Preface: It’s interesting what can be done with just flour, water and a touch of salt. A cook can make tortillas, flatbreads, or pasta. A young mother can make finger paint or paper glue. Similarly, it’s fascinating how different business models turn the three elements of need (when, where, how much) into finished sales. This article will describe how some companies have developed specific approaches to the problem of maximizing sales. The data will confirm some of the things apartment owners already know, and alter the importance of others. It develops that there is one specific characteristic that determines rent levels more than any other amenity.
Airbnb Business Model
Airbnb is an electronic market where some people list their homes and other people rent them, generally on a short term basis. It owns no units itself, instead acting as a transaction agent on the willing seller / willing buyer principle. Airbnb does not establish rents, but it does provide information that may help the host-landlord get a sense of the marketplace. Its predictive pricing algorithm was recently released and has made quite a stir, with articles posted at half a dozen or so sites. (Google)
Airbnb is not alone. There is a need for predictive pricing. Other companies have recognized this and help customers reach a pricing solution. Ebay will tell you what similar products previously sold for on their site. Since Ebay deals in personal property (not fixed to a specific location) it doesn’t make much difference where the product is; it’ll be mailed to you. It doesn’t make much difference when it’s being offered: you can buy it on Super Bowl weekend and it won’t affect the price. Although Ebay helps the seller by disclosing historical winning bids for similar products, its business model is an internet-auction system which is not sensitive to time urgency. The reason, of course, is that Ebay sells things. Fred owned it, and Helen bought it. If an object doesn’t sell on Friday, it’ll still be available for purchase the following Monday. This is not the same as renting a living area for a specific period of time, where a Monday sale means it was vacant over the weekend and somebody lost money.
Unlike Ebay, airline seats are time critical and most airlines price dynamically to maximize revenue on that particular flight. The metric is the number of seats still available. All other elements, including departure and arrival locations, or even time of service are rolled into and impact seat availability. Those items are fixed. The variable, which controls the load factor, is pricing. The cheapest tickets are available when there are a lot of unsold seats. A midweek red-eye between hubs offers the cheapest fares because it’s not even half full. It would be a mistake to think the red-eye is cheap because it leaves late at night. It is cheap because there are unsold seats. The airline can never recover the money lost from an empty seat, just like Airbnb can never regain the income from rent lost because the unit sat empty last weekend. In that time critical way Airbnb resembles the airline pricing model.
vs. Livery Service
Uber uses dynamic pricing when appropriate. The company sets prices after considering what you want, when you want it and how long you want it. You can choose from four classes of service. At the low end of the price scale is UberX (everyday cars less than 15 years old). The other end, UberBlack, features livery-type cars (SUV or luxury sedan). The first step is to pick the service level you wish. Once that’s decided, Uber prices depend on how long you will be using the vehicle, what the total mileage will be and exactly when you will require the service. Busy periods can cost more.
Like Uber’s four service classes, an Airbnb client may select housing ranging from a yurt with a firepit in front and a slit trench in back to a fully staffed castle (not available in every city). In the airline world Airbnb might most resemble something like NetJets. The vendee chooses what size plane he wants for the trip and the amenities (special foods, selection of adult beverages, cabin staffing, etc) he prefers.
The thing is, aircraft leasing programs are at heart mostly a cost-plus model. There is plenty of data available to determine the likely cost to provide each element desired by the lessee. Add them up, include pro-rata overhead, and adjust for a little profit. As complicated as the business of full-service aircraft leasing may be, the pricing model is simple: total costs plus an appropriate profit margin.
Optimum pricing for the sort of product Airbnb markets is more difficult. For example, pricing might not be expected to cover costs and overhead. The property owner is probably quite accustomed to paying the costs of property ownership from her salary, so she could be very satisfied with getting just a couple of hundred dollars extra each month. It’s enough to get her hair done and maybe save a little to pay the property taxes at the end of the year. If there is no allowance made for costs and overhead and if what is required is not an accounting profit but only “something extra”, then pricing could theoretically be almost anything and that complicates matters.
