All Sharing Economy, All the Time

So it’s been a long time since I posted here. Basically, I thought I’d said all I had to say about the sharing economy: it’s being covered in so many places now a blog has less of a contribution to make. So I pondered the meaning of life and thought I’d move on to other things (and thanks to Kieran for counsel on that).

But oddly enough, requests for comment and posts about the sharing economy continue to come in, far more than any other topic I’ve written on, so that’s where the demand is (thanks to Trebor, Cameron, CBC (times 2) and Andrew (even if we don’t get anything worthwhile from that effort)).

What’s more, the dividing lines are being drawn increasingly clearly as labour and community groups line up on one side, and fans of deregulation, conservatives, republicans and Fox News line up on the other. It’s good to be contributing to something worthwhile (thanks to LAANE from Los Angeles for a chance to support their efforts).

So I’ll be spending the next several months focusing on the sharing economy, and hope to build up a useful set of resources here, as well as contributing elsewhere. It’s likely that this blog will be more narrowly focused than in the dim and distant past when I used to write about other things. It will also get repetitive and repetitive. You have been warned.

Airbnb in Barcelona

So Airbnb is in trouble in Barcelona (Guardian, El Pais, El Pais again). The company was fined €30,000 by Catalonia’s local government the Generalitat “for illegally commercializing short-stay apartment rentals that are not listed on the Catalan Tourism Register.” Here are a few notes for context.

Barcelona has a particular problem with tourists: it has too many of them. “The city’s 1.6 million residents have seen the number of visitors to the city skyrocket from 1.7 million in 1990 to more than 7.4 million in 2012.” As one besieged local says, “We’re part of what they’re selling, but they’re destroying it.”

Airbnb has done its usual lightweight study on the city. It’s interesting partly for what it doesn’t say, because as a city with a highly developed tourist industry, Barcelona is a sign of what is likely to be the pattern for Airbnb for the future. So, here’s a quick rundown on Airbnb’s usual economy with the truth.

I actually found more hosts than the 4,000 that Airbnb claims: my recent scan of their site picked up 6442 hosts and 11284 listings. (part of a series of studies: here, here, and also here).

Airbnb avoids its usual claim of “the overwhelming majority of Airbnb hosts in Sydney are ordinary residents who rent out the home they live in” because it’s just not true. 75% of hosts rent out a single listing, which is lower than in most other cities. And the majority of Airbnb listings come from people who are renting out more than a single offering, and I estimate that over 60% of Airbnb revenue comes from multiple listers.

The idea that Airbnb is made up of hosts sharing a part of their home also fails in Barcelona. 59% of listings are “whole home”, and these make up a whopping 81% of Airbnb revenue from the city.

In short, Barcelona is exactly the kind of place that needs to regulate tourism, in order to keep a balance between the income from tourists and preserving the city as a place to live, work (and visit). Airbnb is promoting its “home-sharing” story to justify its lack of interest in these regulations, but the story has even less truth to it than in New York and elsewhere.

Airbnb stats everywhere (and a question)

Nice to see newspapers looking into Airbnb stats:

At The San Francisco Chronicle, Carolyn Said takes a thorough look into the data for that city. She followed up with a blog about my previous postshere.

The Guardian looks at Airbnb in London here.

This after travel site Skift reposted my reports here and here.

Both The SF Chronicle and The Guardian’s data are consistent with mine, so we are all obviously finding the same pages from the Airbnb web site. There’s no doubt the newspapers have better ways to present some of the data, and of course they are looking at an individual city (so they do neighbourhood-by-neighbourhood drill-downs) while I’ve been looking across multiple cities. One nice thing about Carolyn Said’s piece is that she complements it with some interviews that give a peek into likely incentives and motives behind some of the patterns that the data show.

We are all limited by what information is publicly available. One additional piece of the puzzle was given when Airbnb executive Chip Conley wrote that 70% of guests leave reviews, which helps in the conversion of reviews into actual visits.

Question for those who have read this far. In many cities, the number of “whole home or apartment” listings is about a factor of two more than the number of “private room” listings (see below). But in Paris, the ratio is more like 5 to 1 and in London it’s about 1 to 1. What’s happening in these cities — particularly Paris? Any ideas, please add to the comments.


Notes from Personal Democracy Forum 2014

So I was lucky enough to be at the Personal Democracy Forum at the beginning of June in New York. A few scattered thoughts.

I had a prejudice that this would be something of an inward looking conference, and was happy to be proved wrong. The range of perspectives was much broader than I expected, and while the core of the conference is the intersection of the internet with electoral politics and governance, the range of topics was broader than I expected too.

