AndCo Remixes Commercial Real Estate

Plus! Privatizing Emissions; The Put-Seller of Last Resort; Prayer-as-a-Service; (Voluntary) Crypto Regulation; Grinchbots

AndCo Remixes Commercial Real Estate

There’s a set of service industries which rely on very high fixed costs, and very low marginal costs. Demand for their services is fluctuating, and of course they still have to maintain the high fixed costs. Whenever there is volatile demand for businesses with high fixed costs, there’s room for a marketplace business in behind, to smooth the peaks and troughs of their demand cycle. This is most obvious in the airline industry, both at the level of a passenger, and at the level of the planes. At the level of the passenger, the airline is paying $X to fly a plane from A to B, and the difference in costs—whether seat 24E is filled or not—is equal to the marginal teaspoon of fuel.1 So an OTA that fills last minute demand can reduce the variance in flight-level demand. Similarly for the airlines, running an airline is mostly fixed costs: planes, staff, landing/takeoff slots and so the cost of having a plane in the air is (only) the cost of fuel. This means one way to think of the aircraft leasing companies ($) is as a marketplace between Boeing and Airbus, and airlines with volatile demand for their planes (in different areas of the globe). 

This is also true of real estate: the variable costs are things like cleaning a space for the next person who occupies it; the fixed cost of the structure itself dominates. One of the functions of Airbnb is to smooth out demand for residential real estate. It’s slightly weird to think of the demand for rooms in your house, but if someone else’s demand for a spare room in your house is greater than yours, listing it on Airbnb is one way of monetizing this. Renters of commercial real estate face the volatile dynamics more than most. They make commitments to multi-year leases, but demand on their space can fluctuate as much as hour-to-hour. Restaurants exemplify this: a restaurant big enough to be 100% at capacity during dinner on Friday and Saturday night also has enough space to be fairly empty on a weekday afternoon. The demand for real estate around stadiums spikes on game days, and drops off the rest of the time; Arena rock is a music genre, but it’s also a real estate play.

Up until recently in the co-working industry, WeWork and others have solved the problem of volatile demand by taking a long term lease, and splitting it into hour-long, day-long, and quarter-long slots, and charging a premium for this. These co-working spaces, like the airline leasing businesses, function as a market maker between the owners of commercial real estate and the individuals who demand small areas at different times.

These marketplace and market-maker businesses are all providing the same service, covered in the Diff previously ($), to shift volatility off the books of one person, aggregating it on their books (which makes it more predictable on a macro scale) and charging a premium for it. AndCo is a startup that connects knowledge workers looking for a place to work, to underutilized hospitality venues. In the process, they’ve reduced the price of traditional coworking by an order of magnitude—an annual membership is under $200.

Many businesses have a simple truth, which they exploit better than most: Google’s was that the link graph represented a way to measure the context and importance of every single URL, Facebook’s was that people’s default mode of online interaction would be based on their real-world identity, and Netflix’s was that movies are streams of bits and the regularly-moved bits tend to go over wires rather than physical media. AndCo’s simple truth that a set of service businesses are converging on the same bundle of services: desks, food, beds, perhaps a weight rack. To varying extents, this is true of: hotels, fancy gyms, coworking spaces, and airport lounges. Bundles are most often subsidised by one of the elements in the bundle, and the other services are provided as a complimentary complement. Hotels obviously lean most heavily on the beds, but still provide the other functions—a hotel customer is paying up for a set of known defaults in an unfamiliar city. The same bundling is true of the weights rack for fancy gyms, and the desks for co-working spaces. There’s definite economies to bundling “experience” services, but this kind of convergent evolution also produces inefficiencies. One example is the under-utilization of hotel lobbies and conference rooms. The founder of AndCo told me their research suggested occupancy rates as low as 6% for these shared spaces. AndCo unbundles this aspect of the hotel’s service by opening the space to knowledge workers. For no cost to the venue, AndCo is basically lead generation for their amenities. It’s mostly food and drink, where spending has doubled from the pre-Covid average spend of $20, but there’s also the option to upsell private meeting rooms. (The founders told me that there’s been some instances where customers have booked their visiting relatives into the hotel because of their experience working in the lobby.) Of course this model is applicable not just in hotels, but also restaurants, and pretty much any hospitality venue.

