Find out which candidate sites look boxed in before anyone books the flight.

The network model says the next DC belongs somewhere in a metro or along a corridor. The broker sends parcels: a greenfield tract, a brownfield lot with an old warehouse on it, a building that is already standing. The analyst works them from the desk. Parcel boundaries go on a map, the county supplies the zoning envelope, a drive-time isochrone shows which stores the site would reach. Whatever imagery is open on the screen is the last word on what the ground looks like, and it may be years old.

Then somebody travels. The site is walked, the access is driven, and the trip finds what the desk could not: the trees that have grown over half the lot, the frontage that is one narrow road, a row of yards next door that already compete for the same drivers. By the time that is known, the flight and the hotel are spent and the broker is waiting on an answer.

Distribution Siting puts a current look between the desk and the trip. It reads each parcel from a recent pass, so the sites that obviously won't work drop out before the travel budget moves.

Results come in formats a network model and a committee deck can take. On a pilot, that covers the candidate parcels you send.

1. 2.1

   Site screening layer

   GeoPackage or shapefile

   Your parcel boundaries, each carrying what the imagery shows: land cover across the lot and the roads that reach it. It opens in ArcGIS or QGIS next to your isochrones.

2. 2.2

   Shortlist scorecard

   CSV

   One row per candidate parcel, with its land-cover mix, the access read and a count of logistics activity nearby. Paste it into the scorecard your network model or committee already uses.

3. 2.3

   Logistics density read

   Map layer and counts

   Yards, large warehouse roofs and parked trailer areas detected around each parcel, so you see how crowded the neighbourhood already is before you stand in it.

4. 2.4

   Flags for what the pass can't see

   In the scorecard

   A parcel under cloud or heavy shadow, or one the pass doesn't cover well, is marked with the reason and stays on your visit list. Nothing is scored on a guess.

1. 1

   Parcels in

   Send the boundaries your broker gave you as a shapefile, GeoJSON or KML, or draw them on a map. Say which ones are the shortlist for this pilot.

2. 2

   Latest usable pass

   We pull the most recent clear, daylight high-resolution pass over each parcel, at 0.5 m or finer. If a parcel is under cloud, it waits for the next clear pass.

3. 3

   Classification

   Land cover, the roads reaching each parcel and the logistics activity around it are classified. Parcels the pass can't read cleanly are flagged instead of scored.

4. 4

   Delivery

   The screening layer, the scorecard CSV and the flag list. We agree the delivery date against your site-visit schedule at the start, then go through the results with you on a call.

It sits after the desk research (parcel lists, zoning maps, isochrones) and before anyone is sent to the site. Use the scorecard to decide which parcels still deserve a trip and to ask the broker sharper questions about the ones that don't look right from above.

The visit stays. Zoning, utilities, title, and what a parcel will cost to build or lease are still decided by your team and your advisers, with people on the ground. If you run a repeat search across several metros, send one shortlist per search and the same layers come back each time.