Which made the many references throughout to some loosely-defined “we” a bit odd. These were references from the stage and from organizers that implied a community around the internet and its use in (loosely progressive?) politics. This was the 11th PDF, a lot has changed over the last decade, and I’m not sure there is such a “we” any more.

The highlight of Day 1 was the appearance by video of Edward Snowden, a year after the first published leaks. Snowden impressed. He came across as thoughtful, well-read, articulate, and modest, which is a compelling combination. He even looked bashful at the prolonged applause, which was winning. I wish I could say the same for John Perry Barlow, who was “in conversation” with him, but I can’t. Barlow kept dragging the conversation around to himself and kept prompting Snowden to agree with him – prompts that Snowden politely stepped around. Snowden’s main point was a simple one: that, whatever one thinks of states spying on each other (as in, the US spying on Angela Merkel for example), mass surveillance of entire populations is significantly different and crosses a pretty clear line into illegitimacy. In addition, it has not even proved to be useful for its stated goals of preventing terrorist attacks. Snowden may turn out to be as pivotal figure for our time as Daniel Ellsberg was for the Vietnam War era. The video is here.

Some other selected personal highlights from the talks, with links where I can…

The theme for Day 1 was “Save the Internet”, focusing on the role of an open internet for political freedom. I found Katy Pearce‘s talk on “How Authoritarian Regimes Take Advantage of Social Media” and Zeynep Tufekci’s talk on “Movements in a Connected Age: Better at Changing Minds, Worse at Changing Power?” both fascinating. Pearce has been looking at activism in the former Soviet states for years, and she paints a complex picture of the tensions and continuing changes as digital technologies morph and take on new roles in society. Tufekci returned from her native Turkey with interesting ideas about the difference between what is needed to spark sudden uprisings and what is needed to sustain a social movement or to achieve lasting political change. Demonstrations, she argued, act as a signal for a movement’s organizational capacity. When demonstrations happen more easily, governments may misread the organizational strength they imply–and perhaps this happened in Egypt in 2011–but such misreadings won’t persist.

Later on, after my own panel session on the sharing economy was over (written up by Sam Roudman here, I felt it went pretty well but then I’m not the one to ask), more talks. Susan Crawford on the potential for municipal investment in “dark fibre” (ie, optical cables that are open for use) was all new to me – she writes about it here. Sue Gardner, now leaving Wikipedia, is an inspirational figure and her concerns about Building the Public Internet were compelling. You can find similar thoughts by her here from last year.

Day 2’s theme was “The Internet Saves”, the idea being to focus on digital activities that used the Internet for good. The first morning session was great: Matthew Burton and Mike Bracken both talked about the importance of government service – a welcome change from the usual tech focus on entrepreneurship as the solution to social problems – and both delivered far more than they promised, which is the right way to go. If you have 17 minutes, do listen to Bracken: entertaining and novel. Bracken draws his language from Tim O’Reilly, speaking of “Government as a Platform”, but his conception differs from that of O’Reilly (does one write “O’Reilly’s”? that doesn’t look right). The bulk of the O’Reilly “Government as a Platform” talk is more accurately described as “Government as Data Source”: government provides the data, private industry uses it to build services. Bracken was talking about building a complete stack of services within the government, and to me that’s a different take on the topic, and one that is far more promising.

But then came Brad Smith, Microsoft’s General Counsel, and the contrast could hardly be more stark. Lots of overblown rhetoric and ringing phrases but much less substance and little self-awareness. Let’s just say he sounded like a trial lawyer. And then much of the rest of the morning reflected the Omidyar Network’s approach to dealing with social issues by setting up businesses to deal with it, an approach I am not fond of. Omidyar Network is clearly doubling down: one short talk here was about HandUp, a startup that funnels donations to the homeless. Yes, a startup. So there was more for me to grumble about here, but again – that’s the point of a good conference, to hear different points of view.

The format for the whole program was talks of about 20 minutes with no questions (plus breakouts). It worked well: allowing Q&A would have slowed everything down, and a lot of positions got spelled out just by having the variety of presenters on stage.

So good job TechPresident, and thanks for the opportunity to take part.

The shape of Airbnb’s business (II)

Collecting data for several cities both in November 2013 and in May 2014 gives a look at some ways in which Airbnb’s business is changing in these cities. The cities surveyed are:

  • Paris and New York (Airbnb’s two biggest markets)
  • San Francisco (the home of Airbnb)
  • London, Rome, Berlin (three European capitals and major tourist cities)
  • Toronto and Chicago (two smaller markets).

As in Part I, the surveys were carried out by a two-stage method:

  • Stage 1 used the Airbnb search pages for a city to collect Room ID values (a number that uniquely identifies an individual listing).
  • Stage 2 visited the listing page for each Room ID and collected information about it.

