SourcingConcentration

How much of this category sits in one place?

Sourcing Concentration plots your category spend against where your suppliers physically are, once a year, and draws the clusters: the industrial park, flood plain or port catchment that too many of them share.

GSD
0.5–2 m (High-res)
Cadence
Annual
Platform
High-res satellite
Spectral
RGB
Schematic plot: supplier sites drawn as dots, with three clusters marked C-01 shared parkC-02 lowgroundC-03 port approachN
Fig. 1Schematic of how a draft marks clusters. Invented dots on a blank grid, not supplier data and not a map.

The vendor list that nobody has drawn

Before the annual category review, the category manager exports the vendor list for one category and adds spend. Then comes the part nobody likes. The addresses go onto a map in batches, or someone reads down the column and says "that's the same area, isn't it?" The remit-to address is often a head office or a lockbox in another state, so the first plot is wrong and gets redone from plant addresses, if the team has them. Spend sits in one tab and geography in another, and the join is done by eye.

What the risk committee wants is simple to ask and slow to answer: if one place goes down, how much of this category goes with it? One park with a shared substation, one river valley that floods together, one port every supplier ships through. These only show up when someone has drawn the suppliers on the ground, and in most years nobody has. The cluster gets noticed after a disruption, in the post-mortem.

Sourcing Concentration does the drawing. You send the tagged supplier list, we locate each site in current imagery, and you get the clusters on a plan with the biggest ones ranked first.

Container cranes at a port terminal at night
Fig. 2Container cranes at a terminal. A stock photo, not imagery from a draft.

The draft

An early-access draft is one package per category, once a year. What follows is what we are building toward and offering to scope with the first few categories, not a product on a shelf.

  1. 2.1Concentration mapMap image

    A schematic plot of your supplier sites for the category, clustered by physical proximity, sized by the number of suppliers and, where you send spend, by spend. Made for the committee deck.

  2. 2.2Ranked clustersCSV

    Each cluster with its supplier count and share of the spend you supplied, ranked from the heaviest down, keyed to your supplier IDs so it joins back to the vendor master.

  3. 2.3Geography notesShort notes

    For each cluster, the physical feature it sits on: a shared industrial park, low ground beside a river or coast, a common port. A planner's note on what the imagery shows, not a hazard rating.

How it works

  1. 3.1Send the tagged list

    You send the supplier list for one category, with supplier ID, the address of the site where the goods are made if you have it, and your spend if you want it weighed.

  2. 3.2We locate each site

    We find each supplier site in current high-resolution satellite imagery. A site we can't place with confidence is left off the draft and listed.

  3. 3.3We group by ground

    Sites are grouped by physical proximity and by the shared geography around them: the same park, the same low ground, the same port approach.

  4. 3.4The draft comes back

    Map, ranked clusters and geography notes, once a year. If the supplier list changes materially, ask and we redraw it.

Before the committee meets

Use the draft as the evidence behind a dual-sourcing or diversification proposal. It shows where the weight sits. The decision about what to do about it, and the cost of qualifying a second source, stays with you. It does not replace spend analytics: you bring the spend, we bring the ground.

Limits

  1. L1Annual, not live

    One draft a year. It is not a monitoring feed and says nothing about this week.

  2. L2Resolution

    At 0.5 to 2 m a pixel, a factory building or a warehouse roof is usually visible, and a cluster of them can be read as a cluster. Small units at the coarse end of that range may be hard to pick out. Fittings, stock and equipment inside a yard do not. We locate a site; we don't read what is made on it.

  3. L3Who occupies a building

    Imagery shows a building, not the name on the lease. In a multi-tenant park we place the site you gave us; we can't confirm from above which unit is the supplier's.

  4. L4Clusters are proximity, not boundaries

    Clusters are drawn from visible nearness and shared ground, not from administrative or cadastral boundary data we don't hold. Read "one park" as "these sites sit together", not as a legal parcel.

  5. L5Sites we can't place

    A site that can't be located with confidence, because of a vague address, cloud or heavy shadow, is excluded and listed. The draft is never padded with guesses, which also means it may not cover your whole list.

  6. L6Not a risk score

    No probability, no loss estimate, no rating of the flood plain. The draft shows where suppliers are bunched.

Who it's for

For

  • Category managers who own a supplier list and have to show one category's concentration to a risk committee.
  • Supplier risk teams who want a map of the sourcing footprint behind a dual-sourcing case.

Not for

  • Teams that need continuous monitoring of supplier locations or live alerts.
  • Teams that need a hazard model, a flood-risk rating or a loss estimate.
  • Teams looking for tier-2 and tier-3 supplier discovery. You bring the list; we don't find suppliers for you.

Questions before a draft

Is this a live monitoring tool?

No. It is an annual draft, redrawn on request if your supplier list changes materially.

How precise is "one industrial park"?

Clusters come from visible proximity and shared infrastructure on the ground, not from boundary data we don't hold. Treat the cluster as a strong prompt to look closer.

What about suppliers with unclear addresses?

If we can't place a site with confidence, we leave it off the draft and list it for you. We don't guess.

Does it replace our spend analytics?

No. You bring the category spend; we bring the geography.

Does it give us a concentration percentage?

It ranks clusters and shows their share of the spend you provided. It is a reading of your own list, not a benchmark and not a risk score.

How do we check it against what we know?

Pick a category where you already know a few sites share a park or a flood plain and see whether the draft finds them. Then judge whether it earns a place in the next review.

How does it fit with what we run?

It comes back as a map image, a CSV keyed to your supplier IDs and short notes. We don't connect to your ERP or procurement suite.

Who owns the draft and the imagery?

Terms for the deliverables and the underlying imagery go in the early-access agreement before any work starts. Ask early if legal needs to review them.

How is early access priced?

It depends on the category and the number of sites. Tell us the category and roughly how many suppliers, and a person will scope it with you.

Get your category drafted.

Tell us which category and roughly how many suppliers, and when your next risk committee review is. We'll tell you plainly whether a draft can be ready for it.

Request early access