You want even more complication? Houses are far more individual than airplanes or town cars. Houses each have their own character and seemingly small things can accumulate to have an extravagant impact on daily rents.
Imagine a small rental of 400 sq. ft., or about the size of a two car garage. It is one of the ground floor units in a building of 96 identical studio units recently built in downtown Columbia, South Carolina. It fronts a major stop for the municipal bus line, and from the side window has a fine view of the Veterans Administration Hospital parking lot. The floor is carpeted in an affordable manner and the walls are painted sheetrock. The kitchenette has a two burner electric hotplate, a small microwave, and a refrigerator rescued from a storm damaged motorhome. There is a tiny bathroom with one of those freestanding showers enclosed with a circular plastic curtain. The name of the unit is 2267 Busy Hwy, Unit 2-A. The owner historically gets $35 a weekend night for this place.
Still imagining, visualize another 400 sq. ft. studio unit. This one, however, is in downtown Charleston, South Carolina. It was built of brick in 1763 and was at one time a stable. It is listed on the National Register of Historic Places. Being a former stable, it is of course on the ground floor. Charleston real estate was originally taxed by front footage, so lots were platted narrow and deep. This unit sits in the back of the lot where historical stables are traditionally found and looks out a long garden to the street beyond, where horse drawn trolleys make their touristy rounds past the magnolia trees. Like the Columbia unit, it is a studio with a private bath and only one large room. The floor is quarter sawn 12 inch wide white oak, darkly stained and laid just after the Revolutionary War. It has been lovingly maintained ever since. The walls are the original handmade red brick. There is a kitchenette with a stovetop featuring two small gas burners, a built-in ice box converted to electric refrigeration shortly after Nicola Tesla’s alternating current became available, and a private tiled bathroom with an ancient (but perfectly restored) tub suitable for two. There is still a hitching post in the garden where in times past a horse could be kept while an eligible bachelor was inside the main house wooing the nubile young woman, with her father’s consent of course. They married early in those days.
This unit is known as the Anderson Honeymoon Cottage. The Andersons were together 57 years and had 12 children, 43 grandchildren, and 9 great-grandchildren before they died just six minutes apart. Her hand was still in his, her nails freshly done. Naturally, there is a carefully shined brass plaque inlaid in the wall just outside the cottage’s front door with the story in raised letters. At the bottom of the plaque is a cameo of a woman’s aged hand lovingly held in a man’s worn palm. Excuse me for a moment. I think I need a tissue.
What do you think? This former stable is functionally the same as the Columbia property. They both are 400 sq. ft. on the ground floor, have a kitchenette, a bath, and sleep two. We already know the Columbia property can get $35. Does that help us price the terribly romantic former stable in Charleston?
Internet pricing models have been around for a while and we’ve learned a few things. For example, some business models like the Ebay auction approach work best when marketing non-time sensitive items.
We have seen that the Airbrb model is time sensitive like commercial airlines. The airlines offer dynamic (demand-sensitive) pricing, and that seems like it would be a good element thing to incorporate into the Airbnb model.
The airline most similar to Airbnb is something like NetJets, which will provide client requested amenities. In the Airbnb model it might be something like the landlord-host acting as cook or tour guide, or even personal shopper (“And it’ll be one year exactly from the day we met. That’s the “paper” anniversary you know. So we’d like a cardboard box of wine in the fridge, please”).
So we have existing pricing models that can be imported in part into the Airbnb business plan. The thing is, they all prove inadequate. Even after borrowing parts of several different programs there still are important elements of the current Airbnb model left out.