The data were stored in a database and the charts here are based on queries of that database. The survey code (and database) are available on github.

The main observations are:

  • It is possible that the number of listings has reached a maximum in some of Airbnb’s largest cities, particularly in the USA, while other cities do continue to show significant growth.
  • There is a surprising amount of churn in Airbnb’s business. Over a third of the listings on the site in November were no longer there in May, to be replaced by a similar (or, in some cases, greater) number of new listings.
  • Around 90% of all ratings on the site are 4.5 stars or 5 stars (out of five).

These observations suggest some conclusions:

  • Airbnb’s move to professionalize its host base may be a reaction to the churn in the host population, a necessary step to generate continued growth.
  • In April Airbnb expelled 2,000 listings from its New York offerings, claiming that they did not offer a positive experience. Many of those listings must have had either no ratings at all or were rated highly (4.5 or 5) by their guests.
  • The rating system does not provide a reliable assessment of host quality.


Figure 1 shows the net change in the number of listings in each of the eight cities for which data was collected in both November and May, as a percentage of its November value.

Some cities seem to have maxed out in terms of listings. Six of the cities saw either decreases or increases of less than 10% in the total number of listings. Each of the three US cities in the study lost listings between November and May, while Berlin and Rome were big gainers.


Figure 1: Change in number of listings in cities.

The net gains or losses mask a bigger change. Figure 2 shows that in the six month period, well over a third of existing listings vanished from the site, and were replaced by a similar (or, in the case of Berlin and Rome, substantially greater) number of new listings. For both Paris and New York, well over 40% of listings vanished, only to be replaced by new ones. Figure 3 shows the total number of vanished and new listings.

The high churn rate must introduce uncertainty for Airbnb’s future, as it is relying on a steady stream of new listings to replace those that are vanishing from the site. Perhaps this is one of the motivators of Airbnb’s efforts to professionalize its host base: hosts who make a commitment to the business being more likely to stick around.


Figure 2: New and vanished listings between November 2013 and May 2014, as a percentage of the November number.


Figure 3: absolute number of new and vanished listings

[Note: I did wonder whether this high churn rate could be an artefact: whether the displayed room ID values could have changed between the November and May data collections. A query showed that, of the 10597 listings that occur in both the November and May collections from New York, 10529 have the same host. I concluded that room ID values are stable.]

We can look a little closer at those vanished listings. Figure 4 shows that the vanishing listings have fewer reviews than the overall population in the city.


Figure 4: Mean number of reviews for vanished listings and for all listings

Figure 5 shows the ratings that the vanished listings had received from guests. The overall distribution of ratings is investigated later in this report, but Figure 5 shows that, while many of the listings had no reviews at all (Airbnb recently claimed that reviews are left for 70% of visits, so it is likely that these listings had no visits either), of those that did have visits the average rating was overwhelmingly high (4.5–5).


Figure 5: The number of vanished listings by average overall rating

Figure 6 emphasizes the point, showing that the average rating of listings leaving the site is within 0.1 of those that stay on the site.


Figure 6: Average rating of listings that leave the site (red) and which have been on the site for the whole period (green). Note that the y axis starts at 4.5.

It seems likely that many hosts try Airbnb, have a few guests (or none at all) and then simply decide that, for whatever reason, the service is not for them. To repeat the conclusion from the overall churn rate, the high turnover rate does raise questions about what long-term population of hosts the site can support, and how it can ensure that those who try and leave the site are replaced by new hosts.


Trust has long been a key part of the Airbnb business, and technology has been seen as the magic ingredient. Airbnb CEO Brian Chesky expresses the company’s confidence when he says of city-level rules that “they’re primarily set up for screening. To protect consumers. Well it turns out that cities can’t screen as well as technologies can screen. Companies have these magical things called reputation systems… We think government should exist as the place of last recourse.”

There are two sides to Airbnb’s trust system. There is peer-to-peer trust, often described as a reputation system, which consists of ratings, personal profiles, and comments written by hosts or guests and seen by other hosts or guests. The other side is centralized trust, which consists of central payment, verified ID, and a complaint system.

Much of the talk around the sharing economy focuses on reputation and peer-to-peer trust (as in Chesky’s comment above). That’s because it’s the peer-to-peer aspect that is new: after all, in one sense we have forever trusted strangers to deliver services from food to haircuts to regular taxis, and the centralized component of the system is not that different to many hospitality companies. Jason Tanz writes in Wired that

We are entrusting complete strangers with our most valuable possessions, our personal experiences—and our very lives. In the process, we are entering a new era of Internet-enabled intimacy.

