With this post, I return to the topic I got sidetracked from a few months back: where is search going? I’ve had the chance to talk to Stefan Weitz from Microsoft, The Search author John Battelle and, this time around, I had the chance to pose a few questions to Johanna Wright, Google’s Product Management Director for Web Search (by email). If we’re exploring where search is going, at some point we have to poke the 800 pound gorilla and get its thoughts. While I didn’t have the chance to follow up on any of the thoughts from Johanna (as you’ll see, we didn’t agree on all points), I thought it most appropriate to post my questions and Johanna’s responses in their entirety.
comScore has indicated that more and more search is happening outside the typical search portals like Google. In fact, Google itself reminds us that YouTube is the number two search engine. Most of these searches are launched as part of a larger overlying goal, ie getting information and doing something with it. Can Google keep delivering a useful search experience from a one-size-fits-all portal? Will search functionality splinter over multiple destinations and applications?
Wright: In general, we think it’s a good thing that there are many different search options out there. This gives users more choices for how they want to search and discover information. However, on the other hand users also want a more unified and seamless search experience. Users don’t think in terms of “portals” and “properties,” they want information (or “answers” and “to do something with it” as you say). Over time, integrating different content types into the main Google experience has been a clear win for users. We did this a couple years back with universal search, where users can start finding images, news, books and other content right in their main search results. With our more recent results page refresh, we’ve made the interface more consistent across different types of information, so when the user clicks “images” she still knows exactly where to find refinement tools, the search box, and other interface elements she’s used to. We don’t anticipate any kind of splinter of the core search experience on Google.
Mobile search is typically launched from a specific destination. Mobile devices can augment our query with images and sound. Smarter applications can even layer in the context of our activities at the time. How do you see this changing the search experience?
Wright: Mobile is a key part of our strategy and opens up many possibilities for the future. With smartphones getting better and better, people are doing more and more searches. Phones also enable new kinds of search input. For example, today users can search by voice by speaking a query into their phones. They can also search by sight by snapping a picture in Google Goggles. They can even search by location by simply clicking “Near me now” on their mobile Google homepage. The phone is a uniquely intimate device, and this invites further personalization and other exciting possibilities for the future.
Can Google’s algorithm still rely on relevancy as a measure of search success? Are user’s expectations being reset to expect usefulness rather than just relevancy determined by semantic and linking analysis? And how is Google’s algorithm changing to meet not just changing user expectations, but the addition of other signals, such as timeliness and social network activity?
Wright: There are number of different questions there. Whether we talk about “relevance” or “usefulness” is a very subtle difference. Isn’t more “relevant” information also, by definition, more useful? I would also note that relevance is by no means an easy or solved problem. The ordering and presentation of the top web results is a critical part of the search experience, and one of the key reasons users keep choosing to search on Google. This is why today we use more than 200 signals to rank webpages, and in 2009 alone we made more than 550 improvements to the core search experience.
One related question might be, “Is Google moving beyond creating lists of webpages?”—and the clear answer is “yes”—we’ve been doing this for years. Let me give you a couple examples. Last year we introduced something called “rich snippets” which enables webmasters to mark up their sites so that rich relevant (and useful) information shows up in their result text on Google. A year later, and users are finding things like event dates and locations right in their search results—information that someone could “use” to go to an event. With more recent innovations like our updated answers feature, we’ve moving beyond ranking webpages to actually extract and understand structured information from across the web. Do a search today for [barack obama birthday] and you don’t just get a website, you get the answer, sourced from across the web.
To your questions about “timeliness” and “social network activity” and ranking signals, we don’t generally get into the details of our core algorithms. One notable improvement is Google real-time search, a comprehensive and integrated search experience that includes content from across the web. Another is Google Social Search, which surfaces relevant content from your friends and broader social network right in your search results.
Recently, you’ve added a third column to Google search results, following a more Bing-like format (and Ask’s 3D format before that). Why the change? And why do you offer fewer options refinement options than the competition?
Wright: Search is far from a solved problem. We are constantly working to make search better, and we made our latest enhancements to the results page to make it easier to access search tools that help people find what they need. We all search differently today because the universe of online information has grown tremendously. As information changes, so do the tools we use to find it and we will continue to make enhancements to the results page.
Google has been experimenting with a permanent left-hand navigation panel as early as 2006 (take a look at this blog post). We launched our left-hand panel “search options” in May 2009 (check out this blog post ). After studying how our users were interacting with search options and developing new technologies to show only the most relevant search tools, we decided it was time to open the left-hand panel for all users.
I’m not sure I agree with the premise of your last question. You’ll note that the refinements Google offers vary dynamically based on the search query and content type you’ve selected. While I haven’t counted the tools, the key innovation here is not the sheer quantity, but the relevance of the tools we display to users. In addition you’ll find a variety of query refinements in Google Suggest, the new “Not Entirely Unlike” section of the left-hand panel, and at the bottom under “Related searches.”
So what’s going on here is that we’ve developed a broad set of refinement tools, but we only show you the most relevant tool for your specific query. It might be that when you’re looking for [how to tie a bow tie] we’ll prompt the video refinement tool, but not when you’ve entered the query [red badge of courage] which prompts a books refinement.
If search does splinter over multiple destinations, many embedded within online and mobile apps, that changes the whole nature of search. It is no longer a destination, but rather a utility. That also changes Google’s potential revenue model. If you no longer control the real estate that people see the results on, you have to start looking at sharing revenue with partners. How can Google defend against this shift in the marketplace?
(Wright chose not to respond to this question, as she didn’t agree with the basic premise.)
Increasingly, Google has turned its attention to offering different platforms (Google TV and Android) that might support our consumption of digital content. How do these new interests integrate with Google’s search development strategy?
Wright: Our core focus at Google remains on search. We still have more engineers working on search than any other product. Other initiatives, such as Google TV and Android, help move search beyond the desktop and laptop paradigm on to other hardware (television and phones). We’re excited about the potential in these areas. If you think about something like television, years ago when there were only a few channels, there wasn’t much of a search problem to solve. Today, with hundreds of channels, plus hundreds of shows on your DVR, plus endless content on the web, people are now confronted with serious search problems on their TVs, and we’re looking forward to applying our search skills to that space.
Well, it wasn’t the far ranging and blue sky interview I had with Weitz or Battelle, but I knew that was a long shot when I reached out to Google’s PR department. Obviously, when we peek beyond the immediate horizon when it comes to Google’s search plans, they’ll be playing their cards close to the chest.
Opinions expressed in the article are those of the guest author and not necessarily Search Engine Land.