Airbnb’s pricing algorithm and Aerosolve, its open-source machine learning tool

A little machine learning can have a big impact.

Dan Hill, product lead at Airbnb, wrote the company’s pricing algorithm after the British-based rival startup he cofounded, Crashpadder, was acquired by Airbnb, the short-term rental giant, a few years ago.

Hill has published a plain-English article about the factors involved in Airbnb’s pricing algorithm in IEEE Spectrum, a journal for the Institute of Electrical and Electronics Engineers.

In more good news, Hill’s team has released Aerosolve, the open-source machine-learning tool on which Airbnb’s pricing algorithm relies, on the Github code-sharing platform.

The Aerosolve machine-learning package enables people to upload data to improve a set of algorithms in a way that can continuously inform the model.

Aerosolve isn’t just for travel industry applications. Airbnb Engineering suggests other uses, such as predicting household income based on demographic, map, and other data.

Yet travel startups looking for a free tool designed by data scientists to help them achieve scale and leverage may like Aerosolve as a free and powerful machine-learning package.

Airbnb used machine learning to help refine the “weights” it gave to various factors in its pricing algorithm. Hill writes:

“Here’s where the learning comes in. With knowledge about the success of its tips, our system began adjusting the weights it gives to the different characteristics about a listing—the “signals” it is getting about a particular property.

We started out with some assumptions, such as that geographic location is hugely important but that usually the presence of a hot tub is less so. We’ve retained certain attributes of a listing considered by our previous pricing system, but we’ve added new ones.

Some of the new signals, like “number of lead days before booking day,” are related to our dynamic pricing capability. We added other signals simply because our analysis of historical data indicated that they matter.

For instance, certain photos are more likely to lead to bookings. The general trend might surprise you—the photos of stylish, brightly lit living rooms that tend to be preferred by professional photographers don’t attract nearly as many potential guests as photos of cozy bedrooms decorated in warm colors.

As time goes on, we expect constant automatic refinements of the weights of these signals to improve our price tips.”

Hill’s broader argument in the article is urgent. He insists that the entire sharing economy needs machine intelligence to help set prices.

His argument rings true to Tnooz, partly because we’ve written about a few revenue management startups aiming to help Airbnb owners price their lodgings. (See our profiles of Beyond Pricing and Everbooked.)

It makes sense that similar services will debut for other sharing economy services, like RelayRides, a short-term car rental service.

Hill’s article also offers insight into how Airbnb cracked the code of pricing suggestions for users in an easy-to-use way.

One thing the article does not discuss is the user experience. The simplicity of Airbnb’s pricing algorithm user interface deserves admiration. Its slider enables users to pick a nightly rate for a rental based on various factors. Hopefully its designer will write a similarly detailed article sometime soon. (Hint: Tnooz is open to guest articles.)

READ NOW: The Secret of Airbnb’s Pricing Algorithm

Aerosolve (open-source code)



Airbnb buys UK apartment rental service Crashpadder

Beyond Pricing seeks to boost revenue for property rentals

Everbooked brings Big Data to Airbnb rental management

Source: Airbnb’s pricing algorithm and Aerosolve, its open-source machine learning tool

Via: Google Alert for ML