That was the issue facing Dan Hill, a brilliant man who led the team responsible for Airbnb’s pricing algorithm. In a recent article he described how his company addressed the most likely rental rate. (http://spectrum.ieee.org/computing/software/the-secret-of-airbnbs-pricing-algorithm). Maximizing income is important to Airbnb. Lost (or underpriced) rentals impact its business both directly (like the airline’s lost transactional revenue) and indirectly through inventory losses as potential landlords become discouraged (like unsuccessful private livery drivers). Pricing is too important to everyone involved to leave host-landlords rudderless.
Flour, Water and Salt
Through his investigations Mr. Hill determined that the three salient data points necessary for estimating market rental rate within the Airbnb model were (1) similarity, (2) recency, and (3) location.
Similarity includes not only the type of property (home, condo, yurt, Buckingham Palace), whether it’s an entire property or a private room, or the number of people the space sleeps but also how many reviews the property has. People, it developed, are willing to pay a premium for a place with reviews. Mr. Hill writes that even a single review “makes a huge difference to a listing”. When you think about it, that’s probably how it should be. The host-landlord has a vested interest in putting lipstick on the property, but an independent review is more of a third party opinion. An endorsement can make a huge difference, Mr. Hill found, in marketability.
Recency is not what it sounds like. Although markets change quickly data from last weekend may still be relevant. They don’t change that quickly. But travel is a seasonable business, so pricing from even a month or two ago may not be comparable. It might be better to compare today’s activity to what it was this time last year. Counter intuitively, recency can mean comparing this season to the same season twelve months ago.
Location is very specific. It is more than simply comparing a host-landlord’s property with local hotel rates. Hotels tend to cluster in certain areas, while Airbnb has rentals all over town. And even then, neighborhoods can change abruptly. As an obvious example, it would be hard to compare a property right on Riverwalk in San Antonio with another property off Riverwalk, perhaps in the South Flores St and Probrandt area. Or a property in the Canal section of Venice (California) with an otherwise similar property in the Oakwood section. Sometimes a neighborhood consists of only a handful of streets. When Mr. Hill learned this he hired a cartographer “to map the boundaries of every neighborhood in our top cities all over the world . . . we could use (this) to accurately cluster listings around major geographical and structural features like rivers, highways, and transportation connections
Businesses like Ebay, commercial airlines, livery services, and Airbnb each have to deal with the three elements of need (when, where, how much) to turn availability into revenue. How they individually succeed depends on what their customers expect from them. Ebay and commercial airlines benefit when they are the low cost providers. Low cost providers are in a continual race for the bottom. Every buyer likes a bargain, so the cheapest provider wins all the marbles.
If you aren’t the low cost provider, all you are selling are the amenities you offer. That’s what NetJets, livery services, and Airbnb do. The most variable amenity is service. Once a certain service level is reached, the issue becomes one of maintaining that standard. Amenities, however, don’t have the universal appeal of a low price point because not everybody wants the same amenities and it turns out that not every selling point contributes equally to the sale. Even disregarding the romance, that’s why property in Charleston rents for much more than the newer property in Columbia, South Carolina. The most permanent amenity that influences pricing is the neighborhood. And neighborhoods are terribly specific. Airbnb goes to unusual lengths to identify their borders. Please see https://www.airbnb.com/locations.
This is not new to us. Apartment owners learned long ago that units rent better in some neighborhoods than others. What is new is the degree that “neighborhood” outweighs any other single amenity among comparable properties. And also interesting is confirmation that neighborhoods are very specific. Cross the street and you’re somewhere else. We knew that intuitively, but now there’s an entire data base that supports it.
This article is for informational purposes only and is not intended as professional advice. For specific circumstances, please contact an appropriately licensed professional.
Klarise Yahya is a Commercial Mortgage Broker. If you are thinking of refinancing or purchasing five units or more, Klarise Yahya can probably help. Find out how much you can borrow. For a complimentary mortgage analysis, please call her at (818) 414-7830 or email Info@KlariseYahya.com
If you’ve missed some of the prior articles, basic guidelines on successful investing are in my book “Stairway to Wealth” available at www.LuLu.com.