A tool that tells you what a listing won't: does this land flood, can you build on it, can you get to it
A land listing tells you what the seller wants you to know. A photo won't tell you if the parcel floods, whether you can build on the slope, or how far the nearest road really is - and those are exactly the things that kill a deal after you've paid for it.
So I built LandBenchmark. You draw a boundary anywhere on Earth and, in about a minute,
you get a screening report measured directly from satellite imagery and public record — not scraped from the listing.
How it works — three steps:
1. Draw it. Draw a parcel boundary on the map. No account needed to start.
2. We measure it. Radar + optical satellite imagery, elevation models, soil surveys and public registries, computed for your polygon. Every conclusion comes with its working: what we observed, why it matters, how it was computed, when the imagery was taken, and the published method or agency behind it.
3. You get the report. A verdict across Water, Buildability, Access, Hazard, Climate and Surroundings — every value cited and dated.
What's actually in it:
- 84 land factors in the framework; 55 measured live from satellite + public record (35 answer anywhere on Earth, 20 add extra depth in the US).
- 8 human checks — the things a satellite physically can't settle (title, easements, zoning, water rights). We put them on a checklist and route them to you. We never guess them.
- Daily monitoring — tell us a parcel is yours and we re-measure it every day and email you what moved.
- An AI you can ask anything about the report, and clean, shareable PDFs.
The part I'm actually proud of: I publish my own error rate. I check the elevation model against NOAA's published survey marks and post the result on the site — mean error 0.90 m (RMSE 1.75 m) across 22 marks. Most tools in this space never tell you how wrong they are. If a value can't be measured and sourced, we don't show it — you get a "pending", never a dressed-up guess.
Try it: screening a parcel is free (no card). If you want the depth — daily monitoring, the
AI chat, clean reports — there's a 3-day free trial, no charge for 3 days, cancel anytime.
I'd genuinely love feedback from this crowd — especially on the onboarding (draw → report) and
whether the "we publish our own accuracy" angle lands or reads as too in-the-weeds. What would make you trust a tool like this enough to use it on a real purchase?
