Terms vs. Behavior Audit — Customer Payment Discipline

1Executive Summary

The receivables book is in excellent shape on paper — but the interesting findings are in the exceptions, the trend line, and the free float being left on the table.

10.2

Avg days to pay

Simple average, all 1,472 applications

18.7

$-weighted days

Big invoices settle slower than small ones

96%

Paid within terms

1,414 of 1,472 applications on time or early

Jun ’26

Trend break

Avg days-to-pay jumped ~3 → ~18 and stayed there

⚠ Finding A — Payment behavior deteriorated sharply in June 2026

From Dec 2024 through May 2026 the portfolio paid in a remarkably tight 3–5 day average. In June 2026 the average jumped to 17.6 days (weighted: 23.9) and has stayed elevated through August (weighted: 28.7 — right at the Net-30 boundary). Application volume also quadrupled (≈35/mo → 130–189/mo) with ~$790K–$960K/month in settled dollars. This is either a real behavioral shift in the book, a change in mix (new large customers on terms), or a change in how payments are being entered. It deserves investigation before it becomes a DSO problem.

⚠ Finding B — 6 payments are dated before their invoices — including future-dated invoices

Three Alpha Demand invoices (INV1660–1662) are dated Aug 31, 2026 — two weeks in the future — yet were "paid" Aug 1–3. Frutti Di Mare (INV1721, −17 days), Buzzie's Sandwiches (INV503, −4), and Smith Pacific Northwest (#171, −2) show the same inversion. These are data-entry / process artifacts (deposits or prepayments applied to later-dated invoices), but they distort DSO math and are exactly what an auditor's cutoff test would flag.

ℹ Finding C — Your single biggest payer has no terms assigned

Alpha Demand (customer id 1615) accounts for 1,022 of 1,472 paid invoices ($195K) and pays in 6.2 days on average — yet the customer record has no payment terms. Every aging bucket, dunning rule, and due-date calculation treats these invoices as due-on-receipt. Assigning real terms (their behavior fits Net 7–Net 15) makes reporting honest at zero cost.

✓ Finding D — Terms compliance is excellent; the leverage runs the other way

Only 5 of 49 Net-30 customers average past 30 days, and the worst is just +3.4 days. Meanwhile 21 customers pay 15+ days early, including Whole Markets ($302K @ 16.9d), Acme Produce ($221K @ 9.4d), and Dubois Candy ($199K @ 10.8d). That's prepaid float you're receiving without offering any early-pay discount — pure negotiating capital.

2Methodology

Days-to-pay is measured at the payment-application level, not the invoice level — each application of a customer payment to an invoice is one observation, weighted by the dollars applied. This handles partial payments correctly and avoids the "last payment defines the invoice" distortion.

transaction (CustInvc)→ nexttransactionlinelink (linktype='Payment')→ transaction (CustPymt)＋ customer.terms → term

Metric definitions

Days to pay	`TRUNC(payment date) − TRUNC(invoice date)` per application. Measured from invoice date, not due date — this is what "Net 30" promises against.
Avg days	Simple mean across a customer's applications. Every invoice counts equally.
Weighted days	`Σ(days × $applied) ÷ Σ($applied)`. When weighted ≫ simple, the customer's large invoices are the slow ones — the pattern that becomes a collections problem.
Variance	Avg days − assigned terms days. Positive = pays late; negative = pays early.

Assumptions & scope decisions read before quoting numbers

≥ 3 invoices	Customer-level table requires at least 3 paid invoices — below that, an "average" is noise. 51 of 100 paying customers qualify; excluded customers still count in portfolio totals.
Paid invoices only	This measures behavior on invoices that got paid. Currently-open and never-paid invoices are out of scope (that's the A/R aging's job). Survivorship bias: true behavior is slightly worse than shown.
Payments only	`linktype='Payment'`, CustInvc→CustPymt, positive applied amounts. Credit memos (49 in the account) and journal applications are excluded — they represent adjustments, not payment behavior.
Link-table dates	Dates come from `nexttransactionlinelink.previousdate / nextdate` (invoice and payment tran dates). Negative values were kept and reported as anomalies (Finding B), not silently dropped.
Currency	`foreignamount` = transaction currency. Dollar totals assume a single-currency book; if multi-currency is active, cross-currency sums are approximate.
No-terms customers	Treated as 0-day terms for variance math, and flagged separately (Finding C) rather than buried in averages.
"Within terms"	Application counted as compliant when days-to-pay ≤ the customer's terms days. For no-terms customers this is a harsh ≤0 test — the true portfolio compliance rate is therefore understated.

