Alt-data · satellite × retail

Read the quarter from orbit.

Lotwyse counts the cars in a chain's parking lots from overhead imagery and turns fill rate into a demand signal — so you watch the beat or the miss take shape weeks before the earnings call.

Weekly refresh Site-level detail Signal before the print
SUPERSTORE · LOT 04 LIVE SCAN
Cars detected
0
vs 90-day avg
Implied read
Signal cadence Pass in progress
  • Count the cars
  • Skip the guesswork
  • Fill rate → demand
  • Weekly imagery pass
  • Signal before the print
The premise

Demand shows up in the lot first.

A busy parking lot is a busy store, and a busy store is a good quarter. Lotwyse watches that from above and hands it to you as numbers — long before the company reports.

🛰️
OVERHEAD

Cars, counted from space

Every pass over a location, our models find and tally the vehicles in the lot. No surveys, no panels, no reliance on what a company chooses to disclose — just what's physically parked there.

📈
DEMAND

Fill rate becomes a trend

A single count is noise. Weeks of them, benchmarked against a location's own history, become a footfall trend that tracks revenue at the chain level — up, flat, or fading.

⏱️
EARLY

Ahead of the print

The signal lands while the quarter is still open. You form a view on beat, in-line, or miss with weeks of runway — not thirty seconds after the press release drops.

The number is public for ninety seconds a quarter. The lot is public every single day.

— The Lotwyse premise
The shape of it

What the beta already reaches.

0
retail sites in coverage
0
imagery refresh cadence
0
median lead on the print
0
to trial the beta
InputWeekly satellite imagery of lots
ModelVehicle detection + site benchmarks
OutputChain-level demand signal
TimingWeeks before the earnings call

One dumb, durable fact: full lots print well.

Most consumer research is a survey of what people say they'll do. Lotwyse skips the asking. It measures the one thing that doesn't lie — how many cars actually showed up — and rolls thousands of those measurements into a read on the chain.

The pixels don't have a quarter to protect and can't be talked up on a call. They just show the lot as it is, week after week, so your model is built on behavior instead of narrative.

Trading the print blind?

Request a beta seat. We onboard desks in small waves so every one gets the raw feed and a human to walk it through.