Three suppliers, forty miles apart, carry sixty percent of revenue.

Executive summary

The network is small, entirely domestic, and structurally concentrated. Two furniture suppliers in the San Francisco Bay Area are the sole source of 46% of product revenue, and the single Los Angeles distribution center ships 54% of it and performs 92% of manufacturing. Delivery performance has been excellent on paper, but the underlying data cannot yet support that conclusion.

Context and method

The company

TD3016323 is a US multi-channel retailer and light manufacturer operating three subsidiaries in a single currency. Product lines derived from the item master and sales history: furniture and mattresses (Estes Park, Ascend and Baja casegoods; Contour Rhapsody Breeze and Patriarch Luxury Firm mattresses), apparel and leather goods (jackets, satchels, watches, backpacks), beauty (hair care and cosmetics), and a small but growing assembled-goods line (bicycle sub-assemblies, keyboards and flex circuit boards) built at the Los Angeles distribution center since June 2026.

Distribution runs through two DCs (Los Angeles, Miami) serving a small number of large B2B customers (~$6K average order), and two stores (San Francisco, New York) serving many small transactions (~$170). A Chicago DC holds $43K of stock but recorded no sales in the period.

Method

Network map

The structural map below is built from actual inbound receipt lanes (vendor → site), transfer orders (site → site) and outbound sales (site → customer segment) in the period. Line weight follows receipt count; red marks the two sole-source suppliers whose loss cannot currently be covered.

Node register

† Revenue attributed by preferred-vendor flag. Hestra and Flexsteel are flagged preferred on items that Apparel Co and Broyhill actually fulfilled (Flexsteel received $3K of POs against $71K of attributed revenue) — they are alternates on paper, unproven at volume.

Critical nodes

Criticality combines four lenses: single-source dependency, downstream blast radius (revenue and sites affected), concentration, and estimated time to recover or qualify a replacement. Inventory cover is shown as time-to-impact — the days before an outage reaches customers, assuming demand continues at the 12-month run rate.

‡ Time-to-recover figures are industry-norm estimates, not measured. Time-to-impact is aggregated by vendor; individual SKUs will stock out earlier than the aggregate suggests.

Interdependencies

Bedline and Broyhill are each other's only backup. The Estes Park Queen Poster Headboard ($27K revenue) is the single dual-sourced furniture SKU — and its two vendors are the two most critical nodes in the network, 15 miles apart.

The assembly chain is three deep. AS_MBK001 requires AS_WHL001 (×2), which requires AS_PHA0001, which requires BLT001, CAP001 (Crown) and NUT001, TYR001 (Johnson). A single late Johnson or Crown shipment stalls three assembly levels. All 24 purchased components have exactly one vendor.

The beauty trio is the healthy pattern. Lotion Co, Mac Oca and Health & Beauty Supplies cross-source every item between them: zero single-source SKUs, four sites each. This is the template the furniture lines should follow.

Chicago is the unused hedge. Location 8 holds $43K of stock, receives Crown components and transfers from LA, and has already built one SAF001 work order — but shipped nothing in twelve months.

Risk heatmap

Each supplier is scored 1 (low) to 5 (high) on three likelihood dimensions and one impact dimension. The likelihood dimensions are combined with explicit weights; impact is scored separately so the two are never conflated.

Underlying scores

Action plan

Ten actions, ordered by risk reduction per unit of effort. Quick wins are data and configuration work that can start this month; the two strategic items address the only two risks that no amount of inventory can cover.

Assumptions, data gaps and next steps

Assumptions

Data gaps

Next steps

1. Approve actions #1–#4 (data and configuration). Saved searches, an OTIF Suitelet, and itemvendor updates can be drafted with dry-runs for approval.
2. Issue a supplier questionnaire to Bedline, Broyhill and Generation N: manufacturing location, sub-tier sourcing, business-continuity plan, capacity headroom. This converts the Tier-2 gap into data within one quarter.
3. Re-run this heatmap in 90 days. Once real receipt dates exist, the likelihood axis becomes evidence-based and the 40% delivery weight can rise to 50%.

Source queries

All figures derive from the SuiteQL below, run against the production account on 16 September 2026. Period bounds: 2025-09-01 ≤ trandate < 2026-09-01. Queries follow the account's house style: tl.subsidiary <> 4 excludes the elimination subsidiary; mainline = 'F' AND taxline = 'F' isolates item lines; amounts are wrapped in ABS().

