NOTE: GrowthBeat -- VentureBeat's provocative new marketing-tech event -- is next week! We've gathered the best and brightest to explore the data, apps, and science of successful marketing. Get the full scoop here, and grab your tickets while they last.
Marissa Mayer, Google’s vice president of search and user experience is on-stage at the Le Web conference in Paris. I’m live-blogging as we go:
Michael Arrington asks about the search announcements earlier this week (namely Google Goggles, local search and real-time search).
Mayer: We think of search as having four main components. Modalities — how do you search? Most people now type in search queries, but we think people will talk to their phone and they’ll take photos. The other piece is media — as the web has gotten even more and more rich, we’ve gone from text to a very rich web with video, blogs, books, product search and now real-time. Bringing all that together really enriches the result section. Then the other two components have to do with language, translations and personalization. We think the searches in the future will be better because they understand you, who your friends are and where you are.
Arrington asks about Google Goggles (the image search the company launched earlier this week).
Mayer: With Google Goggles, they’re looking at image recognition. Sometimes location doesn’t help. The example that Vic used earlier this week was with a wine label and it picked it up and gave tips on its flavors. As image recognition gets further along, so will the type of indexing you’re talking about.
It seems like voice-to-text, looking at a audio stream and turning it into words is farther along. There’s good progress being made both in academia and at Google. Google Goggles is the first example of something that’s available for consumers.
Arrington asks about the rumored phone Google is launching.
Mayer: I don’t comment on speculation.
Arrington asks about the volume of mobile searches.
Mayer doesn’t reveal exact figures but says its growing.
Arrington asks about Chrome and Chrome OS.
Mayer: There are tens of millions of Chrome users and we’re excited about its growth.
Arrington: We all understand the dire situation of the print media in general. What I want to talk about is how do we fix this. Eric Schmidt talked about this in the WSJ last week. Can you talk about how you see the future of news?
Mayer: We’ve thought a lot about this. We really think about engagement — how do you increase the engagement of users online with news? If we invented news as a delivery vehicle — how does it get delivered on the web. If we reinvented it from scratch, it would look very different from what we’ve had. We released a prototype yesterday with the New York Times and the Washington Post called Living Stories. For a long time, news has been the print article on the web page. But what if that story was alive? You could come back to it, add to it, and get alerts on it? I basically think whenever a media changes over to a new delivery vehicle, it puts pressure on the atomic unit of consumption. It happened with iTunes with the album moving to the song. It happened with YouTube with long-form standards of video to short-form. Now it’s happening with news. People can come in and read one story from the source and then move on. That’s the atomic unit.
Certainly, some readers will follow brands more attentively. If you follow that line of thinking, a lot of things need to change. If you look at Wikipedia, it will beat a lot of news articles on search placement. They’ll have one page. And then newspapers will have hundreds of articles that are competing with each other. If you have one URL, it can really rise up. If you look at engagement and a lot of factors — 1) Personalization and 2) What to do next.
In a newspaper you have a lot of columns and your eyes jump around. If you look at the bottom of an article, where the most engaged reader will reach, there’s nothing there — no ads, nothing to do. And of course, the reader will go somewhere else to be entertained.
Twitter and Facebook have popularized this notion of the stream. So why can’t I have a portable stream of news that I can take with me everywhere? Why can’t it take into account my preferences, my location, the veggies so to speak of news that I should know? I call this the hyper-personalized news stream. What form it takes is unclear.
Arrington: Do you expect the New York Times to understand what you’re saying? Are they going to get this and move fast enough to stay alive?
Mayer: I think so. They’ve been very progressive in their thinking. They want to know how to reinvent themselves in this medium.
Arrington: What about Murdoch?
Mayer: We do partner with Murdoch. As part of our real-time search announcement, we partnered with MySpace. They’re giving us their public updates.
Arrington: Do you think they’ll actually pull their content out of Google?
