Fireside chat with Marissa Mayer, VP, Google and Michael Arrington, Editor, TechCrunch

TL;DR

Contextual discovery aims to deliver useful information based on a person’s location, activity, and broader context without requiring a typed or spoken search. Google was exploring browser panels, mobile recommendations, social restaurant menus, airport guides, and location-sharing features while confronting privacy, interface design, network growth, and product integration challenges.

Transcript

so this is the third year we've done this Y and uh and as L said the first year I think that evening you you got engaged I did I was very surprised but yes um and Zach is here I saw him earlier today and then uh last year you didn't get married at Le web but you got married at some point and then two or three days later yeah and you're still marrie... Read More

Key Insights

• Contextual discovery is an approach that uses location and user context to identify useful information without requiring the user to type or speak a search query. Its purpose is to anticipate the next relevant piece of information and present it proactively.
• Google’s local and geographic organization combines product management with engineering. Its responsibilities include local search, Maps, Earth, Street View, Latitude, desktop products, mobile products, and experiments that connect geographic context with information discovery.
• The central interface challenge is deciding how proactive information should appear without disrupting the user. Mayer suggests that browser-based recommendations could use a panel positioned beside or below a webpage to provide related entities and complementary details.
• Mobile contextual discovery works by considering physical location, recent activity, and information users permit Google products to access. A restaurant could trigger a socially annotated menu, while an airport could trigger a guide relevant to that location.
• Latitude is designed around persistent location sharing, which makes it suitable for a smaller circle of trusted contacts. Mayer distinguishes this model from check-ins, which provide an explicit and potentially more comfortable way for users to disclose their location.
• Location products become more useful as participation grows because their value depends partly on network size. Mayer indicates that adding explicit check-ins to products such as Latitude or Google Maps for mobile could broaden participation and improve user comfort.
• Large acquisitions require customized strategies because integration depends on the acquired company, its leadership, its brand, and its relationship to Google’s products. Some acquired businesses may remain autonomous rather than being fully absorbed into existing operations.
• Hotpot is described as a personalized recommendation engine integrated with Google’s mobile mapping and local search experiences. It allows people to rate restaurants and supports recommendations within Google’s broader collection of local and place-based products.

Questions & Answers

Q: What is contextual discovery in Google’s product vision?

Contextual discovery is the idea of providing useful information without waiting for a person to submit a typed or spoken query. The system considers signals such as location, browsing patterns, recent activity, and permitted usage of Google products. It then attempts to identify the next relevant information, such as details about a nearby place, a restaurant menu, or an airport guide.

Q: How could contextual discovery work inside a web browser?

Inside a browser, contextual discovery could examine the page being viewed and the broader browsing pattern to recommend relevant entities or additional information. Mayer suggests presenting these recommendations through a panel located beside or below the webpage. The panel would complement the browsing experience by surfacing useful context rather than requiring the user to leave the page and conduct another search.

Q: How could location improve information discovery on mobile devices?

Location can help a mobile service infer what information is immediately useful. If someone is sitting in a restaurant, the service could present a menu enhanced with information about what other people ordered, liked, or annotated. If someone is at an airport, it could provide an airport guide. These suggestions would depend on explicit or implicit location information and user permission.

Q: Why did Marissa Mayer move from search to local and geographic products?

Mayer says she had worked on search for a long period and had watched it grow through many innovations. She viewed the move as an opportunity to try something new while retaining a strong connection to her previous experience. Local information, location, and contextual discovery remain closely related to search, allowing her background to transfer into a newly organized product and engineering unit.

Q: What is the difference between Latitude and check-in services?

Latitude supports persistent, always-on location sharing, while check-ins are explicit statements that a person is at a particular place. Mayer argues that people may accept ongoing location visibility only for a small group of trusted contacts, even if they are comfortable sharing check-ins with a broader group. Combining implicit and explicit location sharing could therefore make the product more flexible and comfortable.

Q: Why does network size matter for Google Latitude?

Latitude is described as a network product, so its usefulness increases when more people participate. A user gains more value when relevant friends or trusted contacts are also available through the service. Mayer connects this principle to Google’s interest in expanding Latitude to additional devices and adding layers such as check-ins through Latitude or Google Maps for mobile.

Q: How does Google approach the integration of acquired companies?

Google’s approach depends on the acquired company, its leadership, its brand, and how its products complement Google’s broader strategy. Mayer explains that larger acquisitions can be more complicated because Google must decide whether and how to integrate them. In some cases, an acquired company may remain an autonomous unit with its own brand and internal functions instead of being fully absorbed.

Q: What is Hotpot and how does it relate to Google’s local products?

