Google Flights Introduces New Predictive Flight Information

9th February, 2018 by

Google has largely created the infrastructure that we think of as the internet. It has revolutionized the way we search, even turning its company name into a verb. It’s an email platform: Gmail accounts for more than 1 billion monthly active users. Android mobile software is one of the only systems to successfully rival Apple’s iOS. Acquisitions like YouTube, which the company bought in 2006 for $1.7 billion, have been cultural game changers on their own before being attached to the Google brand.

Flight of Fancy

In the years since its integration in all corners of the web, Google has aimed to intuitively make everyday tasks easier as well, taking a page out of the Amazon model by creeping in and upheaving the mundane. Google Flights, the company’s online travel assistant, has recently developed more intuitive features in an effort to combat the unforeseen annoyances of modern air travel. The software has always pulled information directly from the airlines, but now Google has equipped the feature with an advanced understanding of historical data, using its machine learning algorithms to warn buyers about delays that have yet to be flagged by the airlines.

According to Google, the site’s AI algorithm crunches the data provided, and uses it to predict delays by running through patterns. For instance, weather patterns at certain times of year can be relatively predictable, especially coming out of cities that serve as major travel hubs—think of Denver, which houses the sixth busiest airport in the United States, with 1,500 daily scheduled passengers. Denver International Airport is used by twenty-five different commercial airlines, meaning that the airport provides a healthy amount of data for Google as it tries to predict when a flight may be delayed, especially when considering Colorado’s notorious winters.

Keeping Your Flights On Time

Google won’t actually flag any flights as delayed or potentially problematic until it is at least 80% sure that the prediction will come to pass. This means that the technology won’t be helpful during the buying process—80% confidence requires up-to-the-minute analysis, which will only aide last-minute purchases—but they will help flyers plan accordingly in the lead-up to their departure, eliminating the element of surprise that awaits once you’re already at the airport.

Additionally, the new Google Flights feature allow travelers to understand what facets different stages of Basic Economy provide, a long-standing issue as each airline has its own vernacular and tier system for flyers to parcel through. Low-cost fares are often the only travel options for last-minute flyers or those on a budget, but the rules and restrictions often vary airline to airline, such as limits on use of overhead space or the ability to pre-select a seat, or, the most frustrating, enforced additional baggage fees.

New Industries to Disrupt

These increased features come hot on the heels of major travel startups that, like all tech disruptors, aim to decrease fuss and increase efficiency. Companies like Hopper have recently added hotel search to their mobile apps, and similarly use big data to analyze airline prices. TripIt, also exclusively developed as an app, introduced a new feature that tells flyers the wait time for security checkpoints.  

This later feature feels like an inevitable addition to the Google Flights system, especially since Google is already aware of busy times at businesses thanks to its ability to track people through Google Maps (a creepy feature to be sure, but have you ever seen the “less busy than usual” note under business descriptions? Beyond handy). Google is rolling out these features with select airlines first, starting with American, United and Delta worldwide. These changes come only a month after Flights’ last major upgrade, which added price tracking and deals, as well as hotel search.

Google Flights has the potential to be a game changer for one of our more frustrating industries. And beyond that, it offers a new way to think of big data and machine learning. If technology is meant to disrupt the average industry, and make good on decades worth of promise, then matters of convenience are perhaps worth uphending too.


(Visited 26 times, 1 visits today)