3Portfolio View

How the 1,472 payment applications distribute, and how behavior has moved over 33 months.

Days-to-pay distribution count of payment applications per bucket

Note the bimodal shape: a fast cluster (0–14 days, 82%) and a "pay near the due date" cluster (22–30). The 15–21 day valley (only 6 payments) suggests two distinct payer populations, not one continuum.

Days-to-pay trend by payment month Dec 2023 – Aug 2026, capped at 30 for scale

Simple avg days $-weighted avg days Jun–Aug ’26: weighted avg hits 28.7 — brushing the Net-30 ceiling

4Customer-Level Results

All 51 customers with ≥3 paid invoices, ranked worst-to-best by variance from assigned terms. The bar shows how far behavior deviates from terms (red = late, green = early).

Terms vs. behavior — full ranking variance = avg days-to-pay − terms days

Customer	Terms	Inv	$ Applied	Avg d	Wtd d	Worst	Variance	Flag
BCP Customer 2	—	5	682	12.6	12.6	24	+12.6	NO TERMS
Alpha Demand	—	1,022	195,132	6.2	6.1	13	+6.2	NO TERMS
Baxter Elementary School	Net 30	5	49,054	33.4	33.4	34	+3.4	LATE
Bonita Inn	Net 30	3	8,677	33.0	33.0	33	+3.0	LATE
Sam's Stop N Go	Net 30	5	21,781	32.4	32.3	34	+2.4	LATE
Snaptags Consulting	Net 30	3	36,570	30.0	30.0	31	0.0	AT TERMS
Photolist Foundation	Net 30	3	47,970	30.0	30.0	31	0.0	AT TERMS
Realpoint Co.	Net 30	3	36,570	30.0	30.0	31	0.0	AT TERMS
Skibox LLC.	Net 30	4	320	29.5	29.5	31	−0.5	AT TERMS
Oozz Incorporated	Net 30	5	134,155	29.0	29.6	31	−1.0	AT TERMS
Riffpedia Company	Net 30	5	39,975	28.8	28.8	31	−1.2	AT TERMS
Ntags Associates	Net 30	3	23,689	28.7	30.4	31	−1.3	AT TERMS
Phasellus Vitae Mauris Inc.	Net 30	3	38,070	28.7	27.5	31	−1.3	AT TERMS
Oyope Industries	Net 30	4	44,971	28.5	27.3	31	−1.5	AT TERMS
Wapp Hardware Sales	Net 30	4	75,980	28.3	28.3	31	−1.8	AT TERMS
Realcube Industries	Net 30	4	9,308	28.3	28.3	31	−1.8	AT TERMS
Rhycero LP	Net 30	5	30,475	28.0	28.0	31	−2.0	AT TERMS
Sem Corporation Associates	Net 30	5	200	28.0	28.0	31	−2.0	AT TERMS
Keller PR	Net 30	6	99,275	28.0	28.0	31	−2.0	AT TERMS
Meedoo Industries	Net 30	3	26,739	27.7	27.8	28	−2.3	AT TERMS
Smith Pacific Northwest Store	Net 30	27	151,613	27.0	20.1	66	−3.0	ERRATIC
Dab's Deli	Net 30	3	7,224	27.0	30.3	40	−3.0	WTD>TERMS
Shuffle's Grocery	Net 30	7	62,395	26.1	25.9	28	−3.9
McEdwards & Whitwell Steakhouse	Net 30	4	13,181	25.0	25.0	25	−5.0
Wikizz Industries	Net 30	7	16,448	24.1	26.4	31	−5.9
Chatter's Candy Counter	Net 30	11	104,009	23.6	19.6	31	−6.4
Red Oak Country Club	Net 30	6	17,455	22.7	22.7	24	−7.3
Nightingale Senior Center	Net 30	11	183,262	22.5	23.5	31	−7.5
Vinder Commercial Cleaners	Net 30	5	92,330	20.0	9.4	25	−10.0
Crescent Street Grille	Net 30	7	8,191	19.6	23.7	27	−10.4
Camido Cocina	Net 30	10	44,845	18.6	24.9	28	−11.4
Frutti Di Mare Restaurant	Net 30	21	130,063	17.4	12.4	40	−12.6	ANOMALY
Whole Markets	Net 30	21	302,227	16.9	19.5	34	−13.1	EARLY $
Abbott's Restaurant	Net 30	14	41,044	14.8	18.6	35	−15.2
DynaCare Health Stop	Net 30	7	36,978	13.9	13.9	15	−16.1
Meetz inc.	Net 30	21	86,279	13.9	26.3	31	−16.1	BIG=SLOW
Webster Grill	Net 30	6	25,492	13.7	13.7	14	−16.3
Buzzie's Sandwiches	Net 30	4	6,008	13.0	20.8	28	−17.0	ANOMALY
Cooper Concessions	Net 30	11	51,582	12.7	20.6	27	−17.3
Magna Janitorial Services	Net 30	8	136,481	12.1	5.8	14	−17.9	EARLY $
Restaurant Wholesale Inc	Net 30	11	94,577	11.6	13.2	114	−18.4	OUTLIER
BCP Customer 3	Net 30	10	1,364	11.6	11.6	23	−18.4
Telescope Knoll Country Club	Net 30	11	21,895	11.4	19.0	24	−18.6
Viva Cafe	Net 30	17	51,294	11.4	18.8	31	−18.6
Dubois Candy Emporium	Net 30	12	199,235	10.8	7.1	18	−19.3	EARLY $
Volutpat Industries	Net 30	6	13,001	10.3	10.3	11	−19.7
Acme Produce Market	Net 30	19	220,849	9.4	8.7	27	−20.6	EARLY $
BCP Customer 1	Net 30	5	616	8.8	8.8	22	−21.2
Underwood Produce Market	Net 30	3	37,089	8.3	5.9	22	−21.7
Skipstorm Seafood	Net 30	6	16,792	5.0	5.4	8	−25.0
Moore Foods	Net 30	3	36,773	4.3	11.7	12	−25.7

