Creating a Frictionless Rental Housing Market Using Shared Housing 

Why Does Short-Term Housing Suck Right Now?The short-term housing market is one of the most undervalued markets by both investors and entrepreneurs alike. However, the primary misconception is the perceived trade-off between long-term home ownership and increased demand for frictionless short-term rentals as demographics start to skew towards millennials and Gen Z. Nearly, 85% of millennials still plan on buying a house within the next 5 years in the United States, but at a slightly later age (32.8) than generations before them. So beyond broad macroeconomic factors dictating the overall condition of the housing market, the two markets are probably less of substitutes than you would think. There are big factors that are dictating 

The problems between the two markets (long-term and short-term) are also vastly different. For long-term housing, the discovery and initial buying process are really clear on the consumer’s end but the first problems arise from the excess time spent on completing the process because of inefficient intermediaries (closing, title insurance, appraisals, servicing) which by default also shifts unnecessary end costs on the consumer. The biggest problem points for consumers in residential largely come after the transaction in the form of consequences from uncertain macroeconomic conditions (for refinancing, declining home equity, completing mortgage payments etc.). 

For short-term rentals (i.e. anything <18 months) the problems start immediately at the beginning from discovery and continue throughout the course of a lease. To highlight the main problems from the perspectives of the consumer and the lender (aka landlord): 

Consumer  

  1. No Single Shared Data Source: The problem with short-term housing starts at the very beginning with the lack of a single trusted source of data, which greatly limits liquidity and accurate price estimation. For long-term housing, the local MLSs serve as a single source of data which each platform (Zillow, Redfin, Realtor etc. builds on top off. Each of these platforms also has “non-MLS” listings proprietary to the platform, which increases the value proposition of each individual distribution channel (Zillow has by far the highest concentration). However, this fundamental shared source of data gives the consumer a clear estimate from an aggregation of the platforms gives a clear narrow price range. For short-term housing not only is there not a shared source of data, there isn’t even a single data-driven distribution channel that serves as an implicit price estimator. Zillow and Craigslist both set prices at the command of the owner, and the consumer ends up guessing what the relative value of their rental is!

  2. Lack of Buyer Mobility: There’s a fundamental disconnect between renters, who want increased flexibility and lenders, who want a guarantee of income over the course of over at least a year. Essentially, buyers are looking to sign shorter leases and move easily while the current model works relatively well for lenders.

  3. Urban Density in Evolving Markets: This isn’t necessarily even a solvable problem outside of government building policy but there is clearly just a huge net surplus of rental demand over supply in markets such as SF and NYC, the implication of which is that there will never be a perfectly efficient housing market in these areas no matter what type of tech solutions are implemented. And in a way this is on

  4. Transitioning Roommates/Community: This addresses more of the community aspect of shared housing. From the entry-level

  5. Lack of Innovative Financing Models: The financing of rent and short-term living options is fundamentally broken, which also piggy backs on the lack of buyer and seller mobility. In the residential market, there are many different ways of financing your home given that you are above a baseline credit level, especially in a strong housing market. First of all, the mortgage in itself a basic lending model that doesn’t exist for rentals. As a qualified buyer, you have a ton of good options for mortgages (brokers, direct bank lenders, non bank lenders like Quicken Loans). Even as a non-qualified lender there are decent options (correspondent lenders, FHA, HAMP), though declining home equity in down markets remain a strong concern.

    And for landlords, the financing requirements such as upfront payments (i.e. first + last month) and proof of income are totally necessary, nor should the burden of financing models be on them. But for a majority of entry-level renters with limited credit history, there’s very few flexible financial models besides credit cards and bureaucratic personal bank loans that allow them to rent on the basis of future earning potential or personal network based crowdsourcing. This disproportionately affects entry-level low-income and minority workers and discourages them from pursuing opportunities in high-demand, high-outcome markets such as NYC or SF. Homebrew portfolio company Pinch had a start on this by attempting to use rent as means for establishing credit, but there’s a ton of other innovation left in this space.