The ostensible competitor to AndCo is coffee shops, but it’s the wrong comparison to make. The shift to remote work might have looked for a while a professional free-for-all, but as it’s turned out, there are still going to be constraints imposed by bosses, colleagues, and customers. As such, remote workers still need certainty in the spaces they work, meaning reliable and secure internet connection, noise levels acceptable for calls, and a guaranteed desk. This is why AndCo’s app works by booking work spots and in each location they install their own hardware to create a secure WiFi network across venues. As we mentioned earlier, bundles often involve paying for one element and enjoying the others as complimentary complements. Coffee shops exemplify this: you pay to access WiFi and rent a table, one cup of coffee at a time. The problem for remote workers is there’s no recourse if the internet connection stops working, or there isn’t a table. Whereas when you pay directly for the service you want through AndCo, there’s a different set of expectations; it’s a coffee shop with an SLA.

A possible drawback to being asset-light is that you won’t be able to generate the same insights or data that a company which controls the space is able to. When WeWork filed their S-1, Byrne wrote that their utilization of space was 400% more efficient than traditional office spaces. The trouble with asset-light businesses is that they still need assets, they just don’t need them on the balance sheet—and when someone else owns them, someone else controls them. However, because AndCo installs the hardware at their venues, these sensors are able to provide all kinds of insights about how people are using the space.

With an asset-light business, it’s an interesting thought experiment to consider what you should do with the insights you generate. Running a marketplace will teach you a lot about running the businesses you work with, but to realize the gains of these insights, you'd have to start running the asset-heavy model you explicitly wanted to avoid. There’s a balance between a) keeping the insights completely within your company, and b) sharing them with venues on your platform, or c) making them freely available to everyone, in an effort to get more venues to join. For example, in the early days of Airbnb, the founders knew that the quality of images for an apartment was important to conversions, so they went to hosts’ houses and improved their photos. This initially fit into the category of “sharing insights with venues on the platform”, but now it has become Blitzscaling folklore and Airbnb promotes their network of 1,000 photographers so everyone is aware of the benefits to good photos.

One of the downsides of the asset-heavy model is that it keeps costs fairly fixed even when revenues drop. During the pandemic asset-heavy companies were committed to larger payrolls and multi-year leases which marked the death of many real estate startups like Breather ($), but it was validation for the AndCo business model. When the company began in 2017, for most people it solved the problem of: “I’m in an unfamiliar part of London, where can I work without needing to wander around looking for a suitable coffee shop?” but through the pandemic this has morphed into “Where can I do my meetings today?” When the office got diced up and redistributed to kitchen tables, second bedrooms, and coffee shops, it benefited companies that could indirectly supply this new form of office real estate. Getting through that transitory period was significantly easier when retrenchment meant collecting their hardware from venues, instead of defaulting on leases and relying more heavily on furlough schemes, which asset-heavy companies would be forced into. Going forward certain types of real estate will always be intended for fluctuating use and AndCo will reduce the troughs, for whatever hospitality venues might need it. They’re also making remote-work-friendliness a portable trait, by installing their hardware at businesses with spare capacity in high-demand locations. Essentially, the company is a tacit benefit of dense cities—coffee shops and hotel lobbies as floating desks for remote work—and making it an explicit product with a price tag attached.

Thanks to Sanj Mahal and Tom Wordie of AndCo for talking to me about their business. If you’re London-based, join their service at, but anywhere else they’ll be expanding soon.


