June 2026 — Demand-driver intelligence, an analysis that reasons like a local, and one-click install
The AI analysis now reads the institutions around a property as rental-demand signals — universities tagged with their student enrollment, hospitals with their bed count, research clusters as a category of their own — all on a rebuilt open-data amenity layer. We also hardened the Investor Fit Score around your strategy, made maps honest about approximate addresses, and put the browser extension on the Chrome Web Store.
In April we folded Japan's authoritative data into the platform; in May we put it to work with a personalized Investor Fit Score and a market map. June was about judgment: we taught the analysis to read the neighborhood around a property as rental demand — which universities, hospitals, and research clusters actually shape who would live there — rebuilt that amenity layer on open data so the picture is complete and accurate, sharpened how the analysis reasons about Japanese real estate, and removed friction with honest maps and a one-click extension install. Here's what shipped.
1. The neighborhood, read as demand — not just dots on a map
Knowing what sits near a property is table stakes; what matters to an investor is who that draws as tenants. In June we rebuilt the nearby-amenities layer on open data — Overture Maps, OpenStreetMap, and MLIT — and, more importantly, taught the analysis to read the institutions around a property as demand drivers. Universities and colleges now carry their student enrollment (official MEXT figures), large hospitals carry their bed count (MHLW data), and research institutes and medical-R&D campuses are surfaced as their own category. So instead of a flat "university nearby," the analysis reasons the way a local would: a 20,000-student campus within walking distance means durable demand for compact 1K/1DK units with a sharp March–April turnover; a 600-bed hospital anchors a stable, long-tenancy cohort of medical staff; a research park points to higher-income, less price-sensitive professionals. Every property analysis now includes this demand-driver read, with campus-appropriate distance bands — a university 2 km away is still "close" in a way a convenience store 2 km away is not.
Rebuilding on open data paid off in coverage and accuracy too. The amenity picture is now complete and uncapped — we removed the old "20+" limit, so counts are exact — shrine coverage jumped more than tenfold by drawing on OpenStreetMap, and transit comes from MLIT's national rail and bus datasets.
Source: Overture Maps + OpenStreetMap + MLIT, enriched with MEXT university enrollment and MHLW hospital data
2. An AI analysis that reasons like a Japanese professional
The analysis got materially smarter about how property is actually evaluated in Japan. It now applies the right yield basis (gross vs. net / 表面・実質), checks bank-financing feasibility the way a lender would (積算評価 collateral valuation), flags rebuildability constraints (再建築不可 / 接道義務 road-frontage rules), applies the correct brokerage-commission rules by transaction type, and frames natural-disaster risk in balanced terms rather than alarmist ones. We also corrected the underlying facts the analysis cites — light-gauge-steel (軽量鉄骨) depreciation periods of 19 / 27 / 34 years by steel thickness, the ¥330,000 tax-inclusive commission cap for sub-¥8M properties under the July 2024 reform, and the 2,000㎡ land-notification threshold — and added a name-fidelity safeguard so the analysis names the institution that's actually next door, not a more-famous one nearby.
Source: Japan real-estate practice standards + on-platform analysis methodology
3. The Investor Fit Score, hardened around your strategy
May introduced the Investor Fit Score; June made it sharper and strategy-aware. Returns are now compared on a like-for-like yield basis with the yield type named explicitly, and scoring understands whether you're investing for income or appreciation — so an appreciation-focused profile is judged on growth potential ("Solid appreciation potential for this location tier"), not against a gross-yield target it was never meant to hit. Location matching now considers city and ward, not just prefecture; risk alignment factors in the building's seismic-standard era; and we closed a gap in the demographic-fit scoring.
4. Honest maps, and addresses that land in the right place
Not every listing has a precise street address, and the map now says so honestly. Properties with an area-level or hidden address are drawn as the actual neighborhood (chome) boundary instead of a single, misleading pin, and they're flagged with an "≈" marker on the list and cluster maps so you can tell them apart from exact addresses at a glance; when no chome matches, the map falls back to the ward boundary. On the input side, property forms now include address autocomplete that captures the canonical coordinates at the moment you pick a suggestion — so an address is no longer re-geocoded from free text into the wrong ward — and English-language users now see a clean English/romaji rendering of each address (e.g. "3 Chome-16-9 Torikai, Chuo Ward, Fukuoka") while the Japanese address stays the source of truth.
Source: Government boundary data + Google Places autocomplete
5. Install in one click — extension on the Chrome Web Store
The Tatemono IQ browser extension — the fastest way to pull a listing into the platform — is now published on the Chrome Web Store. Install it in one click with automatic updates; no more downloading a ZIP or using "load unpacked." The in-app download page, the Import Properties guided tour, and the documentation all walk you through the new flow.
6. Two new market reports
We published two more live market reports this month. The newest, the Japan Prefecture Demographic Demand Scorecard, ranks all 47 prefectures by projected household growth through 2040 for the cohorts investors care about most — single adults aged 25–39 and elderly single-person households — using official IPSS 2024 projections. Earlier in the month we added Tokyo vs. Regional Cities: Price Comparison, which sets Tokyo's prices against major regional metros. They join our first report, Japan Residential Property Prices 2020–2025, on the market reports page.
Source: IPSS 2024 prefectural household projections + MLIT transaction data
Also improved in June
- The Market Intelligence map now renders with Leaflet for faster load times.
- Refreshed in-app guided tours — the AI Assistant (Aki) tour now matches the current side panel, the Import tour covers the Maisoku (マイソク) PDF path, and there's a new tour for Lists.
- Clearer property labels — floor area is now labeled by building type: 専有面積 (Exclusive Area) for condominium units, 延床面積 (Total Floor Area) for other buildings.
- Stability and accuracy fixes across the board: AI analysis no longer crashes on unusual commercial listings, Maisoku PDF analysis was restored after Google retired an older vision model, and several map and geocoding errors were resolved — including a false "approximate address" transit warning that could appear on properties with an exact, confirmed address.