Latest / Pricing Power Podcast with Fexingo: How Businesses Set Prices, Raise Margins, and Win Customers / How Airbnb Designed Dynamic Pricing for Hosts
Transcript
- Lucas: There's an assumption baked into the way most of us think about marketplace businesses like Airbnb, Uber, or Etsy — which is that the seller sets the price and the platform just takes a cut. Luna: And that assumption is wrong, or at least it's incomplete, especially when you look at how Airbnb actually nudges its hosts toward certain prices. Lucas: Right. So let's talk about Airbnb's pricing system, which is a fascinating case study in how a platform can shape pricing without fully controlling it. Airbnb doesn't set your nightly rate — you do — but they've built a recommendation engine called Smart Pricing that suggests what you should charge based on demand, seasonality, local events, and booking patterns. Luna: And it's optional, right? Hosts can turn it off. But I'm guessing most don't? Lucas: You'd think so, but actually adoption is mixed. Airbnb introduced Smart Pricing back in 2015, and initially a lot of hosts were skeptical. They felt it undercut their sense of control, and some reported that the algorithm recommended prices lower than what they thought their property was worth. Luna: So there's a trust issue. The platform has more data than any single host — they see the entire demand curve across a city — but the host has local knowledge the algorithm doesn't capture. Lucas: Exactly. And that tension is at the heart of the whole episode. Let's get into the mechanics first. Smart Pricing uses a machine learning model that takes into account things like day of week — weekends are more expensive, obviously — seasonality, local holidays and events, booking lead time, and even the listing's own characteristics like number of bedrooms, amenities, review scores. Luna: So it's not just a blanket algorithm. It's personalized to each listing. Lucas: Right. The model is trained on historical booking data from similar listings in the same market. So if you're a one-bedroom apartment in downtown Austin with a 4.8-star rating, the algorithm looks at what other similar apartments have been able to charge on similar dates. It then suggests a price range — a minimum, a maximum, and a recommended price. Luna: And the host can override it entirely, or they can set a floor and ceiling and let the algorithm play within that band. Lucas: Yes. And that's actually the most common setup — hosts who use Smart Pricing often set a minimum and maximum they're comfortable with. They're willing to let the algorithm adjust dynamically, but within guardrails. That's a smart middle ground, psychologically. Luna: So what does the data show about hosts who use it versus those who don't? Do they earn more? Lucas: Airbnb has published some studies — though they're careful not to overclaim — suggesting that hosts who use Smart Pricing see a revenue lift of somewhere between 5 and 15 percent on average. The biggest gains come during high-demand periods, where manual pricing tends to be too conservative. Luna: Because hosts are afraid to price too high and scare away bookings, but the algorithm knows that if you're the only three-bedroom left in the city during a conference, you can charge a premium. Lucas: Exactly. And conversely, during low-demand periods, the algorithm might drop the price to capture last-minute bookings that would otherwise go empty. A human host might just leave the price static and hope, but the algorithm is constantly optimizing. Luna: That sounds like a clear win. So why do some hosts still opt out? Lucas: There are a few reasons. One is the control issue — some hosts see pricing as part of their identity as a small business owner. They want to make that decision themselves. Another is perceived fairness: if the algorithm suggests a price that's lower than what the host thinks the property is worth, they feel devalued. Luna: And there's also the risk that the algorithm might price too low during a unique event it doesn't fully understand — like a local wedding where the host knows demand is even higher than the model predicts. Lucas: Right. The model is only as good as its training data. If a host has a special feature — say, a hot tub that's very popular in winter — the algorithm might not fully capture that premium. So there's a legitimate case for human override. Luna: But let's zoom out. For Airbnb as a company, why does Smart Pricing matter? It's not directly revenue for them — they take a commission, so higher host revenue means higher commission, but that's indirect. Lucas: It matters for marketplace health. If hosts consistently underprice, they get booked quickly but leave money on the table, and they might get resentful. If they overprice, their listing sits empty, guests get frustrated with limited options, and the platform's liquidity suffers. Dynamic pricing helps balance supply and demand, which makes the marketplace work better for everyone. Luna: And it's a competitive advantage. Other platforms like Vrbo or Booking.com have pricing tools, but Airbnb's is more deeply integrated and data-rich. Lucas: Let's talk about a concrete example. Imagine a host in New Orleans during Mardi Gras. A manual pricer might set a flat rate of $200 a night for the whole week. Smart Pricing would probably recommend $400 for the peak nights — Friday and Saturday — and maybe $150 for the slower Tuesday night. That dynamic adjustment captures more total revenue while still filling the calendar. Luna: And the algorithm learns year over year. So the next Mardi Gras, it might adjust further based on the previous year's booking velocity. Lucas: Exactly. And that's where the network effects kick in — more data leads to better predictions, which leads to better pricing, which leads to more bookings, which generates more data. It's a virtuous cycle. Luna: But there's a darker side, too. Some critics argue that dynamic pricing leads to price gouging during emergencies. During natural disasters, Airbnb prices sometimes spike, and the company has had to step in with price caps or fee waivers. Lucas: That's a real issue. Airbnb has faced backlash for allowing hosts to charge exorbitant rates during hurricanes or wildfires. In response, they've implemented policies that cap prices during declared emergencies — for example, they might limit price increases to a certain percentage above the median rate. But it's a delicate line between dynamic pricing and profiteering. Luna: And that's where human judgment and regulation come in. The algorithm can optimize for revenue, but it doesn't have a moral compass. Lucas: So the broader lesson here is that platform pricing is not just a technical problem — it's a behavioral and ethical one. Airbnb's Smart Pricing is a powerful tool, but it only works when hosts trust it and when the platform sets appropriate boundaries. Luna: And it's a reminder that even the best algorithm is just a recommendation. The final decision — and the responsibility — still rests with the host. Lucas: And that's a good note to wrap on. For listeners who want to dig deeper, Airbnb actually publishes some of its pricing research on their engineering blog. It's worth a read if you're building any kind of marketplace. Luna: And for the hosts out there, maybe experiment with Smart Pricing for a month and compare your revenue to a manual month. The data might surprise you. Lucas: Alright, that's it for this episode of Pricing Power. We'll be back with another one soon.