Hotpot is described as a personalized recommendation engine connected to Google’s local product ecosystem. It is built into Google Maps for mobile, Places, and local search on the web. Users can rate restaurants through the service, creating information that can support personalized recommendations. Mayer does not characterize it as merely a competitor to another review service, but as a recommendation feature.

Summary

In this video, Marissa Mayer, former VP of Google, discusses her new role at the company and the concept of contextual discovery, which aims to provide information to users based on their location and context without them having to search for it. She also discusses the importance of social media and Google's efforts in the space, as well as the release of the Nexus S and the latest version of Google Maps for Mobile.

Questions & Answers

Q: What is contextual discovery and how does it relate to search?

Contextual discovery is the idea of using location and user context to provide information to users without them having to search for it. For example, if a user is in a new place, we can provide information about that location, such as when it was built and its significance. This concept extends search beyond the traditional methods of voice search or typing queries. We are exploring this concept and its potential applications.

Q: Why did Marissa Mayer give up her role in search to focus on consumer products?

Marissa had been leading the search team for 11 years and saw a lot of innovation during that time. She felt that it was a good time to try something new, and the focus on consumer products, particularly in the local and geo space, allowed her to leverage her background and experience while exploring new opportunities. The decision was also driven by the company's organizational structure, which combines product management and engineering to create a dedicated focus on consumer products.

Q: How has Marissa Mayer's team grown since she took on her new role?

Marissa's team consists of between 800 and 1000 people, primarily focused on building a virtual mirror of the world through efforts like street view and Google Maps. When considering the entire organization involved in these efforts, including street view cars and planes, the team consists of a few thousand people. This is a significant investment for Google and reflects the importance of providing accurate and comprehensive information to users.

Q: Why did Google fail to acquire companies like Yelp, Twitter, and Groupon?

Marissa cannot comment on specific instances, but each deal is unique and brings its own challenges. Larger companies often present more complex integration challenges, and the strategy behind each acquisition must be carefully considered. Google has had successful acquisitions, such as ITA and AdMob, but each deal is evaluated on a case-by-case basis.

Q: What is Hotpot and how does it relate to Yelp?

Hotpot is a personalized recommendations engine that allows users to rate and review businesses. It is built into Google Maps for Mobile and Places, and provides recommendations based on users' connections and preferences. While it is not intended to directly compete with Yelp, it aims to provide personalized recommendations for local businesses, leveraging collaborative filtering and social components.

Q: What is the difference between contextual discovery and contextual search?

Contextual discovery focuses on providing information to users based on their location and context without them actively searching for it. It leverages location data and user preferences to deliver relevant information. On the other hand, contextual search uses the same principles but may require users to voice search or type queries. Both concepts aim to enhance the user experience and provide relevant information.

Q: Is Google planning to release Latitude for the iPhone?

Google has been interested in bringing Latitude to the iPhone, and it appears that they are close to accomplishing that. Latitude is a network product, and the more users there are on the network, the better it works. By expanding its availability to the iPhone, Google aims to increase the usefulness and adoption of Latitude.

Q: What are some of the challenges Google faces in the social media space?

Social media is an important focus for Google, and they are working hard to develop and integrate social components into their products. The success of social media platforms like Twitter and Foursquare in raising awareness and engagement is something Google wants to achieve. While there are challenges in the space, Google is committed to getting social right and investing in its development.

Q: Will Google make further acquisitions in the near future?

Google has been active in the acquisition space and has been making strategic deals throughout the year. While Marissa cannot provide specific details, she acknowledges that acquisitions are an important part of Google's overall strategy and that there will likely be more in the future.

Q: Can you provide an overview of the new features in the Nexus S and Google Maps for Mobile?

The Nexus S is the first device to feature the Gingerbread version of Android, which includes the new Google Maps for Mobile 5.0. The new version offers dynamic map drawing, which provides a smoother and more fluid experience when zooming and panning. It also introduces offline reliability, allowing users to access maps even without a network connection. Additionally, the Nexus S features a compass mode which adapts the map's perspective based on the user's physical orientation.

Summary & Key Takeaways

• Marissa Mayer describes moving from search and user experience into leadership of Google’s local and geographic products. Her organization combines product management and engineering across services such as local search, Maps, Earth, Street View, and Latitude, while exploring ways to connect geographic information with users’ immediate needs and circumstances.

• Contextual discovery is presented as search without an explicit query. A system could interpret browsing behavior, physical location, recent activity, and permitted Google product usage to suggest relevant information. Possible experiences include supplementary browser panels, contextual details about nearby places, socially informed restaurant menus, and guides tailored to locations such as airports.

• The discussion also covers Latitude, check-ins, acquisitions, organizational structure, and personalized recommendations. Mayer argues that persistent location sharing serves a narrower audience than explicit check-ins, while combining both approaches could make location products more comfortable and useful. She also emphasizes that acquisition and integration strategies must be customized for each company.