Price Optimization: How Dynamic Pricing Helps Airbnb Hosts Earn Big

Airbnb’s business model relies on a large, active catalogue of hosts. They have to strike a balance between having inventory available for guests to book and keeping properties occupied often enough to make it worth their hosts’ time and effort. The problem is, Airbnb’s hosts aren’t professional hoteliers. They don’t have the hospitality experience to know how to price their listing. When Airbnb first launched, executives noticed that most hosts based their rates on what they wanted to earn instead of what their property was worth. The result? Lower income for both hosts and company. At first Airbnb designed a calculator to suggest pricing based on features and geographic locations. It was an immediate hit with hosts. The tool’s price-setting algorithm was too rigid, though. Even tiny boundary changes had to be put in manually. The company was growing too fast for that to be practical anymore. They needed a better tool, and in mid-2014 they found it: dynamic price optimization using predictive analytics.

Who Uses Airbnb?

Airbnb has more than 650,000 hosts in over 191 countries. It’s most popular in the United States, France, Italy, Spain, and the United Kingdom, but travelers can find bookings in 98% of the world’s nations. 60% of Airbnb hosts view the site as supplementary income instead of a main gig. The average host is interested in building a side hustle or offsetting their mortgage payments to make a big house more affordable. 80% have only one listing: a room in their home, a guest house, or a vacation home. Hosts are split almost evenly between men and women, but the fastest growing demographic leans towards senior women. Guests seem to prefer renting from women over 50. These hosts earn more 5-star ratings, become Superhosts faster, and get more positive long reviews than the company average. Older hosts make about $8350 a year although many are only renting partial properties and guest houses. (As a note, 58% of senior hosts report they would have lost their homes after retirement without using Airbnb.) On the other side of the equation are Airbnb’s 150 million guests. Traveler demographics skew towards Millennials, who make up 60% of the company’s all-time rentals. The percentage of Millennial guests increased 120% in 2017. Guests choose Airbnb to have new experiences, to temporarily become part of a neighborhood in their vacation downtime. Most of them focus on value and space, as well. 88% of reservations are for groups of 2-4 people. Compare that to the hotel industry’s average of 1.2 guests per room and it becomes obvious that these travelers are in groups too big for easy hotel accommodations. They stay longer than hotel guests, as well. 89% of Airbnb reservations are for two or more days; 5% stretch to more than 12 days. Airbnb has a surprisingly positive impact on local economies. Because their guests stay 2.4 times longer, they spend 2.3 times more money than hotel guests. That money stays in more diverse areas of the city, and 45% of spending occurs in the local neighborhood.

Challenges to Profitability

The benefits of Airbnb are pretty clear. Guests get better deals and more space than a hotel room offers. Hosts get extra income while supporting local businesses. However, there are factors that make profiting from the model hard for hosts.


Hosts only have so much free time to manage their listings. Remember, most are using the platform as supplementary income. The majority have full time jobs and families. That doesn’t leave a lot of room for the kind of in-depth market research needed to understand rental pricing. They need tools to make property management less time-consuming.

Market knowledge

Right-pricing an Airbnb listing is complicated. There are static factors at play such as size, condition, location, proximity to transit options, amenities, and local hospitality trends. Unique property features like side gardens or unusual architecture add to the challenge. Demand changes based on the season and local events, too. Unlike large hospitality brands, Airbnb hosts don’t have market experts on staff. They don’t always know what should affect their rates or how often to update prices. This is especially true for the senior hosts preferred by travelers.


Even if hosts have the time to put together a well-researched price on their own, they’re limited in how quickly they can respond to shifting conditions. When inventory falls in an area guests will book at higher rates. If the host doesn’t notice the change in time, they can’t take advantage of the more favorable market and might wind up locked into a reservation at an extreme bargain. These challenges aren’t insurmountable, but they do create an extra layer of complexity for hosts. A lack of confidence mixed with the desire to keep properties booked often led hosts to underpricing themselves for the market. This was great for travelers. However, hosts grew dissatisfied when they couldn’t reach their target annual income. Airbnb was also seeing less profit than projected. The company makes their money by adding a percentage to bookings. Hosts pay a service fee of 3% while guests are charged anywhere from 0-20% of the booking amount. Underpricing hurt Airbnb’s bottom line.