So what do the numbers have to say about peer-to-peer trust?

Figure 7 shows the distribution of ratings for each listing. Airbnb provides ratings under a number of categories (cleanliness, accuracy of description etc) and collects them together as “overall satisfaction”. The site does not give the ratings for individuals (although individual comments are visible) but provides an aggregate, from 0 to 5 stars with a half-star granularity.


Figure 7: Overall satisfaction ratings for listings in each city. The ratings for 0-3 are so few that they are collected together. Listings with no ratings are excluded.

The vast majority of ratings are either 4.5 or 5 stars. The finding is consistent with ratings on other sites (e.g. an earlier look at BlaBlaCar, or more serious studies of eBay listed there). When we rate each other, ratings become more a courtesy than a judgment. Just as restaurant tips only weakly correspond to the quality of service, so a rating of 4.5 or 5 is more a way of politely concluding an exchange than it is of assessing the behaviour of a host or guest.

The ratings obviously fail to distinguish among the people on the site, and so are not providing the service for which they were intended. This is part of the reason why Airbnb and others have moved to emphasize the centralized trust aspect of their systems. But centralized trust is the same reason we trust the fast-food restaurant cook or the hotel cleaning staff. The major sharing economy sites rely on discipline and the potential for removal from the site to provide security for the users, just as traditional industries do.

Churn and Trust in New York

The Airbnb court case in New York raised the profile of the service there, and resulted in Airbnb ejecting 2,000 listings from its service.

Airbnb claims that “we found that some property managers weren’t providing a quality, local experience to guests. These hosts weren’t making their neighborhood stronger and they weren’t delivering the kind of hospitality our guests expect and deserve. In some cases, they were making communities worse, not better.”

With 10,000 listings vanishing from Airbnb in the New York area, it is difficult to know which 2,000 were removed by Airbnb and which left for other reasons. Some individual hosts with large numbers of listings have been removed, but that still leaves many unaccounted for. Figure ny-vanished shows that, despite Airbnb’s suggestion that the hosts are not “delivering the kind of hospitality our guests expect”, most of the removed listings must have had either no ratings at all, or must have had good ratings from their guests.


Figure 8: The “overall satisfaction” ratings for listings in New York that vanished from the site between November 2013 and May 2014.

The biggest single category of the 10,000 New York listings that have vanished from the Airbnb site are listings with no ratings at all (shown as NULL). After that come the 5.0 and 4.5 ratings, leaving fewer than 800 with bad ratings (4 or lower). For at least 1200 of the listings that Airbnb removed, bad ratings is not the reason.

Combined with the very high number of positive ratings shown in Figure 7, this result shows that the significance of peer-to-peer ratings is exaggerated for Airbnb. While the company proclaims that peer-to-peer rating systems set it apart from old-style methods, it is clear that their trust system is essentially a traditional complaint-based, centralized system. The peer-to-peer ratings may give warm feelings on the site, but Airbnb itself clearly does not trust it when it goes to remove listings from its site.


The data set suggests that there is a surprising amount of churn in Airbnb’s hosts. Over a third of the host population in November left the site, to be replaced by new listings.

It is likely that a large number of people experiment with being Airbnb hosts, and then decide that it is not for them.

It will be interesting to see whether this becomes an issue for Airbnb as the number of ratings it has in the North American cities surveyed in November and May seems to have peaked. How will it continue the growth that is demanded of it by its backers in the face of the large number of hosts who drop off the platform?

While much is made of the novel peer-to-peer nature of Airbnb’s reputation system, it appears that the company runs a complaint-based trust system that is not so different from other hospitality companies. In April Airbnb expelled 2,000 listings from its New York offerings, claiming that they did not offer a positive experience, but many of those listings had either no ratings at all or were rated highly (4.5 or 5) by their guests: it appears that Airbnb did not use its own ratings when judging who to remove from its platform.

Uber Drivers Earning $90K/year? More Evidence Needed

Last Tuesday taxi-disrupting tech company Uber posted on the company blog that “the median income on uberX is more than $90,000/year/driver in New York and more than $74,000/year/driver in San Francisco”. I don’t think their numbers add up, but first the story so far…

For the notoriously cheap taxi industry, those are some pretty sweet numbers, and they were hailed (hah!) by Matt McFarland of the Washington Post with the headline Uber’s remarkable growth could end the era of poorly paid cab drivers. Noting that “estimates of the typical cab driver’s salary hover around $30,000″, McFarland acknowledged that “Uber’s numbers don’t account for the costs a driver incurs to own and operate a vehicle. Still, the gap in compensation for providing similar services is astounding”.

