Dwelling removes the biggest barrier to entry for naturally-affordable housing – access to capital for local-scale real estate development.
AI to combat the housing shortage.
01
The problem
There's a massive gap between housing supply and demand.
7.2 million homes, to be exact.
The solution is simple: build more homes. But how?
Here's what we've learned:
- Independent real estate developers want to build for their local communities, but are constrained by the complexities of underwriting the deal and finding investors willing to risk capital on small developers. Read about it.
- Investors are eager to fund meaningful projects that improve neighborhoods without causing displacement, but these deals are hard to find and do due diligence at any meaningful scale. (We'll tell you dozens of stories)
- Governments are desperate for community-driven development to address the crisis, but are both broke and stretched thin – Like Houston.
Put another way:
- One party wants to build.
- One party wants to fund.
- One party wants both to happen at a dramatically higher rate.
If they were working as everyone wants them to, we wouldn't be 7 million homes short.
02
Solution
Dwelling is the glue that helps local development thrive.
We're building software to support an ecosystem of local, incremental development in 3 steps:
01.
AI agents to figure out what you can build on a site and how much money it can make
Real estate developers are stuck using outdated tools in an industry that's anything but simple. LLMs can break this bottleneck by automating previously non-automatable processes while still respecting a developer's local knowledge, preferences, and experience. Each AI agent is trained on a particular outcome, and they can be combined to create workflows incredibly bespoke to that specific developer and how they work.
Developers model their deals on Dwelling in a fraction of the time.
02.
Crowdfunding marketplace for vetted deals
43% of new housing development is controlled by 10 companies, leading to communities everywhere looking the same – not to mention unaffordable. There is a large, growing group of folks that are tired of this. They have capital that they are eager to invest into building more livable, walkable, and affordable communities.
Dwelling lets anyone invest any amount into these previously hard-to-find projects.
03.
(Long-term) Facilitating the fundraise
Governments can control their own destiny by soliciting the types of development that they want to see in their city, incentives they're willing to offer, and use Dwelling for expedited permitting. The rent-seeking behavior that occurs in the permitting stages – from lawyers, lobbyists, and permit expediters – can be captured directly by the government, with developers able to more freely build.
Dwelling helps deals get done, and will help projects get built.
03
Financials
business model
We're building 2 revenue streams:
01
Developers will pay us when they model deals
When developers analyze sites in Dwelling, they'll be billed based on usage. This is important for a few reasons:
- Developers should be billed when they receive value from the platform. The value comes in the form of work getting done on their behalf, not just from having a seat.
- Developers hate subscriptions. They're not capitalized to support ongoing charges and prefer paying as they go.
Each AI agent will have a "cost per run". The amount will vary based on the complexity of the agent, but will be substantially cheaper than the time cost of the developer doing everything manually, and especially less than hiring a team of analysts.
Users will fill up their account with credits, and as agents are run, credits will be withdrawn from their account.
02
Investors will pay a transaction fee when transferring capital
When a user in the crowdfunding marketplace transfers capital (via ACH only), we will take a 1% fee for matchmaking and facilitating the transaction.
In the beginning, we want to avoid ever holding this money–thus, we'll just monetize the transaction. Long-term, we plan to operate more like a fund with assets under management. When investors transfer capital, we will charge an annual management fee (between 1-1.25%) until the principal is returned to the investor.
TAM
Disclaimer: While we've aimed to be reasonable in our market analysis, many assumptions are made regarding areas that are not easily quantifiable. We've aimed to keep our assumptions quite conservative. Ultimately, the US real estate market is massive. We're aiming to be pragmatic regarding what portion is actually addressable in the first iteration of our product.
AI Agents TAM: $6.56bn
The due diligence process in real estate is a massive time and money drain. Beyond the “soft costs” like market analysis, zoning studies, and consultant fees—which can easily hit $40K on a $400K unit—developers sink a lot of their own time into it.
As of June 2024, the U.S. Census Bureau reported that there were approximately 1.446 million building permits issued, 1.353 million housing starts, and 1.710 million housing completions. If we just look at US residential projects that get built that’s roughly a $2.56 billion TAM (1.71 million * $1,500 conservative estimate for software fees paid to Dwelling).
When factoring in sites that are analyzed but never acquired, the TAM increases quite a bit. A goal of Dwelling is not just to analyze properties that are already acquired, but also to help identify properties that should not be acquired much earlier. While those scenarios may not justify the same $1,500 estimate, we could assume 10 abandoned projects for each acquired project at a rough cost of $300 in costs paid to Dwelling, another $4bn of TAM (1.353m housing starts * 10 * $300 estimate for software fees paid to Dwelling).
Crowdsourcing marketplace SAM (Serviceable Addressable Market): $24.9bn
Our initial target market is households earning over $150k in 11 cities grappling with significant housing shortages: San Francisco, New York, Los Angeles, San Diego, Houston, Detroit, Washington D.C. metro area, Seattle, and Austin, TX, Boston, and Chicago.
These cities have populations that are not only high-income but also face political momentum aimed at boosting new housing developments. With 8,317,400 households represented, we estimated a mean income of $200k among these earners, conservatively measured to target the p95 salary amongst this group.
To derive the SAM, we assumed that these households invest 10% of their income, which amounts to about $20,000 per household annually. From that investment, 15%—or $3,000—is directed specifically into real estate. Multiplying the number of households making over $150k (8,317,400) by the amount invested in real estate, we arrive at an estimated market size of $24.9 billion that we believe we can target.
While we believe these 11 cities are particularly good markets for us to launch to, it is only the start. We hesitate to speculate too much on TAM for this piece until after we prove the concept in these 11 cities.
We can provide a financial model spreadsheet upon request.
04
Why now?
It's time to unleash the swarm.
We need 7.2 million more homes, and that’s just the start. We need to maintain that rate of development, which we haven’t been able to do since the 50s. Further, the land around these new homes needs to become greenspace and retail areas, reducing car dependency and urban sprawl.
Developers are chomping at the bit to join this building boom, but the current system is a clogged drain. Governments are changing land use regulation to open up more land for development, but developers need innovation at the earliest stages to fill this housing gap sustainably.
The housing gap will not be filled by large corporate developers and top-down government mandates. It will be filled by many hands over many years. However, the ecosystem for these many hands has sat in rot since the 1960s when the roots of the housing crisis were planted.
We need to unleash the swarm of local residents developing their communities in a way that adapts and evolves with the ever-changing needs. Dwelling is the platform that supports the swarm.
This is what it means to unleash the swarm.05
Team
We met at Webflow when we were both on the product growth team. Over time, we decided to work together because both wanted to work with someone who was as much of a startup sicko as the other.
We’ve driven product growth and built AI and machine learning at billion-dollar companies. We’re technical founders laser-focused on two things: talking to users and building the product.
CEO
Tristan Tarpley
I'm a software engineer, product guy, and urbanist who has been obsessed with all things urban development and urbanism for years. I've been working in product growth for late stage, Series A, and pre-seed companies for nearly a decade. Most recently I've worked on building generative AI features into Webflow's core product.
COO
Aneesh Soni
I'm a data scientist and engineer with a passion for leveraging machine learning and AI to solve complex problems. I've led data science teams, focusing on transforming data into actionable insights across industries in digital tech and finance. Recently, I've worked on building data frameworks for Webflow's first party analytics and built predictive models to optimize marketing spend.