Landlords use software to set rental rates | Hacker News Reader

Landlords use software to set rental rates

nytimes.com

21 pointsolegious5 comments

So how long till we see the same sort of "software triggering responses in other software" patterns that've been found in the stock market from HFT trading, in the apartment rental market? Will we one day see entire city blocks empty of residents, because all the different pieces of rental optimization software are raising prices in response to each other on Craigslist, and all the actual humans ns go live somewhere else?

I doubt it would work out like that. The market (especially HFT) is usually one investment fund software trading against another investment fund software. Humans are rarely involved but do oversee the thing, sometimes.

Renters are usually constrained by income and location, if the price goes too high people will make other choices (buy, move further, or squat). Also, as opposed to investments, properties need to be lived in. An abandoned apartment, even in just a few months, will have water damage, weathering, pests and vermin, and theft. And for the moment, I believe, the US has a 1-2 year backlog of empty houses.

So this looks to be a tool for yield management. Simply put, you manage a fixed production capacity to maximize income: sport and concert tickets, museums, airlines, cinemas, all work like this. The unsold places are "wasted". I could take out the quotes, because in comparison with a storable product, you cannot warehouse cinema or museum hours, so they're really wasted. A really smart tool would know your costs and work to increase profit, not just income.

One could argue web apps could work like this as well. There is fixed capacity (how many users could your server farm handle) that is wasted if underused. This fixed capacity is fuzzy, of course, so defining it is the first difficulty, Then, it's much harder to apply price discrimination for software. People feel cheated if they find out someone paid less for the same thing. There are obvious differentiation mechanisms in real estate that prevent this. These are missing in software. One could try to increase price variation by using time-limited rebates, for example around a holiday, in addition to the classic multi-tier pricing.

Interested in hearing what patio11 has to say about this.

A major difference is that cinema owners, airlines, sports teams etc. aim to maximize their users, and each additional user doesn't diminish the experience of other users (well to some extent, but not enough to make a huge difference to the demand).

A lot of spare capacity in server farms seems to be due to the need to cater for demand spikes, so that the experience doesn't degrade. Each additional user will, after a certain point, harm the experience of the previous users. A stronger analogy would be a power utility.

It would be quite difficult, I think, to do what you propose without being able to predict demand spikes well in advance and with a high degree of accuracy. Even if we could do that there's another issue: a power utlility can sell power to certain types of consumers at off peak times to soak up spare capacity, but electricity has many uses. A web-app isn't likely to have many non-core uses that would appeal to specific users at off-peak times.

Cloud providers can use this sort of strategy, but I don't think it would be useful for individual web apps, unless you are selling something very generic. The best way to deal with this issue, from the perspective of an individual web app, if it bothers one, would probably be simply to use a cloud solution, and let them deal with the excess capacity issues.

The article mentions in passing that similar tech is used to price airline tickets, but the important difference is that airline tickets are now almost universally purchased through sites like Kayak or Travelocity, which allows customers to quickly and easily compare similar trips and make a choice (generally on price). This has resulted in airline ticket prices plummeting in real terms over the past 10 years or so. In apartments, on the other hand, this tech seems to give more power to landlords, unless some smart startup builds a Kayak-for-apartments.

Reply on news.ycombinator.com