Settings

Your plan

Current plan: โ€ฆ. Compare plans

Plans are stored on this device only โ€” there is no account and no payment processor connected yet. See Plans for what that means.

Analysis engine

All analysis runs inside the app โ€” there is no model server, no API key and no external service. The valuation, rent, negotiation and growth figures come from HouseScout's own model, trained at build time and evaluated in your browser; the written per-listing analysis is generated during the build from each listing's numbers. Everything works offline.

Confidence is low while few listings publish an asking price โ€” most NZ stock in this price bracket sells by auction or deadline sale. The model shrinks toward documented valuation relationships rather than overfitting a small sample, and says so rather than presenting a false precision.

What counts as a match

These decide which listings pass and how they rank. The money behind every figure lives under ๐Ÿ’ฐ Your numbers at the top of any page.

Your numbers, in full

Every dollar figure on the site is worked out from these. Edit them here, or from ๐Ÿ’ฐ Your numbers on any page.

Deposit
$50,000
Household income before tax
$120,000
Interest rate
5.19%
Loan term
30 years
Interest-only period
5 years
Room rent
$220/wk
Weeks empty each year
3
House price growth
5% a year
Council rates
$3,200/yr
Insurance
$2,200/yr
Upkeep
1.0% of value
Property manager
self-managed
Your living costs
$2,400/mo
Top tax rate
33%
Projection horizon
20 years

About this build

This is the static build of HouseScout. Listings are scraped from public NZ property sources when the site builds, with photos downloaded and self-hosted; scoring, valuation and all financial analysis then run entirely in your browser. Where a listing publishes no price, HouseScout's estimate is shown and labelled as such. For LINZ land enrichment and a persistent database, run the optional Python backend (see README).

Which districts a build covers is set by SCRAPE_REGIONS in the deploy workflow โ€” any of the 59 verified locations in scraper/regions.py, a whole region, an island, or the entire country. The Listings page filters by whichever districts this build actually scraped.