cornered the market on
extensibility or on innovation.
And we want to make it such that our applications can be
extended very easily.
So here's an example for you.
So I'm a movie buff.
And I always keep track of what are some of the new
movies that are coming out.
So where I want to do that though is in my calendar.
Because when I'm scheduling going to a movie, I want to
schedule it right in my calendar where I can see
everything else that's going on.
So one of our developers, Ryan Boyd, actually thought about
the same thing, and quickly put together a feature using a
new platform called Calendar Gadget.
So it's using the same gadget infrastructure we learned
about in iGoogle, but essentially melding that with
Calendar and allowing you to make very interesting mashups
very quickly.
So the way this works now is that if I go to my calendar
here, I can see on particular days when particular movies
are being released.
So I can see there's one coming out tonight.
But I don't want to go there, because of course the Rockies
are playing in the World Series tonight.
But then on Friday, I have an open slot here.
So I'll go in LC right there in line what are some of the
interesting movies that are happening?
And so this one up here actually looks pretty good.
So I'm going to go ahead and add that to my calendar and
schedule a date with my wife.
This is how geeks ask each other out by the way.

So I'll put that on our joint calendar.
So the last thing is community.
So one last thing we found is that these applications work
tremendously well when they're put into a community where
people are sharing and collaborating and interacting.
So when you have a variety of coworkers that you're
interacting with, a colleague's fellow students
for example, we-- and this has actually been the crux behind
our enterprise strategies.
And we've seen a lot of pull from a lot of organizations
that really want essentially Google branded for the
organization, but get that same train of innovation.
So this actually a video from Northwestern University.
Northwestern is one of the first major university
customers of Google Apps.
And they've actually--
They heard a tremendous amount of feedback from their
students, that their students wanted Google Apps.
They wanted Gmail.
They wanted Calendar.
They want to Docs and Spreadsheets much more than
they wanted the internal tools that the
University was offering.
So they created through Google Apps a branded version for
And now all 14,000 of Northwester's undergrad have
access to Google Apps.
And they created this video to trumpet to the world about the
fact that they now have Google for Northwestern.
So the last thing I want to talk to you about, the last
trend, is really that these products are never finished.
And Jonathan mentioned about how you need to iterate, and
iterate, and iterate.
And one of the things is that we have the luxury because of
cloud computing that we can iterate in lockstep with the
user's needs.
And it's almost as if we're having a conversation with the
end user as we're building these applications.
Now if you ask me where are these applications going to be
in three years, I really don't know.
But I'm pretty confident that our users are going to bring
us to the right place.
It's amazing to see where that conversation has taken us in
the last three years.
And I'm very excited to see where it's going to take us in
the next 10 years.
Thank you for your time.
Thank you.
Next we have Paul, who's going to talk about
our developer efforts.
PAUL MCDONALD: Thanks, Jonathan.
So as Jonathan mentioned, my name is Paul McDonald.
I'm a product manager here at Google.
I have been working here about four years, and recently
graduated from the University of California, San Diego, with
a degree in computer science.
Now I want to talk you about our developer program.
But before we get to that, I want to start with a ritual
that we do every single morning on the developer team.
We call it the developer pledge of allegiance.
So if everyone stands up--
come on.
Stand up.
JONATHAN ROSENBERG: If you don't stand
up, he won't continue.
JONATHAN ROSENBERG: It's good for you.
Sergei always has people stretch.
It's good for you.
Right hand on your heart.
Follow after me.
I pledge allegiance to the web open
protocols, packets and markup.
And to innovation for which it stands: one platform under
DOM, highly scriptable, with Ajax and access for all.
So of course, we don't actually
say that every morning.
But really the things in this pledge are what we really
believe in.
We believe and are committed to one platform.
And that platform is the web.
Other companies are building walled gardens in which their
developers work in.
Google is building tools and APIs for the web.
And to make the web a better place for end users to work.
We're committed to open protocols, open standards, and
open source software.
As we innovate in the open, we feel our developers will trust
us more and build more products with Google Tools and
more sites with Google products on them.
And we're doing this for everyone.
So from the copy and paste scripters to the more
sophisticated developers who are using our GData APIs to
extend our products, to the professional developers using
tools like Google Gears to change the way that their
users are interacting with their software.
Now I want to talk about Google Gears for a minute.
Because it's our latest innovation in this area.
I want to show you a company called Zoho, which is a small
company that does online documents and spreadsheets,
and recently integrated Google Gears with their website so
that you can actually access these documents and
spreadsheets offline.
That's what Google Gears does.
It's an open source piece of software that enables third
parties to develop applications that allow their
end users to access the documents online or offline.
So here's Zoho.

