Nov 2, 2026
Field notes: the week a hundred-site vendor list turned up twelve we'd never checked

Every renewal cycle has a week like this one. Someone pulls the supplier master, cross-references it against last year's register, and finds a gap nobody flagged. This time it was twelve sites out of a hundred-odd on the Tier 1 and Tier 2 footprint, sitting on the list with a city and a country code and not much else.
The pattern is familiar enough that it's worth writing down. A site gets added to the vendor master when procurement signs a contract, usually off a supplier's own address on file. The address goes into the spreadsheet. The spreadsheet gets carried forward from the year before, column for column, and somewhere in that carrying forward the assumption takes hold that if a site was fine last year it's still fine. Nobody goes back and checks the actual coordinates against a flood plain or a fuel load map, because checking a hundred sites by hand against four different hazard layers is a week of someone's life that nobody has budgeted.
So the twelve sat there. Three turned out to be a single supplier's sub-facilities that had been entered under slightly different names, which is its own problem. Two were in a river corridor that had flooded twice in the last decade, confirmed once you looked at where the pin dropped rather than the mailing address on the PO. One was close enough to a wildland interface that the fuel load call was unambiguous once imagery of the surrounding terrain was pulled rather than inferred from a county-level zone map. The rest were a mix: a coastal site worth a storm surge look, a couple on grades that warranted checking for landslide terrain, a few that turned out to be genuinely fine once confirmed, which is also useful to know.
None of this is exotic. It's the gap between what a vendor list says and what a vendor list has had someone look at. A hazard-zone PDF from a regional authority tells you where the zone lines are. It doesn't tell you that Site 47 is eleven meters from a boundary nobody drew carefully enough to settle the question, and it certainly doesn't update itself when a supplier opens a new building three streets over and reuses the old address in your system because that's what's in their letterhead.
The twelve didn't surprise anyone who's run this exercise before. What's worth sitting with is how ordinary the miss was: a routine annual refresh that happened to ask a sharper question this time, whether anyone had confirmed where each site physically sits against current imagery, rather than just checking an address against a flood-zone boundary drawn years ago. A hundred sites is not a large footprint by most standards, and twelve unconfirmed is not an unusual ratio. Scale that to a few thousand sites across a global footprint and the unconfirmed column stops being a rounding error.
What closes the gap is running the full list against flood plain, fuel load, slope and storm surge exposure on a fixed cadence, and getting back a numbered entry per site with the hazard type, the exposure tier and the imagery the call was made against, a register you can defend in a conversation with underwriting instead of one you just inherited. That's the register Supply Chain Hazard Mapping runs once a year, one numbered entry per site, built for exactly this kind of gap.
If your own vendor master has a column nobody's verified in a while, that's worth a closer look before the next renewal arrives.