Tip:
Highlight text to annotate it
X
>> PATRICK: So, today, we have Shivani York and Srivaths Lakshimi from TIME.com, and they
did a great demo of the usual list on TIME.com. So, Shivani, can you tell us a little bit
about yourself and...? >> YORK: Sure. So, I’m the director of Web
Development and Products Strategy at TIME.com, and we worked over the last month or so with
Patrick in Google to implement the new Friend Connects and new APIs and Cloud Computing
for Time list. >> PATRICK: Excellent. Can you tell us a little
bit more about Time list, what it is? >> YORK: So the Time list concept is we do
a lot of our daily list. We do a lot of specials that are daily by nature, weekly by nature,
and quarterly by nature. For our daily list, which is the one that we're going to start
off with, they take the form of top 10. And they can be very–they're sort of like a–some
of them are very fun and personal by nature, their different perspective of the current
day's news. So, I'll give you an example. Some of the recent ones have been top 10 iPhone
Apps for moms, or top 10 movie catchphrases. So when you think of things–list like that,
they're–they lend themselves very well to be personalized, to be ranked by users. So,
our edit team creates those top 10 lists on a daily basis, and we also do other list as
well, but we're going to start with top 10. So the concept here was, the challenge that
we threw at Google was how do we make this list more fun, more interactive and more social
by nature? There are a couple of catches to it, too, as well. We didn't want to create
a big application load at the bottom. We don't want to create a separate community or separate
social network. We wanted to be able to leverage social networks that were already out there.
And we wanted it to be fun, easy to use, interactive and easy to build, which, you know, being
a small development team we wanted to make sure that this wouldn't be a big overhead
and a ten-month project. So that's where you guys came in.
>> PATRICK: Yes. So, tell me about the user experience when they're experiencing the social
list. What does it look like? >> YORK: So, the experience that you’ll
see which will be live on TIME.com in the next few weeks is you will get our editorial
list, which is the full list of sort of it’s top 10 iPhone Apps. You will get the ten iPhone
Apps that are selected by editorial team. And then users will get the ability to rank
the list and make it their own. So you'll get an interface that'll allow you to drag
and drop that list into your own rankable form. You can take three of the top 10 or
you can take all 10 of the top 10. Then you can sign in with any of the open social networks.
You can sign in with Yahoo, Google, Facebook, Open AZ. And then once you sign in, you can
then save and share that list to your Twitter, you can share on your Facebook, you can email
it to friends. And the exciting thing is that once you share that list, that URL is yours
and it's unique to use. So you can come back as a user and look at your list, delete it,
create new ones. So it just becomes–it takes that one list and gives the users the capability
to create hundreds of list that they can make their own.
>> PATRICK: Excellent. What are you expecting from these applications? Do you have any expectations
or is this an experiment to see how it goes? >> YORK: It's something that we've been brainstorming
about for the last few months at least. I mean, it has both advertising implications
as well as user implications. What we're hoping we'd get out of it in the short term is giving
you just fun things to do with our list. I mean, our lists are a huge engaging area on
our site. They're very sort of click through by nature so people come to it their short
form. They usually include a video or a photo on top, and there are maybe two or three paragraphs
on each page. We can easily click through. And like I said, the subject matter for some
of those lists is so interesting. It sort of reminds you of something. So, you know,
top 10 iPhone Apps. I mean, everybody who uses an iPhone App has an iPhone App that
they like. So why not give them the ability to create their own version of it. So what
we're hoping to get out of it in the short term is to build user engagement to make the
list more sticky and to take advantage of the viral nature of Facebook and Twitter and
give people sort of the ability to customize the list and make it their own. In longer
term, I think it has implications both in the advertising side. So I think this is something
that will be a very exciting products--product for advertisers as well. I mean, you know,
imagine if we created an editorial list and we also had an advertising-driven list that
was related right next to it and you can take both and share both and create fun, you know,
concepts out of both of those. And then we also do larger list. We, you know, we do quarterly,
we do best of everything, top 100, Time 100, which is a list of a hundred most influential
people. Imagine taking that and building your own list of my ten most influential people
that have influenced me personally. So–then the time edit list becomes very viral and
very personal and very interesting to each user in a different way. So we're hoping that
it'll drive engagement, build additional page views, and we got to take advantage of the
viral nature. >> PATRICK: That's great. Thanks very much,
Shivani. Srivaths, do you want to talk about the technical aspect, how was this implemented,
how long did it take and what the application is?