Zoho writer product is actually implemented with
Google Gears.
I can create a new document, put some text in here.
You can see they have a go offline link here at the top.
If I click that, I would actually go offline, and can
actually access that document without the use of an Internet

So one thing that Jonathan always says is that everyone
is a developer.
And I truly believe that.
Everyone is a developer.
And Jonathan, today you're going to prove it.
I want you to create a web application for this group in
one or two minutes.
How does that sound?
So Jonathan is a a little scared.
But really, you can do it.
It's pretty easy.
I'll help you along.
JONATHAN ROSENBERG: This this isn't a joke?
So Jonathan, if you click the analyst test tab there.
So this is a HTML page that I've created
in Google Page Creator.
It's just a template that we're going to fill in with
some Google technologies.
So the first thing we're going to do is add a Google Map to
this page so that you guys can figure out to get here.
And we're going to use the Google Maps API to do that.
So if you click on the Google Maps tab, Jonathan.
And type in Google into the search box.
So this finds all Google offices in the United States.
If you click on the first result there, that's our
Mountain View office.
And zoom in a little bit so you can kind of see the
surrounding area.
That's pretty good.
And then if you click on the link to this
page at the top there.
Now that second box there is a way for you to actually embed
that map into another page.
So if you copy that,
PAUL MCDONALD: Control C. Good job.
Back to the analyst page.
And then click in the box where it says find your way to
the Google campus.
And then edit HTML in the bottom right.
And then just past it in there.
So we have a Google Map in this page and also you when we
published it with that actually means the next thing
we're going to add is a gadget that gadgets are the way that
developers actually distribute their content on Google.
Marissa always tells the story of the 12 year old boy who
created the Google Gadget, and the next day had 30 million
page views for that gadget.
There's no other place on the Internet that you can get that
kind of distribution in such a short amount of time.
So Jonathan, back to you.
If you click on the Google Stock Price link there, and
then add gadget at the bottom.
Now there's a Finance tab on the left.
Pretty easy so far.
Click the stock chart.
And then we're going to edit this to show the
Google stock price.
So change the title to Google.
And then the ticker.
Good job.
Now we're going to publish this application so that
anyone can actually access this online.
JONATHAN ROSENBERG: That would be the publish button?
PAUL MCDONALD: The publish button.
You got it.
You're way ahead of me.
And then view on the web.
So once this loads up, you'll see the Google Map and the
Google Stock Price gadget.
You can go now, Jonathan.
Thank you very much.
So you can see how even an old man like Jonathan Rosenberg
can create a web application in just a few minutes using
Google Tools and APIs.
And it's not just Jonathan.
We're going after the long tail of developers.
There's millions of them out there.
And we believe that if we can get a piece of Google on every
single website, we're making the web a better place.
That's it for our development program, Jonathan.
Next we have Rami, You'll survive because you're
bigger than I am.
RAMI BITAR: That was very impressive Jonathan.
So hi.
My name is Rami Bitar.
I'm an associate product manager at YouTube.
I have an undergraduate degree in electrical computer
engineering from Cornell.
l also have a double masters, one in computer engineering
from UC Santa Barbara, and one in management science and
engineering from Stanford University.
I've been at Google since 2006 and recently joined the
YouTube team post acquisition about 8 months ago.
So YouTube is the world's largest video community.

And we want to be a platform that is everywhere you are,
across the web, on your phone, and even in your living room.
And because YouTube is a global property--
we recently launched in Mexico,
Hong Kong, and Taiwan--
we know that YouTube is not just a local phenomenon.
It's an international phenomenon.
So we're very pleased with the overall healthy
growth of the site.
And in addition to acquiring new users, something we
constantly think about at YouTube is how can we continue
keep users engaged on the site?
We know that many of our users come to YouTube to passively
explore content.
So it's our job to surface the right content to you.
So we frame this as a discovery problem.
How can we create a serendipitous discovery
experience on YouTube?
Now what exactly does this mean?
Now all of you have all had a serendipitous discovery
A friend recommends a video for you to watch or
a movie to go to.
Or perhaps you're walking down the streets of San Francisco,
and you see a pair of shoes at a storefront, and you say hey
I'd like the stop in and check out the price or see that
those shoes are like.
Well as it turns out, we have so much watch data that we can
mine at YouTube, we can actually observe the aggregate
user behavior and recommend content to you.
And that's precisely what we do with the
related videos on YouTube.
Let me show you what I mean.

