Keynote: Generation Next: Search In The Coming Decade
Lisa’s still sleeping away, lucky girl so it’s just me and Tamar in blogger central today. Breakfast had both bagels and juice this morning, so call it a win.
Our topic this morning is the same as the last part of yesterday’s keynote. What’s the future of search? Where are we going? Will there be flying cars? These questions and more will be answered by representatives from the four major engines: Brad Goldberg (Microsoft), Dr. Larry Heck (Yahoo!) and Peter Norvig (Google). Chris Sherman (Search Engine Land) and Gord Hotchkiss (Enquiro) will be moderating to prevent an all out brawl.
Gord introduces and gives a short bio on each of the speakers. Chris takes things back for the first question:
Will search as we know it survive?
Brad: Tells a story about how last night the cab driver was telling him about how bad search is. You click and go back, click and go back. Users are getting more sophisiticated. Search will survive, it’ll evolve to become more and more useful.
Peter: I agree with most of that. I think in some ways search will disappear as it becomes more and more ingrained in what you’re doing. It’ll be integrated more into what you think of as a core life.
Larry: What’s going on in the background is a foundation that we’re not going to throw away, we’re going to be building on that. What’s going to change more is going to be from the user’s experience. The way people are accessing the information. It’ll become pervasive. Right now it’s just PC but it’ll be going mobile and further and each will require
How do we interact with this huge score of information?
Larry: The way we access information is going to change based on the interface and that’s going to change things down deep in the stack. Right now it’s based on expanded queries but it’s not going to be that way. [He's really hard to hear. Someone fix his mike.]
Peter: Your average query will be longer on PCs. ON the other end it’ll be hard to get the results back because we all take advantage of having space for ten results. If you have a tiny little cellphone with a one inch display, having your result at 4 or 5 isn’t a good experience.
What about voice activated queries? Changes in Search marketing?
Brad: The biggest changes that we’ll see won’t be in the algo or results. It’ll be in the way that advertisers are taking advantage of it. Right now, mobile search is a fairly narrow reach.
Peter: We as an industry need to get more sophisticated.
Have we been using relevancy in place of usefulness? Are we going to see the artificial division between listings go away?
Brad: I think we need to keep the division between paid and organic. People want to know where things are coming from in the algorithm.
Larry: I agree with that. As you go through a ssession it becomes clear to the search engine that you’re going to buy and it makes sense to move the results in a commercial way but you still need to keep the separation between organic and sponsored.
Back when Vista was still Longhorn, you guys talked about implicit vs explicit queries but that sort of went away. Will that come back?
Brad: We’ll see that become more out of the operating system and more in the context of the application. We’ll start to see search playing more and more of a role in the foreground and in the background.
Do you see search evolving into a research assistant?
Brad: I’m not bullish. I think it’s an interesting concept but it’s technically an incredibly hard challenge. Not in the next 10 years.
Peter: We’re all very good at the 3 second response. Then there’s the 3 minute, sorting through possibilities, we’re pretty good at that. But three months or three years, we’re not good at supporting you there. The whole point of search is about what you don’t know. In that way, personalization is a bad thing.
Larry: Search is about making the user more efficient. Having search go off and do it for you isn’t as good. The reach and memory is better for computers but humans do better putting it together. The user is going to be intertwined. I’m not so sure that you have to personalize to one person. Research done suggests that community is better in that single session. In that moment over the series of searches you’ve made, you’re trying to complete a task. It’s probably better to socialize it in that moment.
How smart can search get in the next five years?
Peter: I think we’re just getting started. We’ll get a lot better at understanding the content and I think we’ll start getting better content. Books and magazines are just starting to be searchable.
Brad: I wonder about how much the first SERP after you do a query will dramatically evolve. That’s really where it starts. I think a lot of changes will come back to the users. The number of users who click on the tabs on the front page is so small even though that’s so simple.
Larry: More a question of how much more effective are people going to be rather than how much more effective will search be. How much more effective will the interaction be.
Are we expecting too much from humans in using their options?
Brad: Most search is horizontal but most results are verticals. I think we need to get over into verticals better.
Local search has been held out as a promising vertical but it’s still horrible. How close are we to good local search?
Peter: We’ve got to tknow about it first. There’s a little bit of yellow page information.
