U.S. Markets closed

How Airbnb plans to use AI to make travel more personalized for you

JP Mangalindan
Chief Tech Correspondent

Let’s face it: despite advancements in technology, booking travel can still be an onerous experience. Booking the right place to stay on most services and figuring out logistics still feels like it takes longer than it should.

Airbnb Vice President of Engineering Mike Curtis envisions a day in the not-so-distant future when that doesn’t have to be the case, when Airbnb offers heavily personalized experiences that include not just where you’re staying, but other details also, including transportation and entertainment. All of it would be powered by rapidly-evolving technologies such as artificial intelligence and machine learning.

Airbnb has employed artificial intelligence and machine learning to power its search results since 2014.

“Imagine if one day, Airbnb wants to be able to offer a true end-to-end trip: it’s personalized, and just right for you,” Curtis suggested during an interview at the Grace Hopper Celebration of Women in Computing. “What’s right for Tim, for instance, is different from what’s right for you, and it’s different than what’s right for me. And if you want to give a true differentiated travel experience that’s really personalized, you’re gonna have to really understand that person and the data behind it, to personalize the experience to them.”

To be clear, Airbnb has already deployed artificial intelligence and machine learning in search results since 2014. Before then, search rankings were determined by a set of predetermined software rules based on a small number of factors, or “signals,” such as the price of a place and how many bedrooms it has.

Airbnb VP of Engineering Mike Curtis envisions a day in the not-so-distant future when his company offers a truly personalized end-to-end travel experience. Source: Airbnb

Now all Airbnb search results are personalized to some degree. The company pays attention to hundreds of different signals, including your preferences, rating and search history to figure out which search results you see on the desktop and mobile app, and how those search results are ranked.

“All of that said, I think we’re just at the beginning of what we could do with personalization and where I think we’ll go,” Curtis said.

Curtis contended Airbnb’s algorithms could be fine-tuned and evolve to a point where it uses object recognition to pinpoint — without the aide of a human worker — the look and feel of a place, and the commonalities among the listings you click and dwell on.

“What are these things that are common between the places you’re looking at?” Curtis explains. “It could be, can we tell that it’s in an area that’s surrounded by nature, maybe there’s a lot of greenery outside? Is it in a very urban environment? Maybe it’s on a high floor? Does it have a pool? Is there a visible coffee maker in the kitchen? Those are signals we can be picking up on to help surface other places that have similar characteristics.”

That kind of forward thinking helps explain Airbnb’s rapid growth since the company’s founding in 2008. Since then, the company has enabled more than 200 million total “guest arrivals,” or stays, and reportedly became profitable in 2016. Earlier this year, Airbnb closed a $1 billion round of funding, valuing the company at nearly $31 billion, according to a CNBC report.

JP Mangalindan is a senior correspondent for Yahoo Finance covering the intersection of tech and business. Email story tips and musings to jpm@oath.com. Follow him on Twitter or Facebook.

More from JP: