How Haven Analyzes
Real Estate
Where the data comes from, what the risk audit looks at, and how the Bull / Likely / Bear ROI model works. No black boxes — and no overstated precision.
Where the Data Comes From
Haven pulls property and listing data through licensed real-estate data APIs and references public datasets across six categories. Below is what each category draws on. Coverage varies by market and by provider.
Property & Transaction Records
- ›County assessor and public records (via licensed data providers)
- ›Active and recently sold listing feeds
- ›Price and tax assessment history for the subject property
- ›Days on market and price-change history
Regional Cost & Hazard Benchmarks
- ›State-specific effective property tax rate schedules
- ›Coastal and hurricane zone insurance risk cost adjustments
- ›Regional hazard and homeowners insurance cost baselines
Neighborhood & Demographic Context
- ›Local school and district reference data where published
- ›Walkability and transit access metrics
- ›City and submarket demographic context
Market & Rental Economics
- ›Local rental comps and estimated market rent yields
- ›Cap rate, cash flow, and DSCR underwriting metrics
- ›Prevailing mortgage interest rate benchmarks
- ›HOA dues and fee context where available
Infrastructure & Property Characteristics
- ›Square footage, lot size, and year built specifications
- ›Property style, structural attributes, and room counts
Tax & Assessed Valuation
- ›County tax assessed valuation records
- ›State property tax schedules across all 50 states and D.C.
- ›Assessed value vs. market asking price comparative ratios
Not all data is available in all markets. Where a signal is missing for a property, the analysis notes it rather than guessing.
The Property Risk Audit
Each property is reviewed across six dimensions. The in-app score reflects how the available data for that address measures against benchmark thresholds — dimensions with no data for a given property are excluded rather than assumed.
Property & Title
Deed history, lien status, ownership chain, permit and code-violation signals
Environmental Risk
Flood zone, wildfire and earthquake hazard references, known hazard proximity
Neighborhood
School ratings, crime and safety indicators, walkability, local economic context
Market Economics
Rental vacancy, cap-rate benchmarks, median rent trends, local demand
Infrastructure & Development
Publicly reported transit, rezoning and development activity, broadband availability
Tax & Financial
Assessment history, tax delinquency and lien signals, special assessments, HOA context
Bull / Likely / Bear ROI Modeling
Haven models a range of outcomes rather than a single-point estimate that hides uncertainty. Here is the process:
Base Case Inputs
Haven assembles listing price, estimated rehab cost, local rental comps, current mortgage-rate context, property tax rate, an insurance estimate, a vacancy assumption, and HOA fees to establish a baseline cash-flow model.
Scenario Assumptions
Key inputs — rent growth, vacancy, appreciation, expense growth — are varied across optimistic, central, and conservative assumptions rather than a single point estimate.
Bull / Likely / Bear Modeling
Haven computes cash flow and return metrics under each set of assumptions to produce a Bull, Likely, and Bear range instead of one number, so the downside is visible alongside the upside.
AI Synthesis
An AI model reviews the property record, the risk signals, and the scenario ranges together and produces a plain-English read plus a 0–10 confidence verdict. This is AI-assisted analysis, not a proprietary quantitative fund model.
Human Judgement
The output is a starting point for your own due diligence and a conversation with a licensed agent — not a recommendation to buy or a guarantee of return.
Important Limitation
These are model outputs, not predictions. Haven's automated analysis is AI-generated from third-party data and current assumptions; it does not account for everything (rate shocks, disasters, regulatory change, or the specifics of a given deal). It is an analytical input for your decision, not a guarantee of return. Always do your own due diligence and consult a licensed real estate professional and financial advisor before investing.
Methodology questions answered
Where does Haven source its property data?
Haven pulls property records and listing data through licensed real-estate data APIs, and references public datasets such as FEMA flood maps, school ratings, and local crime and market data. Coverage and freshness vary by market and by data provider.
What does Haven's risk audit consider?
The audit looks at six areas: property and transaction records (title, liens, tax history), environmental and climate risk (flood zone, wildfire, earthquake), neighborhood factors (school ratings, crime trends, walkability), market economics (rental vacancy, cap-rate benchmarks, rent trends), infrastructure and development activity, and tax and financial records. Not every data point is available for every property; the analysis uses what is available for that address.
How accurate are Haven's ROI projections?
Haven produces range projections (Bull / Likely / Bear), not a single guaranteed number. They are model outputs based on current assumptions and available data, intended as an analytical starting point. Actual returns depend on price negotiated, financing, management, and market conditions. Always do your own due diligence and consult a licensed professional.
How often is Haven's data updated?
Listing and property data is refreshed from Haven's data providers on an ongoing basis. Public reference datasets (climate, demographic, market) update on their source agencies' own schedules, which range from monthly to annually.
Does a human review the analysis?
The automated analysis is AI-generated. When you choose to work with a Haven partner agent, a licensed local agent reviews the property with you before any offer. Haven's tools are decision support, not a substitute for professional advice.
See the methodology in action
Run a live risk audit and Bull / Likely / Bear ROI projection on any US property — free.