>> LAKSHMI: Yeah. What else happening here is all of them appears on cash between–basically
it's a static page and we don’t want to put anymore load by integrating the Facebook
connect and Google connect or Friends connect, whatever. And we are using GQuery APIs from
Google, which really helped us integrating with the Google Friends Connect here. And
most of the logic is in JavaScript. So, the JavaScript is [INDISTINCT] in our server so
the amount of reloading the JavaScript again and again is [INDISTINCT] so it's not going
to put any of our servers or anything. And we don’t want to make a separate database.
What we use are log n authentication information. So one good thing with integrating with Google
Friend Connect was, like, we use App Engine from Google tools. And it’s very good and
we can have all the logs in the App Engine, we know the user information like who logged
in, what's the network, what list they're trying to create, how many discs left, whether
they're just one or the 10 or something in between one and 10. And we are really proud
to be the partner of Google's Friend Connect. >> PATRICK: Okay. We're very happy to have
you as a great use case of this kind of new architecture. So to summarize, you're going
to study HTML page. >> LAKSHMI: Uh-huh.
>> PATRICK: And then you had GQuery uploads that is used to do all the user interface.
And then you have Google Friend Connect for the authentication and security, and then
you use the security token, is that it, that brings connection to you?
>> LAKSHMI: Yeah, the security token with the user ID and the network ID has been sent
to the App Engine on the solar side on the App Engine [INDISTINCT].
>> PATRICK: I see. So then... >> LAKSHMI: Whether it's a valid user or not
so that once the user clears the list, the App Engine gives us a new user information
you need. It's going to be some numeric value, so it's going to be unique for each user.
And the good thing is, tomorrow you can assure that, okay, Patrick created this list and
you shut it with [INDISTINCT] and say, "See, my list is in TIME.com." It's going to be
a good sign for everyone. >> PATRICK: Yes, so you generate these unique
URLs and then all the dynamic comes in to support from App Engine...
>> LAKSHMI: Yes. >> PATRICK: Dynamically on the blind side.
>> LAKSHMI: That's right. >> PATRICK: Yeah, that's a new type of architecture
where you have your data repository somewhere else and you are doing these [INDISTINCT]
peoples. And thanks to the security token, you can do a restful under severed sites to
make sure that user is... >> LAKSHMI: And one good thing is all the
data is going to be installed in App Engine at some point of time, you may want to do
some internal community or whatever inside TIME.com. We can just export everything from
App Engine so we have all the user information and we just dump in with [INDISTINCT].
>> YORK: Right. If you remember, one of the–sort of the challenges that are thrown to you guys
was that this was easy to build. So we didn’t want to have to build a database and then
build their architecture. So it was having Google’s App Engine there was a huge help
to us. So the data is there for us and available for us to use when we have our own community.
But for now, we didn’t have to create a new database, new servers. So, it was just
very quick to get it up and running because of that.
>> PATRICK: Great. >> LAKSHMI: Yeah, we also integrated with
Facebook to come out, and one little thing we missed with that is we don’t squat the
data and we don’t know who come with the–what is happening because there was one disc with
this Google Friend Connect as it's stored in App Engine, we have all the data in there.
>> YORK: And another feature that I forgot to mention earlier was--so as--not just people
being able to share their own list, another thing that you guys implemented that was terrific
was being able to see a user view of our list. So there's the editorial-driven list of top
10 movie catchphrases and then there's a version that people can come to and see that this
is how everybody in the community in TIME.com and Google and Yahoo and Facebook had ranked
it. And this is what they think is the top 10 movie catchphrases or top 10 iPhone Apps.
And even if you don’t play around with the list and even if you're not that highly engaged
users that creates their own list, it's just interesting to come back to see the two different
versions of the list. And I think that’s where I was sort of we'll catch a lot of users
that are skeptical about trying their own version of it.
>> LAKSHMI: And the other good thing is as we have implemented in Google Friend Connect
is site disc. It's not like a list disc or anything. Once you're signed in the future
in TIME.com just like if you need something or anything for this list. You start with
some comments, whatever, it's going to be an easy step. Maybe a couple of pause to a
day to integrate the product. >> PATRICK: Yeah, because you own the data.
