Over the last several months, Google has mapped out a vision of its future where you’ll have a smart assistant following you from phone to wrist to flatscreen and beyond, conversationally answering your questions and proactively serving up advice, like when to leave for the comedy show you searched for the night before.
While tech giants like Amazon, Microsoft, Apple, and Facebook are all also eyeing some version of this AI-assistant powered world, Google CEO Sundar Pichai has consistently highlighted one of the things that sets Google’s efforts apart: Its geographic data.
If Google’s smart assistant ambitions center around giving you the right information at exactly the right time, Google Maps provides the keys to context.
Its combination of rich geographic, business, and personalized location data can be paired with information from all of Google’s other services like Gmail and Search to create an assistant that (Google hopes) you’ll actually want to use.
Figuring out the best ways to improve that data and put it to use falls to Jen Fitzpatrick, who has led the Maps division since fall 2014.
“Maps is a pretty incredible swiss army knife of a product,” she tells Business Insider. “One of the things we’e thinking a lot about is, ‘How do we take all of the goodness packed within it and bring the right part of that tool to the surface in the moment that’s right for you?'”
The ‘deepest, most detailed understanding of the real world that’s ever existed’
Fitzpatrick has worked at Google since before it even acquired the basis for Maps, Where2 Technologies, in 2004. She joined the “crazy little startup” as one of its first interns in 1999 and led software development on products like AdWords and Google News before heading to Maps.
Back then, there were a lot fewer tech companies with an interest in maps.
In just the last two years, major companies like Tesla, Uber, and, most relevantly, Apple have doubled down on their mapping efforts. Google’s on notice.
“It reinforces that we can’t stand still,” Fitzpatrick says. “We feel an incredible sense of urgency to not just take what we have and keep it going, but to set the bar higher and be the ones who define what the next generation of maps is going to look like.”
The first part of her goal to build “the richest, deepest, most detailed understanding of the real world that’s ever existed” is getting all the data right.
When Business Insider spoke to Fitzpatrick she had just completed a presentation thanking a handful of its most-dedicated “Local Guides,” the volunteers who add and edit millions of Maps data points for restaurants, shops, landmarks, and more.
She told them that their contributions are helping Google get closer to being able to serve recommendations for complex queries like “tourist-friendly nightclubs in Vienna that play Senegalese music.”
The second part is finding new kinds of data to take in. Maps already uses anonymized info from anyone with the app open or location services turned on to crowdsource real-time traffic and predictions, but Fitzpatrick thinks that there are new opportunities ahead too. Her division just bought UrbanEngines, a startup that analyzes the “Internet of Moving Things,” and Maps is looking for new ways to incorporate live information about events or natural disasters.
Plus, she says that we’re “just at the beginning” of seeing a third-party ecosystem on top of Maps. Although she didn’t get into specifics, we’ve already seen Google rolling out more integrations in the last year, from letting you use OpenTable to make a restaurant reservations to prompting ride-hailing services like Uber and Lyft into the directions section of the app.
“We are just at the beginning of what I consider to be the ‘next chapter’ of Google Maps,” she says.
While some of Google’s smart assistant ambitions require pretty basic location knowledge – like giving you a satisfying answer to “How tall is that statue in front of me?” or making your phonebleepat you to leave for work early to make your first meeting because there’s traffic – it gets more interesting when you consider things like recommendations.
If you have your location history turned on, Google’s has been building a little profile about all the places you’ve been. (And here’s how toturn it on or off.)
Looking at my history, Google reminded me that I went to Papa Razzi, an Italian restaurant in Concord, MA, on September 9, 2013. Thanks to all the data that Paparazzi staff and various visitors have recorded about that business, Google can make some teeny-tiny inferences about my preferences. Now multiply that by the gazillion stores and restaurants that I’ve visited.
“It’s about being able to understand more about you and your context and your preferences and the places that you’ve been before and using that to be able to make better recommendations and suggestions to you,” Fitzpatrick says.
When future-me asks my Google assistant where I could host my 27th birthday party, Maps could pull info about places that take large reservations and have available tables on my desired night, ranked by how well they fit with my profile of tastes.
It’s easy to picture how ads could fit into this model too: If you always go to Thai restaurants, a new spot that opened in your neighborhood could pay to have your smart assistant suggest checking it out next time you walk by. Google Maps is already testing “promoted pins” that offer deals, but, for now, they’re not personalized.
“We’re trying to think about what are the next opportunities to make location even more central to an advertising experiences,” Fitzpatrick says.
Right now, Google hasn’t actually released its “Home” speaker or messaging app Allo, the first two Assistant-powered products that it announced at its giant developers conference in May. But they’re coming soon and when they do, it will be Maps and its data doing much of the work under the hood.