HomeStats vs Reventure
Both projects publish state-level U.S. housing data; the products are very different. This page is honest, factual, and doesn't trash the competition — Reventure has a deserved audience, Nick Gerli's analysis is solid, and they publish a backtest of their own forecast, which is more than most tools do. The summary below is for someone deciding which to use for what. If you came here looking for the free version of a metric you saw in their app, start with the crosswalk.
Data as of May 31, 2026 (Redfin reporting month). Pipeline last ran July 30, 2026. Those are two different dates and we label them separately on purpose.
Their metric → our free equivalent
Left column is Reventure's published metric name. Middle column is the formula we compute for the same quantity, with the source dataset named, so you can check it against the raw file. Right column is the free page. We do not restate their definitions — see note 2.
| Reventure metric (their published name) | What HomeStats computes for it | Free HomeStats page |
|---|---|---|
| Reventure Price Forecast Score1 | A five-year median-price projection for every state. Base growth is half the state's trailing 36-month Redfin CAGR and half its FHFA 1991–2024 long-run CAGR, with a ±1σ band from blended volatility. That trailing leg is 36 months, not five years: forecast.js asks for a 60-month rate first, but that lookback needs one more monthly observation than the trend files carry, so it returns null on every state and the blend falls back to 36 months. The forecast page says the same thing in its own method notes. Each state is shown next to its own one-year hindcast error, so you can see how the model did before deciding what to make of it. Alongside it, a five-factor crash-risk score whose weights are published. | Home price forecast by stateCrash-risk method |
| Overvaluation % | Price-to-income: median sale price ÷ median household income (Redfin ÷ Census ACS table B19013). Benchmarked against the median of the current cross-sectional snapshot, computed at build time and printed on the page — not against a proprietary long-run anchor, because no such anchor exists in our data and we will not invent one. | Overvalued housing markets → “Price-to-income ratio by state” |
| Rental Market Overvaluation % | Price-to-rent: median sale price ÷ (monthly rent × 12), i.e. how many years of gross rent a home costs. Rent is the Zillow Observed Rent Index averaged across the metros in each state. State grain only: the rent series in our pipeline is state-level, no county file carries rent, and dividing county price by state rent would produce a price ranking wearing a rent label. So there is no county price-to-rent here. | Overvalued housing markets → “Price-to-rent ratio by state” |
| Price Cuts | Redfin's PRICE_DROPS field, untransformed: the share of active listings where the seller has reduced the asking price. It is not the share of homes that sold below asking, and it is not the average size of a discount. Ranked as its own number rather than folded into a composite. | Price cuts by state and county |
| Inventory Surplus / Deficit | Two ways. Inventory surplus = (current active inventory − the 60-month average) ÷ that average. Months of supply = active inventory ÷ homes sold in the reporting month, read against the cut points used everywhere on this site: under 3 a seller's market, 3–6 balanced, above 6 a buyer's market. Both NAR and Redfin publish the months-of-inventory measure, but neither publishes those cut points as a formal definition — they are the long-standing industry reading of it, and the months-of-supply page below says so and links both. | Months of supply, state and countyDistress watch |
| Days on Market | Redfin median days on market, plus a DOM spike figure: (current DOM − the trailing 24-month average DOM) ÷ that average. Twenty-four months, not sixty: that is the window the pipeline actually averages, and it is the window labelled on every state page. The level tells you where listings sit; the spike tells you how far the market has moved off its own recent baseline. | Listing velocity decayMarket movers |
| Home Value Growth (5-year) | The full 60-month Redfin median-price series is charted on every state page, with Redfin's own year-over-year change beside it. The compound annual rate we compute off that series is a 36-month CAGR, not a five-year one, for the reason given in the first row, and it feeds the forecast rather than being ranked on its own. Market movers is a different measurement: latest month against the mean of the prior three, at county grain, so it catches turns rather than long drift. | All state pagesForecast page — trailing window and blended growthMarket movers |
- Reventure names its own flagship metric three different ways inside a single post: “Home Price Forecast Score,” “Reventure Score,” and “Reventure's Price Score.” That is a plain observation, not a criticism — it is worth knowing because searching any one of those strings should land you in the same place. Source: reventureapp.blog, accessed July 27, 2026.
