Sign up for our daily recaps of the ever-changing search marketing landscape.
Google Now, Microsoft Cortana & The Predictive Search World: Recap From SMX East
Contributor Benjamin Spiegel recaps a presentation from SMX East about the growth of predictive search technologies and the resulting opportunities and issues for search marketers to consider.
What will you search for next? That’s what Google and Microsoft are trying to figure out with their Now and Cortana products and, with all the data becoming available, they may know before you do. As a marketer, you need to be aware of these developments so you can improve the chances of your brands and products being proactively presented to people before they even know they want them.
That topic — predictive search — was the focus of a session by Cindy Krum, founder and CEO of MobileMoxie, at last week’s SMX East conference in New York City.
Because of Google’s dominance in the data collection and aggregation field throughout its Googleverse, Krum focused chiefly on Google Now, with only a brief mention of Microsoft’s Cortana, the predictive search engine on Microsoft phones.
Her overall premise was that Google Now is attempting to present answers to people before they even think to ask/search for them. The way Google’s reach and data pipes have grown and the types of answers it offers present opportunities and issues for digital marketers to consider as predictive search engines continue to develop.
Initially, I was worried about sitting through another session that talked about the future of voice search and the different ways we would be yelling commands at our devices. Instead, Krum took us through the full spectrum of data that influences and fuels the information that powers Google Now and provided some actionable insights that can help marketers prepare for future of search.
Krum covered these key topics:
What Is Predictive Search?
Predictive search is based on the semantic prediction of needs; predictive search engines return results based on the current context, the historical behavior, aggregated user behavioral patterns and the active solicitation of information.
By combining these indicators, search engines can predict the user’s current intent and serve the best possible answer. We don’t have to shout at our devices at all when predictive search serves us results we have not even had time to query.
Why Is This Important?
- With the rise of mobile devices, wearables, the anticipated growth of the Internet of Things and the myriad possible device interactions, it is more important than ever before that platforms provide predictive answers to our questions in formats we can easily view anywhere.
- More and more, people are going to different places to search for what they want, from Amazon to iTunes to Netflix. Google wants to retain that search traffic, rather than have users go directly to vertical engines, so it is showing them what they might search for — before they even know they want or need it — based on all the data.
What Signals Are Used To Predict Searches?
Google Now leverages the data it collects and aggregates from its users through Search, Mail, Maps, Calendar and Google Plus — basically everything that is using a Google login.
It understands who you are, what you are doing and where you are doing it to predict what you want based on user behavioral patterns. (Microsoft’s Cortana does not have that breadth of data points and aggregation and bases its results on user-set preferences.)
Geolocation is also a huge context factor for Google Now results. With smartphones and wearables, users are no longer geolocated by an IP address but by the device’s physical location and its motion. Google Now uses location history to learn where you live or work; tracks your movements based on GPS check-ins; and uses date, time and your search history to serve you highly relevant traffic and weather reports, local restaurants, travel recommendations, flight schedules and more.
What Do Predictive Search Results Look Like?
Krum says if Knowledge Graph was the first kind of predictive card, Google Now is Knowledge Graph 2.0, providing a truly predictive search experience that creates and serves cards with information you might search for (music, news, hotels, shopping and so forth).
You can get reminders about your favorite activities, details and snippets about a regularly watched show, recommendations based on relationship analysis and more.
Apps Integrating With Google Now
One of Google Now’s leading-edge features is its deep (and growing) integration with third-party apps such as Lyft, Uber and Open Table. App signals are launching geo-based alerts and recommendations.
While Google Now is on all new Android phones, it is available as a downloadable app for iPhone. Krum warns, however, that it is not well integrated with the Apple OS platform — yet.
Where Is Predictive Search Going?
Mobile searches are cross-pollinating with desktop search, and Google Now is merging the look of the platforms. Crossover between Google Play and YouTube is happening, and more mobile search results are showing up as cards (This is happening somewhat on desktop searches, as well).
Google Now is also integrated with offline devices, bringing the Internet of Things to predictive search. Krum pointed out that location cards are easy to favorite from the browser and that menu cards from restaurants are likely to appear soon.
There’s no question that Google Now is presenting exciting opportunities for search and seems tailor-made for the small screens of wearables and devices. Marketers have yet to find a way to leverage this, but I am confident Google will offer sponsored opportunities soon.
For now, Google Now is holding its dominant position due to the massive amount of data it has about its users’ behavior and preferences, but I am sure some interesting players/competitors will emerge soon.
Some opinions expressed in this article may be those of a guest author and not necessarily Search Engine Land. Staff authors are listed here.