Lender

  1. Inefficient Property Management: There are too man

  2. Disjointed Documentation: This ties into the point below (#5) on qualified leads, but there is a lack of automation in compiling the documentation for potential applicants. Lenders have to manually compile credit history, proof of income, lease signatures on the applicant-end and then have to worry about insurance, appraisal, repair forms on the end. There’s startups that capture part of each of these processes but it’s actually the rare instance of startups not being ambitious enough in their value capture. There probably needs to be an one step process that encompasses everything for lenders almost as Blend does for mortgage originators with help from user input.

  3. Fragmented Distribution: The lack of a single distribution channel is actually a detriment to lenders as well. While

  4. Static Market Pricing: With increased mobility for buyers, there has to be incentives for landlords to be able to set the short-term price based on fluctuating demand. In residential properties, real estate agents serve as an intermediary for gauging total demand and setting a price based on years of experience and data. But in rentals, the price is incredibly arbitrary. Some tools such as Zillow will put together a prediction price upon committed listing, but the thing is that estimate comes from other arbitrary or outdated pricing. Large apartment complexes that are controlled by a single lender will occasionally do this, but often prices are static. Having dynamic prices (i.e. each individual gets the same rate for their duration, but changes when current tenant leaves) will allows landlords to be more flexible on the length of stay of customers. This is especially true for the estimated 40-50% of landlords who don’t use a large property manager.

  5. Qualified Leads: Lenders get tons of applications from applicants who might not be qualified to rent, try to negotiate, come to showings and turn out to have red flags deep in the process, which ends up wasting a ton of time. There needs to be an accessible CRM for lenders that uses Machine Learning and personal data/credit aggregation to auto-filter bad applications in an easy way. Because of the fragmentation of rental distribution channels, something like an embedabble API probably provides the best solution.

Shared-Housing Models 

So why would people hypothetically flock to shared-housing models? The ideal model has to have not only community incentives, but particularly for those with lower incomes and different demographics, provide strong economic incentives through transparent pricing throughout the rental as well as easy mobility. 

  • Startup Buys Urban Real Estate at Face Value, Provides The Community: This is the most basic yet most popular shared-space model. For a small premium, you get a clean room, a community of people your age, and you don’t have to go far to search to find the place. The value prop here is simple: we have a brand name of a guaranteed decently spot, there’s excess demand in these markets so we’ll always have customers, you won’t be lonely because we provide community services that augment our living space.
    Examples: Common, Ollie, Homeshare, Starcity

    Flaws:

    • Winner Take-All/Perfectly Competitive Dynamics: What’s the defining advantage besides input capital? Community matters, but it’s hard to build a value add that’s much stronger than another shared-living building, which probably makes the differentiating factor price (which is not good).

    • Unsustainable Upfront Costs in Crisis: Right now some of these startups are working with developers to build the holdings and taking a fee to maintain/manage. But at some point the developers will

  • Startup Identifies, Buys, and Renovates Undervalued Assets, Builds The Initial Community: This is similar to the model laid out in the point above but taking a page from Opendoor and creating the framework for a real competitive advantage. Instead of buying based co-living spaces based on location or perceived advantage, the startup builds a distribution channel of listings or attempts to pay a premium for a limited subset of existing price data from a company like Zillow. Then the startup starts to use their distribution channel to measure the demand for certain community value adds and locations within a city. This startup then completely buys (instead of working with a developer) the undervalued building, refurnishes it and recruits the early residents. At a certain point, the company has the data system that allows them to arbitrage demand and undervalued assets, know the exact ROI on certain improvements in the space, and even sell back properties to third-parties once it’s off the ground, given there is a strong AVM that can set an appropriate price.

    Examples: None

    Flaws: 

    • Capital Intensive: Goes without saying but the startup needs to pay for a strong data science team, find a way to gather an initial data set, manage inventory, conversion of an existing building and hire employees to help create the consumer facing product and acquire customers. Sounds costly, even for proof of concept.

    • Inventory Modeling: In some way you’re essentially a real estate developer that’s disguised as a tech company because you constantly have to keep track of the value of your properties and how they fit into your entire portfolio given your balance sheet.