Larry: not a fundamental shift in technology required. It’s having the mom and pop shop understand that there’s a utility to this information.
Peter: One of you (audience) needs to come up with a business model to get local to sign up (via high school student door to door.)
Do ads need to get more engaging or louder on the page in order to be noticed as the interface gets richer?
Peter: I think the whole web is getting louder. You need to change with the times and fit in.
Larry: You need to provide the most value to the user. I don’t think the search engines are going to need to push very hard. But they need to keep up in terms of creatives and quality.
Brad: Searhc is successful because it’s quantifiable. I think we’ll see a lot of experimentation because you can see what’s working. We’ll see a lot of diversity in ad types. All the ingredients are there for someone to experiement quite freely.
Are we looking at a Darwinian approach to ads? Just seeing what sticks?
Brad: I think that’s fair.
If you improve search to the point where it satisfies users needs, it makes ads irrelevant. How do you balance that?
Larry: There was a good session the other day that pointed out that it’s hard to spam paid search.
Peter: as algorithmic results get better, the bar will get higher but that’s not that hard.
Brad: I think there’s still a lot more than can happen based on user intent.
I understand why personalization is hard, how good does personalization have to be in order to
Peter: there are a lot of parts to it. One of the things is that persists over time but most searches are new. It’s not like amazon where if you like jazz one week you’re not going to suddenly switch over to hip hop the next week.
Larry: We don’t need to use the signal so strongly. If we can identify what the person is trying to accomplish, we can make it feel personalized by pulling from the user experience.
Peter: Like horoscopes.
It seems that search marketers have less and less influence where their content is showing up. Is this a threat to traditional SEO?
Peter: I think it’s more arenas for them to show up.
Brad: You’re seeing a lot more companies using video. I think people are going beyond SEO and looking at other venues. It’s about who can exploit it in the right way so that it doesn’t feel like atmosphere. It’s some SEO and some how marketers think about what things are available to them. I think in 5 years we’ll look back and think it’s quite crude. There’s a lot of surface areas being created and a lot of opportunity. It’s quite difficult and a little Darwinian.
Larry: As the SERPS become richers, it’s going to move back to traditional internet marketing instead of the very spare search marketing so far. More like a homepage.
Gord: It’s opening up opportunities that search marketers haven’t been playing in so you need start thinking about how to create a relationship with a customer like you haven’t before.
Are you creating tools to help search marketers to reach international/non-english speaking audiences?
Peter: we’re focusing more on the algorithmic side, the translation.
Brad: you need to think about the scenario in which someone searches. Mostly it’s research oriented. So you need to think about what sort of ad you place on that kind of content.
Larry: I agree with Brad it’s a marriage between the incentives (like machine translation) and the search marketers.
Gord: What research area is particularly interesting when you look at this interface between humans and the Internet>
Peter: The mobile space and getting away from being chained to the desktop. (Gord: You said that the browser platform is kind of a roadblock.) Peter: I think we’ve got a ways to go in terms of opening up.
Larry: Openness in terms of leveraging, and opening up the platform in order to best use the content. What we really aught to be doing is opening up in terms of utilizing that expertise. There’s an imbalance between the author of the document and the understanding of the intent. I think on the user side we’re going to be facilitating what task they want to complete and what their intent is.
Brad: In the car what will happen, when you’ve got all the single purpose devices that are very good, there will be a convergence. Right now that’s an untapped area.
Chris: Five years from today what one feature would you kill to have working?
Brad: You’re asking the ones that are developing it. We just want to ship it. I would say more and more enabling people to search in a verticalized way. Whether that happens on the homepage or on the SERP.
Peter: I want to really focus on the connection between the user and the content.
Larry: I want access and functionality wherever I’m going. From the computer to the car. I want to be able to access that tools and technology no matter where we are.
Gord: How do you see that kind of ambition going up against the music and entertainment industry that don’t awant to lose control.
Brad: I think search engines are going to have to be very clever about how they give you content. Snippets that lead to the original content, on their own site and with their own player.
Peter: It’s an opportunity. Rather than publish a book and try to get it on the top ten, you’ll be able to reach smaller targeted audiences. I think publishers will get that.
Gord: I don’t think they will. I’m cynical about that.
Chris reminds us that there’s coffee in the exhibit hall.