-
The middle column states our formula, not theirs. Reventure's glossary
(map.reventure.app/profile/glossary,
the URL listed in their own sitemap, which 307-redirects to
www.reventure.app/profile/glossary) is rendered in the browser and returns no definition text to a fetcher — a server fetch comes back with the title “Reventure App” and nothing else (checked July 27, 2026). Rather than paraphrase a competitor's definition into words they did not use, we print ours and let you compare. If the two measure different things, our formula is on the page and theirs is in their app — that comparison is yours to make.
Feature by feature
| Feature | HomeStats | Reventure | Why it matters |
|---|---|---|---|
| Pricing | Free everywhere. No signup, no paywall, no tracking. | Free tier plus a paid Premium subscription — that is the word in their own pricing page title. We link the page rather than restate a figure we cannot verify from a server-rendered source. reventure.app/pricing | Decision-grade housing data should not sit behind a paywall. And a comparison page that quotes a competitor's price from memory is doing the same thing it accuses them of. |
| Per-region depth | 51 state-level regions, 3,073 counties, 18,052 cities, 925 metros, 16,386 ZIPs, each with its own page. The state count is the 50 states plus the District of Columbia, which is how Redfin ships the file set. A state file carries 53 top-level data groups — price and inventory, census, 60-month trend, rent, taxes, insurance, utilities, hazard risk, water quality, occupations, permits, migration and the rest — and the page renders most of them. What is rendered is on the page, not behind an upgrade. | Not assessed. We only describe what we can load without an account, and the map product sits behind a sign-in. Their published coverage figures differ between their own surfaces, so we are not going to quote one as fact. | You should not have to upgrade to see what each state actually costs. |
| Crawlable and citable | Server-rendered static HTML. Every state, county and city page is in our sitemap, the data pages carry Dataset JSON-LD, and there is no client-side gate between a crawler and a number. | Their sitemap index resolves to one child sitemap containing 12 URLs, all on the map.reventure.app host — /map, /graph, /reports, /report-pdf, /join, /cancel, /reset-password and five /profile/* account routes. No geographic pages are listed. Checked July 27, 2026: reventure.app/sitemap.xml | A number that only exists inside a client-rendered app is harder for search engines, language models, and researchers to reach. That is a structural difference, not a quality judgment. |
| Take-Home Income Map | Yes. $75k single across all 50 states with federal + FICA + state + local breakdown using 2025 IRS brackets. Take-Home Map. | No equivalent we could find on a page we can load. | The same gross income produces wildly different take-home across states. Hidden until we shipped this. |
| Wage-Stagnation Lens (1970 CPI + gold) | Yes. Median income deflated to 1970 dollars + gold-ounce ratio for every state, with the BLS methodology-change history (OER 1983, Boskin 1996, hedonic 1998, geometric mean 1999, chained CPI 2002). Wage-Stagnation Lens. | No equivalent we could find on a page we can load. | Headlines say wages are at all-time highs. Two honest deflators say something different. |
| Veteran / VA layer | VA loan toggle (0% down + no PMI + funding fee), tax-free disability + W-2 equivalent, healthcare savings, 100% P&T property-tax + vehicle reg + permits + CHAMPVA — per state. Veteran / VA housing by state. | No dedicated veteran modeling that we could find on a page we can load. | Veterans are systematically underserved by mainstream housing tools. |
| Interactive sliders | PaycheckSlider, AffordabilitySlider, RentVsBuyCalc with tax-benefit + stay-probability + maintenance-shock layers. | Not assessed. We only describe what we can load without an account, and the map product sits behind a sign-in. | Real decisions need to be testable against your own number, not a fixed benchmark. |
| Public read-only API | Yes. /api/states/index.json for every state row, and /api/states/{slug}.json for one. No key, no auth, CORS open. Free with attribution — terms at /fair-use/. | No public API is documented on any Reventure page we can load, and none of the 12 URLs in their sitemap is an API route (checked July 27, 2026). | Researchers, journalists, and analysts shouldn't have to scrape. |
| Citations on every metric | Inline links to primary sources (Redfin, Census, FRED, FHFA, BLS, IRS, VA.gov, EPA, FEMA, NOAA) on every page, plus the formula in plain text. Full list at /sources/. | Not assessed. We only describe what we can load without an account, and the map product sits behind a sign-in. Their public writing does cite its sources — see the forecast review linked in the FAQ below. | Verify everything against the source. Trust is built by being checkable. |
| Distribution | Web-first, tools-led. SEO-targeted state, county and city landing pages. | Strong YouTube + X presence (Nick Gerli). | They win on personality + video. We win on tools + cites. |
| Editorial stance | Data and tools only. Conclusions are yours. | Analyst commentary with a stated point of view. | Both have a place; we keep ours neutral so you can build your own thesis. |
Common questions
Is there a free alternative to Reventure App?