As the story spread, many just ignored those pesky costs. CNBC led with “Uber’s $90K salary could disrupt the taxi business”. The New Orleans Times-Picayune headlined its story “Uber drivers in New York City earn more than $90,000 a year, newspaper reports” and claimed “The Median Income of an Uber Driver in NYC Is Nearly $100,000″. From the technology industry, CEO Mike Jones laid it out at Code Conference: “You’re qualified to drive a car, but not professionally doing it. Congratulations, boom, you’re making [a] $90,000-a-year average Uber salary.”

Among all this enthusiasm, a few voices did raise some questions. In Time Magazine Dan Kedmey emphasized the costs:

Unfortunately, the figure excludes many of the costs of running a business, including gas, insurance, parking, maintenance and repairs and the original sale or lease price of the car which can take some hefty bites out of the driver’s take home pay. It also measures a median income among a particularly dedicated set of drivers, logging a minimum of 40 hours a week and sometimes much longer hauls.

He asked Uber, but they shrugged their shoulders.

Just how much those costs eat away at a driver’s take home wages is not easily gauged, according to Uber spokesman, Lane Kasselman. They can vary depending on the age of vehicle, the density of app users in the city, how many hours the driver puts in and what sort of customer ratings the driver receives. And for now that data, Kasselman says, is proprietary.

At Mashable, Jason Abbruzzese asked whether the income is sustainable (Uber is looking to entice drivers and has deep pockets, after all), and at The Atlantic’s CityLab Eric Jaffe pointed to the Uber driver protest in San Francisco, where drivers claimed they “work for less than minimum wage” and asked how these stories could fit together.

Back in December, economists Felix Salmon (here) and Tim Worstall (of Forbes) had both had fingered the culprit for taxi drivers’ appalling incomes: the medallion system that many cities use means that medallion owners get to take the money. Now, in the wake of the new claims, Salmon asked Uber for details of these extra expenses and they actually sent him numbers (take that, Time Magazine!) showing that business expenses would be around $15K, so the drivers are still making twice the norm: not bad. Salmon deemed these numbers “reasonable and entirely intuitive”.

Uber, by the way, takes 20% of the fare, plus a $1 “safety fee”.

Now I’m no expert, but I thought I’d take a look at some reports into the taxi industry and see what I could find out about the Uber claims. The short version is this:

  • The medallion owners in some cities take roughly the same amount of the fare as Uber. They may be ripping off the drivers, but the lease costs that they charge don’t seem to be the main reason for the difference between the numbers. I think Salmon and Worstall have this wrong.
  • The Uber claims over car maintenance costs are under what other reports say about the costs. I suspect they are cherry-picked, but they are also not the main reason for the difference.
  • The big difference comes from the claim that an uberX car takes in $110,000 in fares over a year, while driving 40,000 miles. A regular taxi takes in about half that, and drives about 50% more. It’s not clear where this difference comes from (different cities? different ways of counting? bad guesses?) but if Uber is going to stick by its claim, it needs to explain the difference.
  • The media writers who take business expenses as a minor factor in the driver’s overall income are way off.
  • Put this all together, and the driver incomes look too high.

First let’s look at the plight of taxi drivers. I found relatively recent reports on the taxi industry in three major North American cities: a UCLA study on Los Angeles (2006), a San Diego State University report on San Diego (2012), and two reports about Toronto (2008, 2012). Obviously these are not San Francisco and New York, which is what Uber was writing about, but the point is not to ask if they are telling the exact truth, but to see if the picture they paint is a representative one.

The three reports paint a grim picture of a taxi driver’s life. In all three cities, most drivers work six 12-hour shifts a week for less than minimum wage, even after tips. These are mainly immigrant men, and most are between 30 and 50 so many have family responsibilities. Health insurance is non-existent, and the job is dangerous with assault, vomit (which they have to pay to get cleaned) and petty aggravations as perpetual companions. I don’t think anyone wants to paint a rosy picture of the taxi driver’s working conditions.

But now to the numbers. There is a surprising consistency to the three cities (which I write as LA, SD, and TO).