    • Initial Data Advantage: How does a company go about getting an initial amount of data to measure the demand for certain areas? There’s no centralized source of rental data so there would have to be some unique ways about going to find this or through the buying of data from an existing site.

  • Startup Leverages Financial Identity and Community, Controls Full-Stack Process: As mentioned, one of the big discrepancies in the market is the lack of innovative financing models for short-term housing. And in a way, I would argue that the economics of shared housing are what’s making it succeed; the comfort for shared housing specific buildings allows developers to build smaller rooms and less shared spaces, which allows for more capacity and in turn more revenue.

    But as mentioned, one of the biggest pain points in lending is young people with little credit history and a lot of student loans being excluded from the system. There is probably an opportunity where the housing startup serves as almost a bank for young person with flexible payment plans on housing and using rent as a driver for the person’s credit history. Even something like a Lambda School model, where the person pays back a committed percentage of their income at a premium as rent, would be pretty innovative. In addition, partnerships with different restaurants and suppliers (i.e. for essentials) could help extend the “friendly credit” model further. At some point, the underlying anonymized user data could allow for the complete elimination of a premium on rent and be used to build a different type of customer lending platform or sold to an existing distribution or credit platform. In essence it’s the data -> free model that could be used to give more non-wealthy people the opportunities they need.

  • Startup Serves As Intermediary for Crowdfunding of Resident Interest: This is based on the popular German housing model of baugruppen. In this model, a group of future residents comes together to collectively finance, build and design their future homes. There are some heavy benefits; there’s money saved from billing a developer, the designs are custom to each resident and there’s a community built before the residents even move in. With the residents being the owners, there’s also no premium that’s billed to the landlord; the rent prices are breakeven, designed to suit residents. An article that explains how a variation of this might work America

    Examples: Almenr

    Flaws:

    • er a place to keep your kids under the same umbrella. From having a fully-integrated view, there are many data advantages from having a comprehensive understanding of your current and potential customers on a macro-level that allow for cheaper customer acquisition, high levels of personalization on a micro-level, and extremely high switching costs. This is pretty much just a description of WeWork if you’ve followed so far.

      Examples: WeWork 

      Flaws:

  • Startup Integrates Your Entire Life: This third-party startup aims to control your entire life, from a good place to work, to where you live, and when you grow older a place to keep your kids under the same umbrella. From having a fully-integrated view, there are many data advantages from having a comprehensive understanding of your current and potential customers on a macro-level that allow for cheaper customer acquisition, high levels of personalization on a micro-level, and extremely high switching costs. This is pretty much just a description of WeWork if you’ve followed so far.

    Examples: WeWork 

    Flaws:

  • Shared Corporate Housing on Steroids: Essentially, there exists an opportunity for certain startups/corporations to pool housing options on behalf of their employees. For example, Facebook and Google are starting to invest in temporary pre-fabricated housing for their employees. It’s not hard to imagine that this trend moves downstream, where a pool of 15-20 high-capped companies that have a low chance of short-term failure (i.e. Coinbase, MixPanel level) buy up 3-4 large buildings throughout a city and take care of the housing process in conjunction with not having to manage the complexity of the back-end and employees having a three second process of filling out a preferences form. Corporate housing has always existed, but it hasn’t existed with options: if you work for Google you live in Google housing where you have to spend more time with people from work. But if you pool 20 startups, where some of your college friends will also be around and your co-workers friends are also here, in a location of your choice, it exists

    Examples: Facebook, Google?

    Flaws: 

    • PR Nightmare: With this model, you are one bad drinking experience from a group of 25 year olds from 20 startups having to go on a PR apology tour.

    • Employee Turnover: How do you manage the complexity of employees leaving and helping to replace them? In a way, this could also be an increased benefit for employers: making the switching costs higher.

