How can we discover homes before they hit the market?
In +Choosing a location for the Neighorhood, I chose five approximately good locations and analyzed weather, crime, transit, centrality, nice houses, and walkability. 

Then it became abundantly clear that housing was going to be the bottleneck. How will we find homes that are large enough for coliving (4 or more bedrooms and 2K+ sqft), that are nice enough and in the right price range (between $2.5-4K per bedroom), and how often will they become available? And can we have nice apartment buildings nearby, too?

Viability of the Neighborhood

Zillow shows very few properties with all of these qualities right now - I put some screenshots in Appendix 2: what does Zillow look like right now? So how are we supposed to fill a (desirable) square mile with hundreds of friends and acquaintances? Here are two strategies:

  1. We’ll coordinate investors to purchase homes and rent them to us
I can build a list of hundreds or thousands of real estate investors that have already invested in large, luxury homes in San Francisco and rented them out, and ask them if they want to buy a specific property and rent it to us. An investor I talked to about this said such an offer would be “music to his ears”. Of course, we could own homes ourselves, too. This is described in more detail in the +The BYO Landlord idea. This is important because many of the nicest homes are only for sale, not rent, and many of our new residents will prefer to rent, at least in the beginning.

2. We’ll discover homes before they hit the market
Arram found 283 Dolores this way by asking the owner of an unrelated property if they had any other large properties. I can do the same at scale. I’ve got a list of hundreds of large, nice homes and know how to find the email addresses for most of their owners. If I email those owners and offer to rent/buy them once they’re vacant, I bet we can win some of them before they ever hit the market. 

Here’s my proof-of-concept spreadsheet for this strategy, which I combine public tax assessor records (which contain addresses, square footage, and numbers of bedrooms), Zillow information that I scraped (which contains current prices or price estimates), and ownership information (scraped from a tool a realtor friend shared with me). 

Locations considered

All areas are about one square mile.

Pac Heights, based around Alta Plaza, Lafayette Park, and Fillmore St. Fillmore St between Jackson and Bush is charming: it’s a lightly-trafficked two-lane street, dense with greenery and bougie shopping and restaurants. Pac Heights is full of large, beautiful homes with a far higher price ceiling than the other two locations.

Alamo Square. It’s two blocks from the Embassy, Genesis, and the Center, all of which throw regular fantastic events. It’s six blocks from Duboce Park, the Panhandle, Buena Vista, and Duboce Park. Page St is a walking-only street and the lack of car noise is incredibly calming. Haight St and Divisadero St are cool in a slightly artsy/grunge way -- more like Valencia St than the yuppie 24th St in Noe. Alamo Square itself supports as many activities as Lafayette or Dolores; it has beautiful views, picnic tables, barbeques, a dog park, a playground, and a tennis court. There’s an undeniable energy and liveliness to Alamo Square - the question is whether we’d be able to get our hands on homes that are actually nice.

Dolores Heights. Dolores Park is beautiful, and the homes in the hills above it are too. Josh and Kanjun’s office is right in the middle of this map. This area’s south boundary is 24th St in Noe, which has a charming, small-town feel - the shops are DIY pottery studios, bakeries, and sports bars. Great views, too. Noe is significantly less invigorating than either of the other two areas, and the nearest downtown, Castro, is somewhat crime-ridden.

Availability of quality housing

The goal is to find homes that are large enough for coliving, that are nice enough and in the right price range (between $2.5-4K per bedroom), and regularly available.
  • large: how many houses have >= 4BR and >= 2K sqft?
  • nice: how many houses are between $2.5-4K?
  • I haven’t scraped all of Zillow to get reliable metrics for this yet (it’s also really difficult to estimate rent and sale prices, see: Zillow), but the proportion of Alamo Square houses that are affordable is about 42% of all large homes, so I’ll just use that ratio for every neighborhood (even though it’s bound to be misleading - Pac Heights homes are much larger, on average).
  • available: I predict annual turnover for the neighborhood and estimate how many of those freshly vacated houses we’d be able to win. 
  • Using Joe’s realtor tool, I can break down a group of buildings by owner vs non-owner occupied, which have annual turnover rates of 5-10% and 15-20% respectively, which means I can predict annual turnover (see Appendix 1: annual turnover rates).

I also model how many of these homes we could actually get, based on Rose’s estimated 20% response rate and 50% win rate for those that do respond (both WAGs).

Large
Large & nice
Annual turnover rate
Large, nice, & available
Homes per year we would want and can get (multiply by .1)
Coliving population after 2 years (avg size 8 ppl)
Coliving population after 5 years
Pac Heights
578
242.8
10.8%
26
2.6
42
104
Alamo Square
611
280
12.6%
35.3
3.5
70
175
Dolores Heights
645
270.9
10.9%
29.5
2.9
58
116
Outstanding questions:
  • Why, at the Archive, were we able to gain both floors above us and almost the house next door within a single 3 year period? That’s supposed to be extremely unlikely according to the table above, so I suspect this table is too pessimistic, perhaps by a lot.
  • What are the proportions of homes in Pac Heights and Dolores Heights that are affordable? That’s the big reason that each row looks kinda samey: I’m just multiplying the number of homes in that square mile by 42%. 
  • What does the actual sales funnel look like? When I email an owner or knock on a door, will my response rate really be 20%, and of those, will we successfully win/buy 50%?
  • What proportion of people in the Neighborhood will live in coliving houses vs not (in single-family homes, condos, apartments, etc)? This will have a big impact on our ultimate population.


Nearby apartments

Ideally, there will also be large luxury apartment buildings in the area. This would make it easier for anyone that doesn’t want to live in a big house to join the Neighborhood. They’d also be good “staging areas” if you wanted to move to the Neighborhood but didn’t want to commit to a single house, or you could take over a floor and turn that into a community, too. Apartment buildings also have nice amenities, like coworking spaces and pools, and they’re often underutilized.