Recently, one of our writers cycled into Saltillo, Mexico, a town known for its museums, cowboys, gorditas and pan de pulque, a sweet bread made with fermented agave cactus. He wanted to get to a restaurant he had seen earlier in the day, but in the bustling historic downtown he had no idea where to find it. He entered the restaurant’s exact name into Apple Maps and waited for it to show up.
Apple referred him to a cafe in San Francisco.
Later, in the town of Cedral, he tried a simpler search: “hotels in Cedral.” Despite detecting his exact location in Cedral, Mexico, Apple Maps offered him results in London.
These kinds of flubs are exactly why Apple, and other map apps, have failed to overtake Google as the popular favorite—not that they aren’t trying.
The Limits of Smart
We like to think of apps as being able to solve any problem and answer any question. Technology marketers tell us that devices will know what we want before we know it ourselves, that they will be smarter than we are. In a way, that’s true—many apps have access to reams of data far greater than any human being could memorize or even sort through. But no matter how “smart” the app, that data has its limits.
A big part of those limits is how well it is interpreted. The apps that seem smartest are those that “learn” as they go: If millions of users all search a certain way, or tap first on certain results, that’s a strong indicator of what results an app should offer first.
That “learning” is exactly what makes a map service useful. When you click on the “directions” mode, are you looking for a place across town or across the country? If you search for “restaurants,” how big of a radius are you likely to consider? And if you look for a business by name, are you looking for an exact match in a different city, or did you just misspell the name of the place down the street?
These are the sorts of questions that a map algorithm has to predict answers to—and the more users that have used it, the better those predictions can get.
Age vs. Innovation
Currently, Google Maps wins that battle on two fronts. Google indexes more web pages than any other search engine, and likely has the most raw data to populate a map in the first place (that’s why it can find hotels and restaurants that don’t always show up on its competitors). But more importantly, of the major map sites and apps in the industry today, Google Maps has been learning from its users the longest—since 2005, compared to 2012 for Apple Maps or 2009 (with roots in technology going back earlier) for Bing Maps. For a service that depends on predicting user needs, those extra years of use represent a major edge.
But age and experience aren’t all that matters in the map wars. MapQuest, for example, was around for years before Google Maps but lost prominence thanks to Google’s better UI, more robust features and superior marketing. It is exactly these kinds of differences—better features and usability—that Apple and Bing hope to use to steal market share from Google.
Apple Maps, for example, enjoys the advantage of being built in to the world’s most popular mobile devices (an advantage similar to what Google once had over MapQuest, being built into a major search engine). The fact that it’s already at the touch of a user’s finger, and is designed with Apple’s traditional emphasis on simplicity, gives it a large user base despite its occasional flubs. And Apple is trying to make up for its smaller pool of data, recently allowing business owners to add or edit their own local listings on Apple Maps.
Bing Maps has nearly as much user history as Google Maps, and Microsoft is trying to distinguish it in other ways. The latest feature is unprecedented in the mapping world: Bing offers traffic condition information even when no live traffic data is available, in multiple countries around the world. The move is based on predictive algorithms that Bing has developed specifically for this purpose, leveraging its own big data in ways Google hasn’t thought of (yet).
Of course, the early stages of predictive traffic conditions are going to be a lot like the early stages of driving directions: prone to errors. But Bing, like all smart apps, will learn and grow over time as users report errors or user behavior drives changes in the algorithm.
Eventually, all of the major map apps will develop large, robust data sets to learn from. Google’s advantage will likely fade, and the benefits of easier UI and better features will woo more users. Which map app will emerge as the champion? There’s no way to know, but one thing is clear: in 2014, it’s already important for a business to make sure their information (written address, location as shown on map, and contact info) is correct on all of the major map apps—not just Google.