>> YORK: Yeah. I mean, I was very impressed by the time that it took you guys to build
it and get it up and running. I mean, I seriously didn't think that you guys will be able to
do this by the conference. So... >> LAKSHMI: The conference was done in less
than three weeks. >> YORK: It took three weeks. That included
use cases, designs, building, building the database, you know, doing all the backend
work. So it's–the time to get it up and running, the speed of it is quite amazing.
>> PATRICK: It feels really comforting. >> YORK: Staying up till 3 am doing it, so...
>> PATRICK: Yeah, as usual. >> LAKSHMI: It wasn't though, see? That is
good response work here and many people liked it. So we hope we will launch the VL version
some time [INDISTINCT] so that we give users the ability to make their own [INDISTINCT].
>> PATRICK: That's great, and I look forward to see live [INDISTINCT].
>> YORK: Yeah, us, too, are very excited about it.
>> PATRICK: I know. So, tell me about the statistics. So how do you see the–how the
users have rated this list? >> YORK: So you have to do some work to see
how other people have rated the list. So the user interaction right now or when we launched
will be–you have to create your own list in order to see how the rest of the community
has ranked the list, which we think is a better user experience because you sort of get something
at the end of creating a list. And then you also get a histogram which you implemented
recently. So you get a version of how users have ranked this list and they also get little
charts, which are very interesting to see on how each item has been ranked by the user.
So it's just another view of how that list was rated by other people.
>> LAKSHMI: Yeah, it's going to be [INDISTINCT] sort of, like, so, you know, [INDISTINCT]
said that it's something on the flash point–flash disc made by editors and maybe the users didn't
like it. It's going to be–many people would have rated that it’s number ten, or something
like that. It's going to be a good and clean graphical representation of how the other
one was rated. >> YORK: Right. And as I mentioned, you know,
this lists are highly engaging for our users. I mean, you know, we get a tremendous page
view numbers on this. So we're just hoping that by launching this, we increase that engagement
on our site. >> PATRICK: Okay, we're excited to see how
it will go. Actually, it would be really interesting to get some numbers and you can share some
like the moms after launching that or... >> YORK: Definitely.
>> PATRICK: [INDISTINCT]something great of trying to see how more engaging these lists
are. >> YORK: Yeah. And modify it if something
isn't working. >> PATRICK: Yeah. Sure.
>> YORK: Yeah. This is a new [INDISTINCT] released by Google Friend Connect. It was
released two months ago, so. >> PATRICK: Yeah.
>> LAKSHMI: So I think that is somewhat involved in both TIME.com site and the Google site
to make it more optimized. It was the test what happens.
>> PATRICK: Oh, yeah, the application is quite cutting edge. So here, you're using the Google
Friend Connect JavaScript TPI... >> LAKSHMI: Yes.
>> PATRICK: That's only two months old. >> LAKSHMI: And [INDISTINCT].
>> PATRICK: [INDISTINCT], which is also two months old.
>> LAKSHMI: Yeah. >> PATRICK: There's Google App Engine...
>> LAKSHMI: And I believe this is the fast site using both of those.
>> PATRICK: Yeah. Yeah, I think you are the first big site who's going live with this,
except maybe for Washington Post. They did a demo as well. I think they're using–I
don’t know if they're using both. I think they're using the [INDISTINCT]. I need to
double check with them. But, essentially, you–the Washington Post and TIME.com will–are
the first ones to go right with this. >> YORK: That's great.
>> PATRICK: And then you're using App Engine and you're using the [INDISTINCT] for...
>> YORK: Right... >> PATRICK: It's full of Google products.
That was–that's what made this list. >> YORK: Right. And I think, you know, once
we get to the larger list, at the 100 list, the UI will be very interesting because we're
going to have to make some changes, too, to let people be able to easily browse through
the 100 and create their own version of each list. So we'll come across that bridge over
the next few months. But we're excited about launching it for the daily top 10 and then
expanding it out to the rest of our specials. >> PATRICK: Okay. Great. I look forward to
see that live. >> YORK: Great. Thanks, Patrick, and Srivaths
for all your help. And we also have a third person which is actually missing from the
interview, Henry Chan, who tremendously helped and he did all the front-end work.
>> PATRICK: Henry should have been there. He spent lots of nights coding that...
>> YORK: Yeah, he's maybe getting–catching up on sleep at this point. So, I can't blame
him. >> PATRICK: I'm pretty sure he needs it. Okay,
thanks very much. >> YORK: Thanks, Patrick.
>> LAKSHMI: Thank you.