Skip to content
Air Automations
All work

Radical Storage · B2C travel · 14,552 listings · 2026

The listings sweep that found 3,271 storefronts invisible on Google

Radical Storage could only see health data for about 3% of its listings. We built a sweep that checks all 14,552 of them — and ranks what to fix by the revenue sitting behind each one.

Radical Storage

14,552

listings under continuous health monitoring, up from ~400

Context

Radical Storage's foot traffic depends on being found on Google Maps. Across 14,552 listings in dozens of countries, that visibility is managed through a listings platform — and when a listing quietly goes inactive, nothing breaks loudly. It just stops appearing, and the bookings stop with it.

The problem

The ranking tool the team relied on scanned around 400 listings — roughly 3% of the network — because every scan costs credits. Everything outside that sample was a blind spot. Nobody could answer the basic question: how many of our listings are actually live right now, and which of the broken ones are costing us the most money?

What we built
  • Built an incremental, resumable sweep that reads the health status of all 14,552 listings straight from the listings API — no ranking credits spent. It works through the book at 3,000 a day, checkpoints as it goes, then keeps rolling so the picture stays current.

  • Cross-referenced the sweep against contract expiry dates and sync settings, which surfaced a specific failure mode nobody was watching: listings flipping to inactive when a contract end date lapsed — including several flagship city-centre locations.

  • Joined booking revenue from analytics onto each listing, so every problem carries a euro figure. Priorities are scored as visibility gap × revenue at stake, not by whichever listing happens to rank worst.

  • Wrapped it in an operator dashboard — headline numbers, a 'needs your attention' list, dead listings, and plain-language fixes — with a daily job that re-checks expiries and a weekly ranking cycle. Read-only by design: the system proposes, a human approves.

How it fits together

14,552 listings

the whole book, not a 3% sample

Resumable sweep

3,000 a day, checkpointed, then rolling

Expiry & sync check

why a listing went dark, not just that it did

× revenue at stake

bookings joined onto each listing

  • dead listings
  • needs your attention
  • ranked priorities
Results

14,552

listings under continuous health monitoring, up from ~400 (3%)

3,271

silently dead listings surfaced — invisible on Google until the sweep found them

€1.9M/yr

of booking revenue mapped onto listings, so fixes are ranked by money at stake

Stack
  • Local Falcon
  • Uberall API
  • GA4 Data API
  • Python
  • Flask
  • Railway

More on what we do for

Multi-location hospitality

See solutions