SELECT v.id, v.entityid, v.companyname, v.isinactive, BUILTIN.DF(v.category) AS category,
       v.subsidiary, v.terms, ea.country, ea.state, ea.city
FROM vendor v
LEFT JOIN entityaddress ea ON ea.nkey = v.defaultbillingaddress
ORDER BY v.entityid

SELECT t.entity AS vendor_id, t.type, COUNT(DISTINCT t.id) AS txn_count,
       ROUND(SUM(t.foreigntotal), 2) AS total
FROM transaction t
WHERE t.type IN ('VendBill', 'PurchOrd')
  AND t.trandate >= TO_DATE('2025-09-01', 'YYYY-MM-DD')
  AND t.trandate <  TO_DATE('2026-09-01', 'YYYY-MM-DD')
GROUP BY t.entity, t.type
ORDER BY t.type, SUM(t.foreigntotal) DESC

SELECT iv.item, i.itemid, i.itemtype, i.class, i.matrixtype, i.parent,
       iv.vendor, v.entityid AS vendor_name, iv.preferredvendor, iv.subsidiary, iv.purchaseprice
FROM itemvendor iv
JOIN item   i ON i.id = iv.item
JOIN vendor v ON v.id = iv.vendor
WHERE i.isinactive = 'F'
ORDER BY iv.item, iv.vendor

SELECT t.entity AS vendor_id, tl.item, i.itemid, i.itemtype, i.class,
       COALESCE(i.parent, i.id) AS family,
       COUNT(DISTINCT t.id) AS po_count, ROUND(SUM(ABS(tl.netamount)), 2) AS po_amount,
       SUM(ABS(tl.quantity)) AS qty, COUNT(DISTINCT tl.location) AS n_locations
FROM transaction t
JOIN transactionline tl ON tl.transaction = t.id AND tl.mainline = 'F' AND tl.taxline = 'F'
JOIN item i ON i.id = tl.item
WHERE t.type = 'PurchOrd'
  AND t.trandate >= TO_DATE('2025-09-01', 'YYYY-MM-DD')
  AND t.trandate <  TO_DATE('2026-09-01', 'YYYY-MM-DD')
GROUP BY t.entity, tl.item, i.itemid, i.itemtype, i.class, COALESCE(i.parent, i.id)
ORDER BY SUM(ABS(tl.netamount)) DESC

-- PO lines
SELECT t.id AS po_id, t.entity AS vendor_id, t.trandate AS po_date, t.duedate AS po_due, t.status,
       tl.uniquekey AS lk, tl.item, ABS(tl.quantity) AS qty, tl.quantityshiprecv AS qty_recv,
       tl.expectedreceiptdate AS exp_date, tl.location
FROM transaction t
JOIN transactionline tl ON tl.transaction = t.id AND tl.mainline = 'F' AND tl.taxline = 'F'
WHERE t.type = 'PurchOrd' AND tl.item IS NOT NULL
  AND t.trandate >= TO_DATE('2025-09-01', 'YYYY-MM-DD')
  AND t.trandate <  TO_DATE('2026-09-01', 'YYYY-MM-DD');

-- PO → item receipt links (nexttransactionlink is empty in this account)
SELECT DISTINCT l.previousdoc AS po_id, l.nextdoc AS rcpt_id, r.trandate AS rcpt_date, r.entity AS vendor_id
FROM nexttransactionlinelink l
JOIN transaction r ON r.id = l.nextdoc
WHERE r.type = 'ItemRcpt' AND l.linktype IN ('ShipRcpt', 'PickPack', 'OrdBill');

-- Reducer: per vendor — lead time = PO date → last receipt; on-time = last receipt ≤ due date;
-- fill = Σ qty_recv ÷ Σ qty; open past-due = unlinked POs with due < today.

SELECT tl.item, i.itemid, i.itemtype, i.class, COALESCE(i.parent, i.id) AS family,
       ROUND(SUM(ABS(tl.netamount)), 2) AS revenue, SUM(ABS(tl.quantity)) AS qty,
       COUNT(DISTINCT t.entity) AS n_customers
FROM transaction t
JOIN transactionline tl ON tl.transaction = t.id AND tl.mainline = 'F' AND tl.taxline = 'F'
JOIN item i ON i.id = tl.item
WHERE t.type IN ('CustInvc', 'CashSale') AND tl.subsidiary <> 4
  AND t.trandate >= TO_DATE('2025-09-01', 'YYYY-MM-DD')
  AND t.trandate <  TO_DATE('2026-09-01', 'YYYY-MM-DD')
GROUP BY tl.item, i.itemid, i.itemtype, i.class, COALESCE(i.parent, i.id)
-- joined in-worker to itemvendor (preferredvendor = 'T', subsidiary = 2), itemmember (kits) and A4

-- Revenue by fulfilment location
SELECT tl.location, loc.name AS loc_name, tl.subsidiary, t.type,
       ROUND(SUM(ABS(tl.netamount)), 2) AS revenue, COUNT(DISTINCT t.id) AS txns,
       COUNT(DISTINCT t.entity) AS customers
FROM transaction t
JOIN transactionline tl ON tl.transaction = t.id AND tl.mainline = 'F' AND tl.taxline = 'F'
LEFT JOIN location loc ON loc.id = tl.location
WHERE t.type IN ('CustInvc', 'CashSale') AND tl.subsidiary <> 4 AND tl.item IS NOT NULL
  AND t.trandate >= TO_DATE('2025-09-01', 'YYYY-MM-DD')
  AND t.trandate <  TO_DATE('2026-09-01', 'YYYY-MM-DD')
GROUP BY tl.location, loc.name, tl.subsidiary, t.type;