  • Startup Fully Streamlines the Discovery/Signing Process: A lot of startups have experimented with creating a successful roommate match service, with limited success so far. But the problem with the roommate search as a standalone service is that it’s only the first step of the extensive short-term housing search. There also is the problem of compiling documentation, finding the right apartment, etc. Shared living spaces are technically a solution to this, but they don’t actual

    Partial Examples: Rumr, Roomi

    Full Example: Bungalow 

    Flaws: 

    • Network Effects Needed for Initial Roommate Portion in Market with very high CAC

    • Need to Establish Liquidity For a Two-Sided Marketplace:

    • Strong Inventory Needed:

    • User Data Entry/Approval/Scraped (Blend for Short-Term Rentals)

 

Economics of Shared-Housing Models 

From an operational P&L perspective, there’s certain factors that need to be considered when funding a shared housing company. 

  1. How far away from the transaction is the platform: Like for residential housing, the further away from the transaction a platform is, the higher the margins and value added for consumer are, but the stronger the competition and the threat of losing big if a full-stack platform can be much more operational efficient by optimizing the entire funnel for consumers (i.e. a WeLive). The obvious upside is that of course, there is a ton of initial capital that’s being saved.

    And while local regulations aren’t as stringent on the buyers side for short-term housing as they are for residential real estate (a significant barrier to creating a full-stack company), as Zillow has shown it’s possible to win big without putting up huge initial capital.

    The big question is how do you create a platform that’s far from a transaction (i.e. without owning or leasing any property). Essentially, you have to hope get significant supply, which will in-turn attract buyers. Zillow, for example, was able to do so when the real-estate market was in a tailspin and realtors were willing to do anything for qualified leads. Simply: when you own the supply you generate the demand, something anyone whose running a funded startup probably understands. My personal belief would be that

  2. What’s the contingency revenue model: It’s not exactly a secret that real estate is cyclical but it doesn’t seem like a ton of these shared housing startups have thought about what happens when the market turns and they’re holding tons of long-term inventory or lease agreements. Saying something such as we'll rent out on AirBnb really doesn’t work because your day to day operational expenses rise exponentially from having to constantly manage your property and you don’t control any of the underlying resident data anymore. And it’s not even on the same (very-high) concern level of Opendoor in residential, where you have a strong AVM from flipping houses all the time that you can combine with public data to figure out how to adjust inventory for a market correction. There has to be a back-up plan for the company when demand falls drastically.

  3. How does the company obtain economics of scale if it’s not purely software-based? As a company moves from market to market, how does their advantage scale from buying physical inventory in each market. It’s simply not possible to grow continuously while paying face-value for buildings in each market, even while subsidized by VCs. It’s not entirely positive that brand name scales in geographical real estate market, and I suspect that’s why WeLive has had some trouble scaling in each individual market and slowed down expansion. I heavy suspect that there needs to be some type of almost network of these locations that allow for cross-location mobility and almost undermine AirBnb

Long-Term Defensibility: 

So this could probably be it’s own category, but what’s the “moat” that sets apart the winners in the space? Personally, the two big defensibility factors I see are the way over-used term “network effects” or more likely, a strong data advantage. 

  • Network Effects: There is definitely potential for network effects here both in pure software projects or dual software-physical location platforms.

  • Data Advantage: As I mentioned earlier, one of the primary problems with evaluating real estate models is often using residential as the same problem drivers for rental. However, if you look at the

The Facebook Effect: 

Facebook, to be blunt, has a ton of concerning shit going on that they have to worry about. With privacy concerns, falling engagement on the core platform, monetizing Instagram, slowing user growth, etc., they haven’t exactly been focused on real estate. But their early returns and potential as a rental aggregation platform are a huge threat. 

As Mike DelPrete points out, Facebook has started to syndicate inventory from the big rental sites such as Zumper and ApartmentList. A combination of ample inventory combined with deep customer knowledge for ad targeting and an easy to use UI that drastically bring down customer acquisition costs possibly allows them to become THE distribution from channels instead of a company like Zillow or Craigslist. Again citing DelPrete, the ROI for agents building awareness on the platform and promoting specific houses has been sky-high so Facebook has been mostly focused on using residential real estate as an advertising channel.

But if the core platform usage continues to fall and Facebook moves to almost an quasi-e-commerce discovery model while shifting social to Instagram (sounds crazy, but it’s not as far-fetched as imagined), it’s easy to imagine them becoming the single source of income