  • In their 72 hours of driving each week, a driver will cover about 1200 miles (LA), of which just under half (LA) are “on the meter”. The LA study gives an average of 6.2 miles-per-paid-gallon.
  • The total income (fares + tips) that comes into the cab over a week is about $1500 (LA), $1100 (SD), or $1150 (TO).
  • Gas is a significant expense, and cab drivers have to pay it. The weekly cost is about $250 (LA), $260 (SD), and $250 (TO). This suggests that San Diego drivers cover the same distance as LA, while Toronto (where gas is more expensive) cover significantly less.
  • Most taxi drivers pay a lease that covers car insurance and maintenance as well as earning money for the medallion owner. The weekly lease is about $500 (LA), $400 (SD), and $260 (TO). There’s quite a bit of variation in each city because there is a mixture of owner-operators, shift drivers, and others at work and their situations are all different.
  • The Toronto report separates the car maintenance, depreciation, repair and insurance out. For LA and SD this will come out of the medallion owner’s income. It is about $300/week ($70 insurance, $70 maintenance and repairs, $175 on car financing).
  • That leave a weekly income for the driver of about $600 (LA), $320 (SD), and $450 (TO). This comes to an annual income of $31,000 (LA), and a terrible $16,60 (SD) and $23,400 (TO).

We can put these in a table and compare them to the Uber estimates:

[Update: changed these figures on June 3 to be better averages of the driver kinds in each city.]

Quantity (weekly) LA SD TO Uber (est)
Fares + tips 1500 1100 1150 2070
Gas 250 260 250 115
Lease/Operator 250 140 260 414
Car Operation & Depreciation 300* 300* 300 76
Other expenses 100 70 90 0
Driver Income 600 330 250 1465

* = uses Toronto estimate, as no estimate given in report. Is paid by the lease-holder for those who lease.

Do the Uber numbers make sense? At the very least they need some explaining before we take them seriously. Here are the questions that need answering:

  • According to the Washington Post, Uber’s sample is “drivers working over 40 hours per week”. Are they working the same 72 hour weeks that taxi drivers are?
  • Taxi mileage is way higher than Uber’ estimate of 770 miles per week. Maybe Uber is not counting the “off meter” miles?
  • Can they justify the low gas costs that they estimate or are they not counting time between rides (which is half the mileage for taxi drivers)?
  • Uber seems to take more than the leaseholder (after expenses). Given the slagging off that medallion holders get, this surprised me.
  • The car maintenance fee for Uber is much lower than for taxi drivers. Does this reflect a lower standard for Uber or where do they get this number?

In short, a lot of questions to be answered before Uber’s claims can be justified.

The shape of Airbnb’s business

When Airbnb talks about its legal troubles in New York, Berlin, Amsterdam, and elsewhere, it claims that existing laws were never designed for its new brand of disruptive peer-to-peer business.

There were laws created for businesses, and there were laws for people. What the sharing economy did was create a third category: people as businesses… They don’t know whether to bucket our activity as person or a business.

In 2010, the State of New York passed a law designed to crack down on bad actors that operate illegal hotels—a goal we all share. Unfortunately, the 2010 law also had the unintended consequence of impacting regular New Yorkers.

There are more. You get the point.

In a series of studies designed to address regulators’ concerns, Airbnb talks about its hosts as “regular people” and focuses on the ways its business is different to the existing tourist business. It highlights the hotel industry as a point of comparison and emphasizes just how different Airbnb is. Here are a few typical quotations:

  • “87 percent of hosts rent the homes they live in” (Amsterdam)
  • “87 percent of Airbnb hosts rent out the home they live in, and the typical host earns $7,530 per year” (in New York)
  • “Airbnb is complementary to the existing tourism industry in Paris. 70 percent of Airbnb properties in Paris are located outside the central hotel corridor.”
  • “73 percent of Airbnb properties in Amsterdam are located outside the eight central tourist districts.”
  • “Airbnb’s 5,600 local hosts are regular people who occasionally rent out their homes and use the income they earn to pay the bills.” (in Berlin)
  • “About 80% of Airbnb hosts rent out the home they live in” (in London and Edinburgh)

There are more. You get the point.

The claim of novelty and of hosts as “regular people” has been widely accepted. For example, the thoughtful Kevin Roose wrote this in New York magazine the other day:

There are no laws governing Airbnb because until very recently, there was nothing like Airbnb in the world—not of the same scale, not with the same guiding philosophy. And when Airbnb came onto the scene, regulators were forced to slot it into existing categories where it, arguably, didn’t belong—treating a bachelor renting out his spare room to make rent, for example, with the same rules as a scuzzy landlord operating an illegal hotel. They’ve been playing catch-up ever since.

Or as Wired Magazine writes:

We are hopping into strangers’ cars (Lyft, Sidecar, Uber), welcoming them into our spare rooms (Airbnb), dropping our dogs off at their houses (DogVacay, Rover), and eating food in their dining rooms (Feastly).

There are more. You get the point.

So when it comes to thinking about and dealing with this and other sharing-economy companies, the kind of business that Airbnb operates matters. Does it match the company’s self-portrait? Is the company as novel as it claims to be? Are its hosts “regular people”?