-- On-hand by item × location (value column is onhandvaluemli in this account)
SELECT il.item, i.itemid, il.location, il.quantityonhand, il.quantityavailable,
       il.onhandvaluemli, il.quantityonorder, il.reorderpoint, il.preferredstocklevel
FROM inventoryitemlocations il
JOIN item i ON i.id = il.item
WHERE il.quantityonhand > 0

SELECT v.entityid AS vendor, tl.location, loc.name AS loc_name,
       COUNT(DISTINCT t.id) AS receipts, SUM(ABS(tl.quantity)) AS qty
FROM transaction t
JOIN transactionline tl ON tl.transaction = t.id AND tl.mainline = 'F' AND tl.taxline = 'F' AND tl.item IS NOT NULL
JOIN vendor v ON v.id = t.entity
LEFT JOIN location loc ON loc.id = tl.location
WHERE t.type = 'ItemRcpt'
  AND t.trandate >= TO_DATE('2025-09-01', 'YYYY-MM-DD')
  AND t.trandate <  TO_DATE('2026-09-01', 'YYYY-MM-DD')
GROUP BY v.entityid, tl.location, loc.name
ORDER BY v.entityid, tl.location

SELECT tl.location AS from_loc, t.transferlocation AS to_loc,
       COUNT(DISTINCT t.id) AS orders, SUM(ABS(tl.quantity)) AS qty
FROM transaction t
JOIN transactionline tl ON tl.transaction = t.id AND tl.mainline = 'F' AND tl.taxline = 'F' AND tl.item IS NOT NULL
WHERE t.type = 'TrnfrOrd'
GROUP BY tl.location, t.transferlocation
ORDER BY COUNT(DISTINCT t.id) DESC

SELECT tl.item, i.itemid, tl.location, loc.name AS loc_name,
       COUNT(DISTINCT t.id) AS work_orders, SUM(ABS(tl.quantity)) AS qty,
       MIN(t.trandate) AS first_wo, MAX(t.trandate) AS last_wo
FROM transaction t
JOIN transactionline tl ON tl.transaction = t.id AND tl.mainline = 'T'
JOIN item i ON i.id = tl.item
LEFT JOIN location loc ON loc.id = tl.location
WHERE t.type = 'WorkOrd'
GROUP BY tl.item, i.itemid, tl.location, loc.name
ORDER BY COUNT(DISTINCT t.id) DESC

SELECT b.id AS bom_id, b.name AS bom_name, brc.item AS component, c.itemid AS component_name,
       c.itemtype, brc.quantity AS qty_per
FROM bom b
JOIN bomrevision          br  ON br.billofmaterials = b.id
JOIN bomrevisioncomponent brc ON brc.bomrevision = br.id
JOIN item c ON c.id = brc.item
ORDER BY b.name, c.itemid

SELECT c.id, c.itemid, c.itemtype,
       (SELECT COUNT(DISTINCT iv.vendor) FROM itemvendor iv WHERE iv.item = c.id) AS n_vendors,
       (SELECT MIN(v.entityid) FROM itemvendor iv JOIN vendor v ON v.id = iv.vendor WHERE iv.item = c.id) AS a_vendor
FROM item c
WHERE c.id IN (SELECT brc.item FROM bomrevisioncomponent brc)
ORDER BY n_vendors, c.itemid

SELECT v.entityid AS vendor, t.status, COUNT(DISTINCT t.id) AS pos,
       ROUND(SUM(ABS(t.foreigntotal)), 2) AS value,
       SUM(CASE WHEN t.duedate < TRUNC(SYSDATE) THEN 1 ELSE 0 END) AS past_due
FROM transaction t
JOIN vendor v ON v.id = t.entity
WHERE t.type = 'PurchOrd' AND t.status IN ('A', 'B', 'E')
GROUP BY v.entityid, t.status
ORDER BY SUM(ABS(t.foreigntotal)) DESC

SELECT t.type, i.itemtype, COUNT(DISTINCT t.id) AS txns, COUNT(*) AS lines,
       ROUND(SUM(ABS(tl.netamount)), 2) AS revenue, MIN(t.tranid) AS sample_tranid, MIN(t.memo) AS sample_memo
FROM transaction t
JOIN transactionline tl ON tl.transaction = t.id AND tl.mainline = 'F' AND tl.taxline = 'F'
JOIN item i ON i.id = tl.item
WHERE t.type IN ('CustInvc', 'CashSale') AND tl.subsidiary <> 4 AND tl.location IS NULL
  AND t.trandate >= TO_DATE('2025-09-01', 'YYYY-MM-DD')
  AND t.trandate <  TO_DATE('2026-09-01', 'YYYY-MM-DD')
GROUP BY t.type, i.itemtype
-- → 27 invoices, $681,677, memo "TEST — Copper Software" (Service items): excluded

Key derived figures

Verification log

Claims that were tested against the account after the initial analysis, and what changed.