Tools
Pick a tool and see it run on a sample property, then read the methodology behind every figure it produces — the actual formula, the actual thresholds, the actual assumptions. Nothing on this page is described in terms Mesa cannot show you the working for.
Comps-based valuation with percentile bands, similarity and outlier filtering, defensibility scoring, and offer-price strategy derived from market conditions.
| # | Address | Price Comp sale price The recorded sale price, or an estimate when the source is LIST or AVM rather than SOLD. Check the SRC column: only SOLD is a settled transaction. | Adj. price Adjusted price What this comp implies for the subject: comp price per sqft (after the same recency adjustment the valuation band applies) × subject living sqft. Size is the only per-comp adjustment; no bed, bath, lot or condition adjustments are made per comp. The median here is over all listed comps, so it can differ from the band mid, which uses only comps that pass the similarity and outlier filters. | Src Price source SOLD is a recorded sale. LIST is an asking price, which is not what the property sold for. AVM is a model estimate with no transaction behind it. Non-SOLD comps widen the valuation band automatically. | Sqft | $/FT² Comp price per sqft Sale price ÷ living sqft, truncated to whole cents. This is the quantity the percentile band is built from. Not the total price. | Dist Distance Straight-line distance from the subject, not driving distance. It does not cross-check school zone, subdivision, or any barrier like a highway. A near comp can still be a poor comp. |
|---|---|---|---|---|---|---|---|
| 1 | 1427 Oak St | $412,000 | No data | - | 1,840 | $224/ft² | 0.20mi |
| 2 | 891 Maple Ave | $435,000 | No data | - | 1,920 | $227/ft² | 0.50mi |
| 3 | 2103 Pine Ct | $398,000 | No data | - | 1,780 | $224/ft² | 0.77mi |
| 4 | 56 Elm Dr | $449,000 | No data | - | 2,010 | $223/ft² | 0.94mi |
| 5 | 3308 Birch Ln | $421,000 | No data | - | 1,890 | $223/ft² | 1.16mi |
Illustrative example — sample property, not a Mesa result.
Why you can check the working
The Methodology tab is not marketing copy sitting next to the tool. It is the same text the in-app panel reads, generated from the engine's own field registry. When a formula changes, that text changes — this page cannot drift from what the software actually does.
Every field the engine can produce carries a prose methodology entry: how it is computed, which thresholds apply, and what it does not account for. The DSCR entry tells you it is net operating income divided by annual debt service, that lenders typically want 1.20–1.25×, and that it reads zero on an all-cash deal. You get the same answer here as you do in the app.
The county count is read from the engine's dispatch tables
(GET /api/v1/coverage). A county is counted only after a recorded live
fetch returned data for it, and the headline is asserted at build time to follow from
the snapshot rows. A hand-edited coverage number fails the build.
If a field has no methodology entry, it does not get claimed on this page — the Methodology tab says so explicitly rather than filling the gap. If a tool ships before its explanation does, you will see which fields those are.
Illustrative examples Sample data throughout this page is a worked demonstration, not a live Mesa result. Real analyses run in the app against your subject property.
Mesa