Questions, answered.

Yes. The core verdict — tap a spot, see how long it's shaded — is free and will stay free. No trial, no account, no credit card.

No. It runs in your phone's browser. Add it to your home screen if you want the app feel.

Building shadows are available across the Sunbelt states shown on the map. Our first local test is Gwinnett County, Georgia. Heights use public laser scans where available and are labelled as estimates elsewhere. Tree shade is excluded for this test.

If you open the map somewhere else, it stays put and says “You're outside the map area.” We don't fly you to your location on purpose: an empty map with no buildings and no shadows is a dead end, not a feature.

Yes — every verdict has a share link. Send “this spot is shaded until 4 PM” to whoever you're meeting. The link carries the area, so it opens on the right map wherever the other person is.

• Navy — shadow cast by a building.

The current test shows building shadows. Tree shade is excluded.

No. Tree shade is excluded from the map, verdicts and recommendations while its accuracy is being reviewed.

Really. We measured it: across 2,215 points in the Georgia test area, 63% are shaded at 8:45 AM and 63% at 6:30 PM — but only 20% at noon. Low sun makes long shadows, even from small buildings and trees. Noon is when shade is scarce, and that's where the app earns its keep.

After sunset everything is in shade, so the map says so — the overlay darkens and the verdict reads All night shade. It comes back at sunrise.

That's the loading indicator (screen readers hear “Loading this area”). It means the map is fetching building data. It disappears when the data arrives; a verdict waits for nearby data too.

The building-shadow model predicts when direct sun reaches this spot. Check the surroundings: estimated heights, missing structures and terrain can change the result.

At least one building in that calculation has an estimated height rather than a measured one. Estimates tend to run tall (when we checked, over-tall guesses outnumbered under-tall ones by about two to one), which means real shade can end earlier than the time shown. Tap the badge to see which buildings.

Because “nothing is blocking the sun here” is the verdict most likely to flip if a nearby height is wrong. A sunny verdict with estimated buildings around it is the least certain kind, so it gets a badge too. ✓ Measured building heights means everything involved is measured — that's the most reliable state.

You tapped a building, not a parking spot. We won't call the inside of a building “shaded” — it would read as the best spot on the map, and you can't park there. Use Find the best spot nearby instead.

That's a parking structure. There's a roof, so the sun doesn't matter and there's no “until” time.

There's no building data within 300 metres of that point. We'd rather say nothing than guess — a guessed verdict that sends you to a sunny spot is worse than no verdict.

The mapped parking candidates did not offer enough additional shade over your selected stay. Your selected spot stays on the map. This does not mean there is no shade nearby; missing parking information and loading failures have separate messages.

It's at the bottom of the verdict card, so it's there whenever you've tapped a spot — but it's hidden while the card is collapsed. Dragging the time slider collapses the card to a single line (Sun until sunset · More ⌃); tap that line or drag the card back up and the button returns.

Use Find shade to search, or Recheck nearby after a recommendation. Set Parking duration first: drag its sun-and-shade bar in 15-minute steps, or all the way to sunset. The search uses that whole stay, starting at the time selected on the bottom slider.

It checks up to 60 candidates within about 150 metres of your selected spot, using mapped ground-level parking areas. Candidates too close to buildings, area boundaries or mapped road and path centerlines are filtered out. It does not invent roadside parking or fill missing parking data with arbitrary map points.

It follows the sun and building shadows over your selected parking duration, including shade that arrives later. A candidate needs at least 20 minutes of modeled shade and at least 15 minutes less modeled sun than your selected spot. These thresholds are product choices, not guarantees of temperature reduction or measured timing accuracy. Eligible candidates in the same parking area come first; where shade totals are within 10 minutes of the best in that group, the closer candidate wins.

The comparison ends at sunset. Changing the selected time or parking duration makes a recommendation stale and offers a recheck. If no suitable improvement is found, your pin stays where it is. Unknown parking or building data is reported separately.

The search uses mapped parking areas and excludes those marked private, no-entry or permit-only. Customer-only areas are considered only when your search starts inside that same area, and the result keeps the customer-only condition visible.

A mapped parking area is not a surveyed individual stall. It can include unmapped aisles or landscaping, and access tags may be missing or outdated. The app cannot confirm an empty space, maneuvering room or your permission to use a particular space. The result therefore says Check marked spaces; use the actual marked spaces and posted conditions.

If it sends you somewhere clearly unparkable, tell us where — that's exactly the kind of report that decides what we fix next.

Every tap is recalculated for that exact moment. As the minutes pass, the sun moves and the answer moves with it — usually by a minute or two. If it jumps by half an hour without you touching the slider, that's worth reporting.

Tap the little pill next to the time — the one that reads ↺ +2h 10m. It jumps the slider back to the current time.

Yes. Every time on the map is shown in your phone's time zone, whoever shared it and wherever the spot is. If the spot sits in a different time zone from you, the map adds your time after the time so there's no guessing — the shade itself is happening at the spot, but the clock you read is yours.

Shared links always recalculate for right now — a line under the verdict says Shared at 3:10 PM — recalculated for right now. We never show an old verdict as if it were current.

You opened a link that pins a specific time (used for checking the app against reality). Calling that “Now” would present an old answer as a live one, so it's labeled differently on purpose.

The sun's position is exact (astronomy is reliable like that). Building heights are the variable — where the data is measured, verdicts are solid; where it's estimated, we label it. We don't give a margin of error, because we haven't found an honest way to measure one: checking how many minutes off a verdict is would mean watching the same spot from sunrise to sunset, over and over. We'd rather tell you what the uncertainty depends on than hand you a number we can't stand behind.

Yes, and we checked rather than assumed. The app's sun is compared against an independent calculation — the same equations the US weather service publishes for its solar calculator — at three cities, across the equinoxes, the solstices and both daylight-saving switch days, at every hour of daylight. The two agree to within a third of a degree everywhere, and daylight-saving days are no different from any other day. For shade, a third of a degree moves a shadow's edge by about the width of your hand.

The clock isn't a weak point either: the sun is computed from the exact instant, not from a local time that could be off by an hour. If a verdict is wrong, the cause is almost never the sun — it's a building height.

Three places, in order of trust. Measured: aerial laser scans (the same public survey data that gives us tree canopy) — the roof height of each building, read directly. Recorded: a height or storey count someone wrote into the public map. Estimated: when neither exists, a typical height for that kind of building, or failing that, a guess from the size of its footprint.

The estimate is the weak one, and we say so — that's the Some heights estimated line. A footprint doesn't know how tall a building is: the same slab of ground holds a one-storey shop or a five-storey block, and shadow length grows in step with height. Where our estimates have been checked against the scan, they mostly ran low, so the map showed less shade than was really there — which is why we're replacing estimates with measured heights area by area. Every building carries a note of which of the three it is, so we always know how much to trust it.

The current test uses buildings only. A real tree or shade sail can provide extra shade that the map does not predict.

Not yet. The ground is treated as flat. On a slope, shadows are a little longer or shorter than we draw them.

Yes. When the forecast says 60% cloud or more, we mention that shade won't matter much today. The verdict itself is still the geometric one — where the shadow would fall if the sun were out.

Email hello@shadespot.app, or if you park downtown, tell the maker directly — 15 minutes, and you pick what we fix next.

Most of the surprising things are answered above. The three that are always worth reporting: a “shaded” spot that's actually sunny, the app not loading or never giving a verdict, and a time that can't be right (a sunset at 10 PM, a verdict dated yesterday).