Privatizing Emissions

Mining giant Vale has sold its Moatize coal mine to a private company for $270m, about $2.1bn short of what the mine's value on Vale's books as of 2016, when they took a writedown that wiped the mine's value out. They project that the mine will produce 18m metric tons of coal in 2022, up from 9m in 2020. So Vale has gotten a problem off its books, at a massive cost, but the mine itself will produce a growing quantity of pollutants. In fact, this sale arguably increases emissions, since the private company’s shareholders clearly object to coal less than Vale’s do.

One interesting question to ask is: if there are so many distressed sellers, but there's still demand for coal, will the real money in ESG be made by people who bet against it? It's unclear what Moatize's costs look like, but applying their projected production to current coal prices implies that it's potentially a $2bn+/year business. In real terms, Henry Clay Frick will probably always be the person to get richest from coal, but in nominal terms it's entirely possible that deals like this will mint a few fortunes.

The Put-Seller of Last Resort

Turkey has a new plan to deal with high inflation: offering consumers savings vehicles that have a guaranteed interest rate and whose value will be topped up by the central bank if the lira falls. In other words, lira depositors who lock up their currency will get a free put option on it. In one sense it's natural that this would slow the currency's decline: writing puts is equivalent to buying, and prices are set by supply and demand. Meanwhile, locking up deposits means reducing the circulating supply of the currency, which controls inflation. On the other hand, it's high risk; one reason Turkey's currency has dropped is that there's little faith that the government can maintain its value, and by selling puts (for free!) the government is doubling down on this: implicitly, they're borrowing other currencies to buy their own, a tactic that can work temporarily but that will also make further declines more costly to the government.


The WSJ highlights the trend of higher funding for religious apps ($), which raised $175m in funding in 2021 compared to $49m last year. Religions don't use the same language, but they have been mastering network effects, flywheels, and retention for longer than other human institutions, so mapping some of that to software products can make sense. There is, naturally, a debate about whether or not it's proper to achieve unicorn status by productizing religion, but religions have been finding ways to interact with the secular world since the beginning, and since the big religions have outlasted all previous economic arrangements, the right bet is that they'll figure something out.

(Voluntary) Crypto Regulation

ISDA, which among other things creates templates for derivatives, is working on standardized contracts for crypto that will take into account quirks of the industry ($, FT). Corporate actions can sometimes create some interesting problems for derivative structuring; Matt Levine has written frequently on the topic of CDS contracts and manufactured defaults, for example. The crypto twists are different—a fork is a bit like a spinoff, except that it can happen without permission. And enumerating all these details is important for derivatives contracts, because one of the cleanest arbitrages around is to buy something priced as what it looks like in order to own something whose payoff is based on what it is like.


I wrote almost exactly a year ago about "grinchbots", programs that automatically buy goods that are in short supply. They're back in the news ($, FT); stores are using technical countermeasures, but the secondary market has only gotten bigger, so sniping deals for barely-in-stock items is still valuable. This may be one reason Walmart made the PS5 and Xbox available to Walmart Plus members, not the general public: it's a way to ration the supply and deliver more of it to end customers rather than resellers. Grinchbots are not the main reason stores are trying to add membership models and collect more first-party data, but they're another force pushing in that direction. Much of the world's most valuable ad targeting data has been owned by advertising companies rather than advertisers, but the shape of the market is changing in a way that makes that less tenable.

  1. The price for seat 24E right before takeoff would theoretically be just above the cost of this fuel. However, practically, it would be higher than this, because selling airline seats is an iterated game; if consumers believe they can wait until the last minute and still be able to book seats, they will do that. This involves a prisoners dilemma, where a critical mass of  the consumers would have to believe this. One way of thinking about airline pricing is that you’re paying $X to cover the airline’s costs, and then $Y for the certainty of  travel when you want to. If you’re time rich, but cash-strapped, you can save money by travelling at a time on the airline’s terms, whereas business travel is the opposite.