Can we get the right?
So here's the video of an individual named Paul Potts
Now what I want to highlight in this part of the screen is
a video we're recommending of Pavarotti.
Now if you look closely at this video, you'll notice that
the term Pavarotti isn't associated with any of the
meta data on that video.
However, we at YouTube know that there's a strong
association between those two videos.
Now if I play this video for you, what you'll notice is
that the individual Paul Potts is actually performing
rendition of a song that Pavarotti also did.
So it's likely that the users were actually trying to
compare and contrast both performances.
Now something else that's interesting about this video
is how did I actually discover this Paul Potts
video to begin with?
Well it turned out I cheated.
I was actually looking at the query logs, and suddenly Paul
Potts was a top query term in the US and the UK.
Now at the time I didn't realize that Paul Potts is
actually the winner of the Britain's Got Talent contest
which was airing about the US and the UK.
But it did strike an interesting idea.
What if we could try to predict which videos on
YouTube were going to become viral before they actually
became popular.
So we set out to build a predictive model.
And we used a part of the homepage as an experimental
framework to actually test this hypothesis.

So you'll notice these videos on the top of the screen here.
It turns out that our users agree.
We were able to predictably improve the click through rate
on those videos by improving those algorithms. So in
addition to algorithmic approaches to servicing
interesting content on YouTube, social discovery is
also something we think a lot about, the idea being that
you're more likely to be interested in content if you
find it through an individual who has like minded interests.
So a classic example of this is more videos from this user.
So if you see a video uploaded by a user, you'll see what
other videos they have also uploaded.
Another example are playlists.
So playlist is a form to allow users to actually compile
interesting videos in a topic or category, and publish that
for other users to view.
But something else I'd like to highlight is an interesting
product we call YouTube Streams.
So the idea behind Streams was to recreate the TV living room
So what I mean by this is something we're all very
familiar with.
You're at home.
You're in the living room.
And you're socializing while you're watching TV. And so
what I'm going to show you here is something which sort
of tries to mimic that same experience.
So at the top left hand screen here is a video.
And what you'll notice is that on the right hand side over
here is actually a chat room where I can actually message
back and forth with other people also
watching that video.
And what's also interesting is that there's a search box down
here where I can actually search for videos I think
would also be cool, and
contribute that to the community.
So it's a very interesting social experience.
And finally the last form of discovery I'd like to talk
about is search.
Search has always been an important part of how users
find video content on YouTube.
But something interesting I'd like to highlight is that on
YouTube, many users we found are actually using search
similar to how they channel surf.
So the behavior that we observe is that they'll enter
a query on a specific topic, and then they'll start to
paginate very deeply to sort of explore all the different
content on the site.
So in addition to investing in pure video search quality, we
actually try optimized for that passive browse
And of course with all these page views and user
engagements which we are trying to optimize renders
tremendous modernization opportunities, which we'll
talk about in a later panel.
So with that thank you very much.
Thank you Rami.
Next we have Amin who's going to talk about Maps.
Thanks, Jonathan.
So I get to do the last show I guess.
I'm Amin [UNINTELLIGIBLE], and I graduated from Stanford
University's grant program in management science and
engineering last summer, and joined Google and joined the
Geo team prior to which I was a computer engineer.
And when I joined the team, I started looking at the vision
that really started exciting me.
And it was geographically organized all the world
information that we have.
So you could think about what you're trying to do is you're
taking all the information that's available online and
offline, and put it on a map.
This is a new way of browsing and consuming information in
its geographical context.
Now when I started thinking about it like well this is a
big problem solved for a handful of engineers and
product people at Google.
But as we went along, we figured out that it's not
really that tough.
What you need to do is build the base map which is an
accurate physical representation of the world,
put the tools out to the people in the world who want
to annotate that map, and then use all that content and
expose it through search, which we seem to
be doing well at.
So we've made innovations in all of these places.
And I'm going to start by showing you--
can we have the left--

by showing the base map.
So a few years ago we introduced this fluid motion
based on Ajax.
And then we introduced satellite imagery after
acquiring Keyhole.
And we've been making steady progress.
And all of these have now become de facto.
So the next innovation on this page was what we called street
view which is high resolution imagery of vehicles driving
around the streets that give you what a street will look
like when you get there.
So as you cans see.
This is extremely high resolution imagery.
You can zoom around and pan some.
But this really is the beginning of
what we want to do.
Then about six months ago, we launched a tool called My Maps
which allows people to start annotating on top of this in
simple drag and drop manner.
So here's a simple map of the 49 Mile Drive in San Francisco
created by somebody called Peter.
And he's put some useful information that people can
now find which wasn't present in a map.
So this is the most basic form of information
that you can show.

