What is Predictive Search?
Predictive search, as the name suggests, is based on the preemptive solicitation of information to a user based on their historical behavior and available data points. In other words it’s search engines, like Google, essentially bypassing the need for a “query” to deliver personally relevant information.
In doing so Google & Microsoft are hoping to regain traffic that is otherwise being lost to in-app searches from products & services with well-developed recommendation features (looking at you Amazon)!
What Does It Look Like?
In its earliest form, predictive search auto-completed our search queries as we typed them. Since then, it has taken a more sophisticated turn with the launch of Google’s Knowledge Graph and Google Now in 2012 and Microsoft Cortana in 2014.
For those unfamiliar, Google Now is a feature built into the Google Search application for Android & iOS that will proactively deliver information in “cards” or push notifications based on search habits and other information from integrated Google products (i.e. Maps, Calendar, and Gmail) and third-party apps.
For example, if you have a dinner reservation confirmation (that was hopefully made with PMG favorite, OpenTable), Google may send you a push notification letting you know, based on current traffic patterns when it’s time to get moving, which is pretty neat!
“Cards” are the tiles that can be within the Google search app that will deliver up to date weather, news & stories, personally catered to you and your search habits. I for instance am served information on the latest Cowboys & Mavericks news, the latest releases from some of my favorite bands (2015 was an amazing year!), and a quick piece of news on the Daily Show (that I will regularly watch). Cards will also include any relevant information around upcoming flights & hotel reservations (making it a must-have travel resources), as well as countless other features, due to its ever-growing third party app integration’s.
Why does it Matter?
For the time being, it is still unclear how digital marketers can or will leverage predictive search in the future, but there’s no doubt we should all be cognizant of its growth & evolution. By preemptively delivering information we may be seeking out, not only are our interests being shaped, but our destinations of choice for those interests is predetermined as well.
At the moment, Google cards (an obvious example of predictive search’s evolution), contains no sponsored content at all, however Google is constantly testing new shopping related results such as price drops, in an effort to regain traffic from shopping oriented sites like Amazon. These Price Drop Shopping cards aren’t built off of PPC campaigns, but rather leverage existing Google Shopping feeds, and are served based on a user’s browsing behavior across devices. While the marketer isn’t directly paying for the card impression, the value of a regularly updated feed is paramount.
Ultimately, we may be seeing the first steps into a future in which the “searches” become an afterthought, as machine learning across an infinite number of data-points become more adept to knowing what we need before we’ve even gotten there. This of course carries massive implications not only for search marketing but for overall online shopping patterns and it may all come faster than we expect!