It turns out there is an element of wishful thinking in the portraits of Airbnb. Last November, I took a look at Airbnb data from New York (here and here). In February, travel web site Skift carried out a similar study (here and here), which was part of the New York Attorney General’s case against Airbnb. The two studies took similar approaches, collecting listings from Airbnb’s public web site and gleaning what we could from that imperfect data set. Both studies concluded that, while it is true that a large number of hosts rent the homes they live in, hosts with multiple listings make up almost half of Airbnb’s business. Also that, while Airbnb makes great play of its origins in renting out an airbed, such rentals are now a negligible portion of its business. Even “spare rooms” are a minority of the business: the majority of Airbnb’s business in New York comes from the rental of entire homes.

The data showed a company that was closer to orthodox models such as HomeAway and its subsidiary VRBO than the narrative would have it. There are differences—HomeAway is focused on vacation rentals, and many of its properties are run by property managers—but the similarities cast doubt on Airbnb’s claims that existing regulations are inapplicable.

Now here we are: it’s six months on, and interest in Airbnb continues. Airbnb has kicked 2,000 New York listings off its site (10% of the total for the city). It handed over host data to the Attorney General (anonymized, the company says). Meanwhile, the company is valued at $10 billion, having raised $450 million in a new round of venture capital. The New York dispute is now over, but the sharing economy poster child is still here, bigger than ever, and still a leading light in the wave of digital disruptors looking to shake things up and make a lot of money.

So during May I collected data on over 90,000 hosts and 125,000 listings—about 20% of Airbnb’s 600,000 total, according to this TechCrunch estimate—from 18 cities around the world, to sketch a portrait of Airbnb’s business. The main questions I had in mind are the straightforward ones, starting with the same ones Skift and I asked about New York:

  • Is Airbnb’s business based on “regular people” in a way that other part of the hospitality industry are not?
  • Is Airbnb’s business based on spare rooms and airbeds?
  • I looked again at several cities I had collected (but not posted about) in November, so that I could look at how the business has changed in some of Airbnb’s key markets.
  • I hoped that looking a second time at New York might have something to say about the 2,000 listings that Airbnb removed from the site in April, during its run-in with the Attorney General.

For those who don’t want to read the whole thing, here are the quick answers.

  • While a good part of Airbnb’s business is based on “regular people”, over 40% comes from hosts with multiple listings. This is different from Airbnb’s self-portrait. Airbnb’s claim that existing regulations don’t apply to it is at least exaggerated.
  • The majority of Airbnb’s revenue comes from whole-home rentals. This makes the company much more like HomeAway and other vacation rental businesses. It casts further doubt on the company’s claim to be a new class of business.
  • In some of its biggest markets, Airbnb may have maxed out the number of listings it can achieve. What’s more, there is a high rate of churn as individual hosts put a property on the market, have a few guests, and then take the property off again.
  • Airbnb does not appear to believe its own claim that customer ratings provide an assurance of good experience. Airbnb says it removed 2,000 New York listings from the site because of bad experience, but at least half of those listings had good (4.5 or 5 star) average ratings from customers.

I’m going to post this in two parts.

  • Part 1 looks at the split between hosts with a single listing and those with multiple listings, and it also looks at how far Airbnb has moved from its “origin myth”, after which the company is named—the hosting of people on couches and in shared rooms.
  • Part II looks at the change in Airbnb’s business over the last six months, including the changes in New York where the legal strife has been the loudest.
  • If I have time and if there are requests, I’ll collect these three together and post a single PDF.

Background and Method

Just to set out some basic information, Figure 1 shows the number of listings and of hosts in each city. The data was collected using a fairly straightforward two-stage search. The first stage goes through all the search pages for a specified city and collects the room_id values. The second stage visits the room page for each value and gets details about the listing. The code is available on github.

A few notes:

  • Airbnb has regularly said that is has about 20,000 listings in New York city. I find 19094 (and have over 20,000 from November) so the collection seems pretty complete.
  • In 2013, Airbnb claimed 5600 hosts in Berlin. I find 6141.
  • In the run-up to the World Cup, Airbnb claims to have 9,000 listings in Rio. I find just over 10,000. Perhaps the borders of the search are different, or the number is growing. Still, 90% accuracy is not bad.

In short, the surveys for each city seem pretty accurate.


Figure 1: Number of hosts and listings in the surveyed cities

The host perspective

Figure 2 shows the percentage of hosts in each city that have a single listing on the site. The values for New York and Amsterdam match Airbnb’s claim for New York and Amsterdam to within a percentage point, which suggests that the sample is realistic and that the use of a single listing is a pretty good proxy for “regular people who occasionally rent out the home in which they live”. The graph also shows that the claim applies to most of the big cities. Barcelona, Rome, and Tokyo are the only Airbnb locations surveyed that have fewer than 80% of hosts with a single listing. So far, so good for Airbnb’s self-portrait. From here on I will call hosts who have a single listing “regular people”.


