An estimate should
show its work.

One price is not the whole picture.

We preserve whether a number is a full cash price, an auction bid, a down payment or annual rent. Search sorting and comparisons use the same price basis. A closed bidding period is labeled closed; a missing page is never treated as a completed sale.

A record needs an origin.

Every ingested offering retains its source URL, observation time, source identifier and status. Crawler runs retain response metadata and content hashes. Information attributed to a source is not represented as an independent parcel verification.

Value begins with appropriate transactions.

The current workbench uses reviewed, source-linked closed transactions with matched rights and use. At least three distinct eligible comparisons are required. The result is a transparent summary of the supplied adjusted values and their observed range. It is not a statistically calibrated prediction interval or a validated national automated valuation.

Asking-price summaries and seller expectations do not become independent sale-value evidence. Users must record the rationale for adjustments. Thin evidence or a mismatched interest calls for more research.

Keep current and future conditions separate.

Current as-is value, proposed-use income and development residual value answer different questions. Income and development outputs remain conditional scenarios. Acquisition, site work, holding, financing and exit costs are explicit so that the same expense is not silently counted twice or omitted.

A discount needs context.

The opportunity screen compares the relevant value scenario with the corresponding cost basis. It withholds a strong opportunity conclusion when critical access, title or intended-use evidence is unresolved. A positive arithmetic spread is not a guaranteed profit.

What the map does and does not say.

Markers use supplied coordinates and are labeled approximate. We do not invent parcel boundaries from acreage or a location pin. Public basemaps are credited to their providers. No national parcel fabric, paid owner database or completed flood/wetland analysis is implied by displaying a map.

Access is source-specific.

The source registry includes research candidates, verified websites and enabled adapters. These states are separate. Crawlers use a fixed host allowlist, check robots rules, enforce budgets, and record failures. Permissioned feed imports require explicit rights confirmation. Third-party photos and full marketing copy are not automatically republished.

Private work stays private.

Workspace notes, service records, scenarios and files are scoped to the signed-in account. Document downloads require ownership checks and are served as attachments. A file’s hash identifies its contents; it does not authenticate the claims in those contents.

What would justify a stronger estimate?

Licensed sale outcomes, reliable parcel features, historical snapshots, geographic and temporal holdouts, and published error by land cohort. Model accuracy has to be measured before an automated estimate earns stronger language. The application currently supplies the evidence and calculation workflow.

Read the detailed valuation research or open the analysis workbench.