5Deep Dives

Restaurant Wholesale Inc — the 114-day outlier customer id 483 · all 11 payment applications

Invoice	Inv Date	Due Date	Inv Total	Payment	Pay Date	Applied	Days
INV622	2026-05-02	2026-06-02	10,428.20	PYMT377	2026-08-24	10,000.00	114
INV1766	2025-11-11	2025-12-10	6,162.76	PYMT1452	2025-11-13	6,162.76	2
INV1767	2025-12-02	2026-01-02	6,279.88	PYMT1453	2025-12-04	6,279.88	2
INV1769	2026-05-18	2026-06-18	4,111.84	PYMT1455	2026-05-20	4,111.84	2
INV1768	2026-01-06	2026-02-05	3,115.06	PYMT1454	2026-01-08	3,115.06	2
INV1825	2025-05-02	2025-06-02	10,264.60	PYMT1511	2025-05-03	10,264.60	1
INV1779	2025-04-22	2025-05-21	10,428.20	PYMT1465	2025-04-23	10,428.20	1
INV1780	2025-10-17	2025-11-17	10,428.20	PYMT1466	2025-10-18	10,428.20	1
INV1781	2026-01-15	2026-02-14	13,093.54	PYMT1467	2026-01-16	13,093.54	1
INV1824	2025-02-28	2025-04-01	10,264.60	PYMT1510	2025-03-01	10,264.60	1
INV1778	2024-11-16	2024-12-15	10,428.20	PYMT1464	2024-11-17	10,428.20	1

Verdict: dispute, not distress

This customer pays in 1–2 days, every single time — except INV622, which sat 114 days and was then settled with a round $10,000 against a $10,428.20 invoice, leaving $428.20 apparently short-paid. A near-perfect payer suddenly short-paying one specific invoice is the signature of a disputed line item or pricing disagreement, not credit deterioration. Also note PYMT377 is dated 2026-08-24 — 8 days in the future relative to this report, another cutoff artifact. Action: confirm the $428.20 residual on INV622 and whether a credit memo is pending.

Payments dated before their invoices 6 applications — Finding B detail

Customer	Invoice	Inv Date	Payment	Pay Date	Applied	Days
Alpha Demand	INV1660	2026-08-31 ⚠ future	PYMT1381	2026-08-01	321.30	−30
Alpha Demand	INV1661	2026-08-31 ⚠ future	PYMT1382	2026-08-02	236.40	−29
Alpha Demand	INV1662	2026-08-31 ⚠ future	PYMT1383	2026-08-03	94.90	−28
Frutti Di Mare Restaurant	INV1721	2026-08-27	PYMT1423	2026-08-10	1,170.00	−17
Buzzie's Sandwiches	INV503	2026-08-18	PYMT239	2026-08-14	1,326.71	−4
Smith Pacific Northwest Store	171	2026-08-15	PYMT211	2026-08-13	4,050.00	−2

6Recommended Actions

Ordered by impact-per-effort. Items 1–3 can be executed directly from this session.