Figure 2: Percent of hosts with a single listing

The marketplace perspective

Imagine that a city has 100 hosts, 99 are “regular people” with a single listing and one host has 99 listings, then the percent of hosts who are regular people would be 99%, but the percent of listings on the market that come from regular hosts would be only 50%. Both percentages are important in gauging the kind of business that Airbnb is. Figure 3 shows the percent of listings offered by regular people. The overall figure is 62%: still a significant majority, but a number that is 20% lower than the percentage of hosts.

We can see that for a few cities, notably Barcelona (11,000 listings) and Rome (growing quickly, at 8,000 listings), the majority of listings come from hosts with more than one offering.


Figure 3: Percent of listings from hosts with a single listing

The traveller perspective

The percentage of listings that come from different types of hosts corresponds to the experience of the potential guest browsing or searching the Airbnb site. The traffic generated by Airbnb is different, because not all listings are equally popular. There may be areas with many Airbnb listings but relatively few actual visits, while other areas may have listings that are visited very frequently. Airbnb does not give the number of actual bookings for each listing (or, nearly equivalently, the number of visits to the listing), but it does give the number of reviews that each listing has received, and this should be a reasonable proxy for the number of visits.

Over a third of all listings have no reviews at all, and some have many (the most-reviewed listing in my sample is this San Francisco treehouse, a novel place to stay with 460 reviews.)

Figure 4 shows the percentage of visits that are to rooms listed by regular people. The numbers are getting significantly smaller now. No city has more than three quarters of its visits at properties of regular people, and several have a slim majority of visits to hosts with multiple listings. Overall, 45% of visits happen at places offered by hosts with multiple listings.


Figure 4: Percentage of visits (bookings) to listings offered by hosts with a single listing, using reviews as a proxy for visits

The Airbnb perspective

There is one more step to take, which is to look at Airbnb’s actual revenue. What fraction of its business comes from regular people and what fraction comes from multiple listers?

To get here, I multiply the bookings (proxied by reviews) with the listed nightly price. Again, it’s not a perfect measure but so long as regular hosts don’t overall have longer- or shorter-stay guests compared to other hosts, it should give a reasonable picture.

Figure 5 shows the percentage of Airbnb’s revenue that comes from regular cities. Only three cities have over 60% of their revenue coming from hosts with a single listing. Overall, 44% of Airbnb’s business comes from hosts with more than one listing, which is slightly up from 42% in November (the increase may not be significant).


Figure 5: Estimated percentage of Airbnb revenue from hosts with single-listings

Airbnb listing types

The story of Airbnb emphasizes the casual “airbed” rental, but this is a very small part of Airbnb’s business. Again, there are a couple of ways of looking at the data.

Listings by room type

Every listing on the Airbnb site is listed as one of three categories: a private room, a shared room, or an entire home/apartment. Figure 6 shows the breakdown in each city. It is clear that shared rooms are a negligible portion of the total: about 1 in 50 listings are shared rooms.


Figure 6: Number of listings that are private rooms, entire homes or shared rooms

Visits by room type

Figure 7 shows the visits by room type, using reviews as a proxy for visits again. Shared rooms are an even smaller percentage of the whole: only 1.4% of Airbnb visits are made to shared rooms.


Figure 7: Number of visits to private rooms, entire homes or shared rooms

Revenue by room type

Using reviews * price as a proxy, Figure 8 shows the percentage of revenue from each room type. The revenue that comes from spare couches and shared rooms is a mere 0.56%.


Figure 8: Revenue from different types.

So what?

  • The data show that Airbnb is consistently economical with the truth when it describes its own business. It’s a long way from being just “regular people” and there is a lot more business coming from multiple-listing owners than they let on.
  • The data say that it’s time commentators and the media stopped using the”Couchsurfing” narrative for Airbnb. At less than 1% of its business, the couchsurfing model is irrelevant for what Airbnb is today. It’s far more like HomeAway than it is like Couchsurfing.
  • If there is a novelty to Airbnb’s business, it’s that it collects property managers, individual renters, and occasional renters under one roof, just as Amazon can offer best sellers, midlist authors and self-published obscura in the same place. And while much of the talk will be about the long tail of rooms, the reality is that Airbnb is pushing for the professionalization of its hosts, and we’ll see more of that over time.
  • Other conclusions to come after we see the rest of the data.