Of course there is static data, there's dynamic data
happening all around the world like for example, I have here
a weather feed from a company called Weather Bug, and from
USGS a feed of live earthquakes around the world.
In fact, there is one in somewhere near central
California yesterday.
So just easy as dynamically looking at a map, you can find
out what's happening around the world at any
given point of time.
So you can expand this to news, events, sports, and
anything that you're interested in.
So this builds on the same concept that apply to the web.
Take the content, put it in a feed.
The next part of this was really about apps, and as
you've heard today, gadgets and APIs, and
exposing all for this.
So we have two products.
One is Google Maps API, which allows you to take this base
map and embed on your website.
And the second products which I'm going to show here is
called Maplets, which is actually geek speak for maps
plus Java applets.
You put them together, it's Maplets.
And what this is a sort of a love child of Google Gadgets
that you've seen all today.
And Google Maps API allows you to create interactive
applications, build functionality on
top of Google Maps.
So what I have here is by far the most
popular Maplet on Google.
And it allows you to simply click and find distance
between two places.
So using this, I'm going to answer a question that has
bothered the greatest thinkers of our time.
Just how many football pitchers can you fit between
New York and San Francisco.
There's your answer.
So when you do get to New York, you're probably going to
want to find some hotels.

And in this point of time, most people go to
or Hotel or Kayak or whichever your favorite
hotel website is.
Instead, what we allow you to do is just
bring that search here.
So you can just turn on your Orbitz Maplet.
And you can do your search that you're looking for.
Or you can turn on your Maplet.
And it will show you the results of that search.
And you can compare prices and do all of the good things that
you can do.
So there are lots of these Maplets being treated just
like gadgets, and inherit all the benefits that you see in
the gadgets platform.
They are extensible.
You can disaggregate your content and your applications
just from your website, and organize things there.
So that's about how you annotate the world.
The third piece, and what I spoke
earlier was about search.
So think about it.
There are 6.6 billion people in this world who are local
experts about some area that they live in, and have
information to share.
And when all that information starts coming on the web, how
are we going to parse it?
How are we going to expose it to users?
That again becomes a search problem.
So about three months ago, we rolled out in Google Map
something called Geo Search, which essentially is an index
of KML, and GeoRSS, and all kinds of geo reference content
on the web, and allows you to find it on
Google Maps and Googlet.
So recently I was in Spain.
And I was in Madrid on Sunday afternoon looking
for stuff to do.
So I typed in a query called just sightseeing in Madrid, as
you can see here.
Now the point note here is not normally a query that you
would go into Google Maps or any other mapping application
and type in.
You'd probably go to or your favorite
search engine to do that.
But what I found here was that I looked up the first result,
and this was exactly what I was looking.
It had been created by a company called
It had all the places it was going to take, and when the
bus was going to leave, and the numbers
that I need to call.
And that's when it struck me that the geoweb that we spoke
about earlier is actually right here and happening, and
it's here to stay.
And just to give you some facts.
There are 50,000 websites that use Google Maps API, and are
built on top of it.
And there are about 4 million maps that were created in the
first three months after we launched the product.
So you can imagine the rate at which this is going to grow.

I'd like to leave you with two thoughts.
One, we know local search is big, right?
When you look at US, there are 25 million
businesses in the US.
And that's about as much content is going to
be there for local.
But when you think about Geosearch, we need about 1.3
billion annotations just to cover the
base map of US alone.
So you can imagine.
There's more content to see.
And there's more content to search for.
So where the opportunity lies.
And of course, with more search is going to bring in
more monetization opportunities.
And if you still don't believe that Geoweb is here, there's
help at hand from Hollywood.

So that's Angelina Jolie with the tattoo of the landlocked
positions of her kids and where they were born.
So there are more ways beyond Google Maps that you can
express your Geo content.
Thank you.

JONATHAN ROSENBERG: Thank you to all the presenters.
I'm almost speechless.
These are the people I'm betting on, and the people who
I invest my time in.
I hope after watching the demos that you just saw there,
all of you guys feel likewise.