Mayer: I hope not. We want to be comprehensive. But we have to respect the copyright owners. That’s what Google is based on. Without content, we don’t have the search engine. I hope for the sake of comprehensiveness and the quality of the search engine, that won’t happen.
Arrington: Would you pay to index content exclusively?
Mayer: That’s hard to say. But we’ve developed a lot of products that help with the monetization problem. Google AdSense paid about $5 billion out to publishers. We also have an Ad Exchange and the display advertising network. We’re very excited about those things. Search and news deliver itself literally billions of clicks to these sites each month.
Arrington asks about music.
Mayer: We did this very interesting computation over all our searches and lyrics was actually our number two most searched term of all time. Music was number 9. We did see a big user need. When someone has something like iTunes or Rhapsody open, the most common concurrent activity is searching for lyrics on Google.
Arrington: I wrote a series of posts about scammy ads appearing on Facebook. Have you ever played Farmville?
Mayer: I haven’t.
Arrington: There’s like 60 or 70 million people on Farmville. But there were really scammy ads and people being ripped off. But all of those companies advertise on Google and I haven’t looked at it too much. What are you doing about it?
Mayer: I can’t comment on it a lot. But we did launch a lawsuit on the scammy ads that use our brand and logo.
Arrington: Because it’s litigation you can’t talk about it. [Asks about Twitter]. I click on a lot of links people share on Twitter and nine times out of 10 that’s really interesting. It’s about discovery. How are you going to evolve your search engine to accommodate that?
Mayer: We had two announcements. One was social search — where you can search something like “New Zealand” and you get the 10 best links about the country but also photo logs and posts written by your friends about that. We definitely think using your social circle is really useful. Search can help social networks that way. People ask questions like, “Is this movie good?” “Where should I go for dinner?” And instead of shouting that out to the world, search can be a part of that.
The second part of that is real-time search, which we launched earlier this week. I was using Twitter search to find out about snow conditions.
Arrington: These aren’t even necessarily your friends.
Mayer: Yes, because someone on the hill is telling the truth. They’re there. They’re taking the time to tweet. If you combine that notion with search — you’d want to see the public updates and the private updates from your friends. The perfect search engine would credential you and then show you updates with privacy considerations taken into account.
While one of your friends give a status update on a particular topic, it might not be useful to you now. But three months later, when you’re going to that movie or to that place, it will be.
Arrington: The biggest problems around real-time search have to do with authority and authenticity and filtering. Can you solve these problems?
Mayer: We’ll have a lot of fun with this. The backbone of Google has long been PageRank. With real-time search, you don’t have that. We’re looking at 12 factors at things like query volume and other things like — does a person get retweeted a lot? Do they get replied to a lot?
I’m an expert at Google, but not on Paris, for example. So that notion of authoritativeness should be weighted into search. We want to avoid something false getting posted and being propagated very quickly.
Arrington: I remember terrorist attacks and CNN was pulling Twitter. There was so much retweeting and it was hard to figure out what was going on and who was really there.
Mayer: That’s true, but when you find them it’s magic. The Bay Bridge had a cable snap. We actually had the Twitter firehose at that time and we had a user post a twitpic. And they beat the news source by 18 minutes. And the first news source was the San Francisco Chronicle, and they were asking about something happening at the Bay Bridge.
Arrington asks about the Wave.
Mayer: Well we’re launching early and iterating fast. The important thing is critical mass. It’s useful when there are more people on it.
Arrington: Are you using it internally?
Mayer: It varies from team to team. We have some teams, like the one that did the new fade home page, they’ve used it almost exclusively.
Arrington: It seems to be with the Google Phone. If it didn’t exist, you’d just say it.
Photo from Blogowski Consulting
We're studying digital marketing compensation: how much companies pay CMOs, CDOs, VPs of marketing, and more
, with ChiefDigitalOfficer. Help us out by filling out the survey
, and we'll share the results with you.