A Game-Changing Solution

In 2014 Airbnb hit on what would become one of their trademark technologies: using predictive analytics to generate optimal prices for any given day. Their AI-powered toolkit provides fluid suggestions for nightly, weekly, and monthly rates. The interface was designed with busy hosts in mind. It’s easy to read and intuitive enough to be understood without a data science degree. The probability of renting the specific space at its current price point is displayed on a color bar that shifts from green (strong rental possibility) to red (unlikely to book). Hosts can experiment with rental options to get immediate feedback on how changes affect rates. Price recommendations are viewable on a calendar stretching to Airbnb’s maximum advance booking date. Some hosts use the tool’s input as a guide. Others take advantage of Airbnb’s Smart Pricing dynamic price optimization tool. This feature automatically adjusts booking rates in response to changing conditions. Hosts can set their minimum and maximum prices to ensure they’re never stuck with a price they can’t accept. Smart Pricing can be set up on schedule. For example, a host might rely on it for weekday pricing while keeping their weekend prices fixed. Suggested prices can be adjusted individually as well. Once manually adjusted, Smart Pricing won’t change it again.

How Price Optimization Works

The initial 2014 price optimization tool was developed over 6 months by a multidisciplinary team of project managers, designers, engineers, data scientists, and researchers. They put an incredible amount of work into the system. Users at all stages of the buying cycle were surveyed to determine how to rank features and amenities. Based on their answers, the Smart Pricing algorithm considers more than 70 factors when calculating rates. Some of these factors are static or change slowly.
  • Number of beds and bathrooms
  • Location
  • Amenities
  • Proximity to public transit
Other factors shift on a monthly or even daily basis.
  • Season
  • Local events
  • Time remaining until stay
  • Available inventory in area
  • Number and length of views on a specific listing
  • Past successful bookings
Reviews also affect bookings. Good reviews raise the suggested rate. Bad reviews might lower the price to keep the property booked and attract guests who might leave better reviews. Airbnb doesn’t have to make adjustments manually anymore. The Smart Pricing algorithm uses machine learning techniques to feed information from bookings back into itself, refining the model so new predictions are more accurate. If a property tends to sell at a higher rate than suggested, for example, it gets a boost to future estimates. The tool even behaves differently for different types of rental purposes. Hosts have the option to set their rental frequency to either “As often as possible” or “Part time”. With “as often as possible” the tool drops to the minimum price more often and is more proactive about adjusting rates. The “part time” option rarely drops bookings to the minimum price but is also less responsive to local conditions. These features aims to generate the maximum potential annual value of each property while keeping hosts happy with their income. Smart Pricing solves another problem Airbnb was struggling with: pricing in new areas where they don’t have enough data for accurate predictions. The tool creates theoretical neighborhoods based on similar real-world neighborhoods with more active rental markets. As more data is gathered the local models change to reflect specific behavior. Doing this gives Airbnb a running start to developing their host network in a new area.

Measuring Results

Pricing is one of Airbnb’s biggest draws for travelers. Using dynamic price optimization to keep properties booked at higher rates has enabled the company to win $450 million in direct annual revenue from traditional hotels. Users are up 13% from last year, too. Hosts reap the benefit of the program, too. Airbnb hosts make more on average that anyone else in the online gig economy. A typical host earns $924 a month, compared to the $380 made by users of the second most lucrative gig platform TaskRabbit. That’s nearly triple the income. Individual properties produce anywhere from $200-$10,000 month depending on location, but more than half of active hosts bring home at least $500 a month regardless of where they are. That’s astounding in context: 85% of gig workers make less than $500 a month through gigs. One North Carolina couple saw a 7% increase in their earnings after beginning to use Smart Pricing. Their property is now occupied 92% of the time, and popular dates book up to a year in advance even at favorable rates. This year the gross rental income for all hosts is predicted to pass $30 billion. Even with lowering fees to stay competitive with travel sites and hotels using their own predictive pricing, Airbnb stands to earn around $3.8 billion in 2018.

Taking the Long View

One of the biggest roadblocks to adopting artificial intelligence is skepticism. The technology sometimes seems like science fiction, and executives aren’t convinced that the potential business value outweighs the risk. Airbnb is a compelling argument against this view. They’ve demonstrated the value of applying targeted artificial intelligence strategies to address specific business problems. That is the real strength of artificial intelligence. It’s not a cure-all or a magic wand that fixes everything. Instead, it’s a tool that- with a little executive guidance- can reduce or eliminate major pain points. For Airbnb that pain point was pricing individual properties. For another business it might be raising email open rates or trimming miles off delivery routes. Whatever the application, artificial intelligence has established itself as a functional tool for modern enterprise. Companies that want to stay viable in a global market should begin exploring how it could benefit them. They might want to think fast, though. With enterprise AI usage rising 60% over the past year, it’s a safe bet their competitors are already moving to adopt.

Tools like embedded analytics make artificial intelligence accessible even to small and medium businesses, but managing multiple data streams can be a hassle. Schedule a free appointment to learn more about how Concepta’s unified analytics dashboards can make reporting a snap!

Related Articles

A Faster Way Forward Starts Here