Yes. HomeStats publishes the same categories of market data with no signup, no email gate and no subscription: the home price forecast and crash risk by state, price-to-income and price-to-rent overvaluation, seller price cuts, and months of supply. The crosswalk above maps each paid metric name to the free page that answers it. Every figure is computed from Redfin, Census ACS, Zillow ZORI and FHFA files, and the formula is printed on the page.
What does Reventure App cost?
Reventure has a free tier plus a paid subscription their own pricing page titles “Premium.” We are not printing a dollar figure here. Their pricing page renders in the browser, so a server fetch of it returns only that title and no prices (checked July 27, 2026), and the figures circulating on third-party sites disagree with each other. Check reventure.app/pricing directly. HomeStats costs nothing and has no paid tier to compare against.
Is Reventure's home price forecast accurate?
Reventure publishes its own backtest, which is more than most tools do. In their forecast review they report that their score achieved a 74% correlation coefficient in predicting price growth among metros above 100,000 population from October 2023 to October 2024, that the correlation falls to 0.56 when the population filter is removed and all of the roughly 1,000 metro areas are included, and they say plainly that accuracy diminishes at the county and ZIP level. Those are self-reported figures, computed by Reventure over a window and against an index they chose, with no third-party audit. Read it and judge: reventureapp.blog, accessed July 27, 2026. Our own forecast page prints a one-year hindcast error for every single state next to its projection, for the same reason.
Can I download the data?
Yes. /api/states/index.json returns every state row as JSON — no key, no auth, CORS open — and /api/states/{slug}.json returns a single state. Reuse terms are at /fair-use/: free with attribution. There is no paid tier and no rate limit to buy your way out of.
The honest summary
Use Reventure if you want a daily YouTube/X dose of housing market analysis with a strong editorial voice and you're comfortable paying for the deep app data.
Use HomeStats if you want a free, cited, tools-led data layer — Take-Home Map, Wage-Stagnation Lens, Veteran/VA modeling, public JSON API — that you can build your own analysis on without a subscription.
They're complementary, not strictly competitive. Many serious housing-data users follow Nick on X and use HomeStats as their daily dataset.
How the numbers on this page were produced
- Our coverage counts (51 state-level regions, 3,073 counties, 18,052 cities, 925 metros, 16,386 ZIPs) come from
data/meta.json, read at build time. They move when the pipeline moves. - “53 top-level data groups” is
Object.keys()on a real state file at build time, not a hand-written figure. They are groups of related fields, not a count of individual metrics, and we are not going to inflate one into the other. - The forecast's trailing window (36 months) is measured at build time by running
src/lib/forecast.jsover every state file and recording which lookback returned a value, not by trusting the variable nametrailingCAGR5y. The DOM spike window (24 months) and the inventory-surplus window (60 months) are the windows the pipeline averages, and they are the same windows labelled on the state pages. - Data as of May 31, 2026 is the Redfin reporting month carried on every state file, while the pipeline itself last ran July 30, 2026. Market figures should be read against the reporting month, not the run date.
- Every Reventure fact above is either something we loaded on a Reventure-owned URL — linked, with the date we loaded it — or it is not on this page. We removed a subscription price that had been sitting here as an estimate, because their pricing page is client-rendered and no server-readable figure exists to cite.
- What we did not do: we did not quote app-store reviews, did not restate their coverage counts, and did not repeat their comparison framing against Zillow. Their self-reported backtest numbers appear above verbatim, attributed, with the link.
HomeStats sources: Redfin Data Center, U.S. Census ACS, Zillow Observed Rent Index, FHFA House Price Index, FRED. Full list at /sources/, methods at /learn/ and /methodology/crash-risk/, reuse terms at /fair-use/. Reventure surfaces referenced, all accessed July 27, 2026: pricing, sitemap.xml, glossary, forecast review.