5 MINAssign terms to Alpha Demand (id 1615) and BCP Customer 2 (id 1178)1,027 invoices currently age as due-on-receipt. Alpha Demand's 6.2-day behavior fits Net 7–Net 15; pick the term and it's a one-field update. This alone fixes the biggest distortion in your aging report.
15 MINInvestigate the June 2026 trend breakSlice the Jun–Aug 2026 payment population by customer to see whether the 3→18 day jump is a mix shift (new slow-paying customers), a process change (batch payment entry), or genuine deterioration. Weighted days hit 28.7 in August — one more push and money starts arriving late.
10 MINResolve the INV622 short-pay ($428.20) with Restaurant Wholesale IncModel 1–2-day payer, one 114-day invoice settled at a round $10,000. Confirm the dispute and clear the residual before it fossilizes in aging.
30 MINCorrect the 6 date-inverted applicationsFuture-dated invoices (INV1660–1662 dated Aug 31) applied against earlier payments distort DSO and would be flagged in any audit cutoff test. Likely deposits/prepayments applied to later-dated invoices — reclassify or correct the invoice dates.
STRATEGICMonetize the early-payer cohort21 customers pay 15+ days early on ~$1.4M of volume with no discount incentive. Options: leave it (free float), formalize 2/10 Net 30 selectively where it buys loyalty, or use it as goodwill capital in the next price negotiation. Their credit limits can also safely rise.
WATCHPut Meetz inc. on a quiet watchlistSimple average 13.9 days but dollar-weighted 26.3 — their large invoices are the slow ones. That divergence is the earliest observable signal of a future collections problem.

7Appendix — Source Queries

Every number in this report traces to one of these SuiteQL queries, run live against the account on Aug 16, 2026. Rerun any of them to reproduce or refresh the analysis.

Q1 · Customer-level terms vs. behavior (primary dataset — Section 4)

SELECT
  c.id AS customer_id,
  COALESCE(c.companyname, c.entityid) AS customer,
  COALESCE(tm.name, '(no terms)') AS assigned_terms,
  COALESCE(tm.daysuntilnetdue, 0) AS terms_days,
  COUNT(DISTINCT l.previousdoc) AS invoices_paid,
  ROUND(SUM(l.foreignamount), 2) AS total_applied,
  ROUND(AVG(TRUNC(l.nextdate) - TRUNC(l.previousdate)), 1) AS avg_days_to_pay,
  ROUND(SUM((TRUNC(l.nextdate) - TRUNC(l.previousdate)) * l.foreignamount)
        / NULLIF(SUM(l.foreignamount), 0), 1) AS wtd_days_to_pay,
  MAX(TRUNC(l.nextdate) - TRUNC(l.previousdate)) AS worst_days,
  ROUND(AVG(TRUNC(l.nextdate) - TRUNC(l.previousdate))
        - COALESCE(tm.daysuntilnetdue, 0), 1) AS variance_days
FROM nexttransactionlinelink l
JOIN transaction inv ON l.previousdoc = inv.id
JOIN customer c ON inv.entity = c.id
LEFT JOIN term tm ON c.terms = tm.id
WHERE l.linktype = 'Payment'
  AND l.previoustype = 'CustInvc'
  AND l.nexttype = 'CustPymt'
  AND l.foreignamount > 0
GROUP BY c.id, COALESCE(c.companyname, c.entityid),
  COALESCE(tm.name, '(no terms)'), COALESCE(tm.daysuntilnetdue, 0)
HAVING COUNT(DISTINCT l.previousdoc) >= 3
ORDER BY ROUND(AVG(TRUNC(l.nextdate) - TRUNC(l.previousdate))
  - COALESCE(tm.daysuntilnetdue, 0), 1) DESC

Q2 · Portfolio summary + distribution buckets (Sections 1, 3)

SELECT
  COUNT(*) AS applications,
  COUNT(DISTINCT l.previousdoc) AS invoices,
  COUNT(DISTINCT inv.entity) AS customers,
  ROUND(SUM(l.foreignamount), 2) AS total_dollars,
  ROUND(AVG(TRUNC(l.nextdate) - TRUNC(l.previousdate)), 1) AS avg_days,
  ROUND(SUM((TRUNC(l.nextdate) - TRUNC(l.previousdate)) * l.foreignamount)
        / NULLIF(SUM(l.foreignamount),0), 1) AS wtd_days,
  SUM(CASE WHEN TRUNC(l.nextdate) - TRUNC(l.previousdate)
        <= COALESCE(tm.daysuntilnetdue, 0) THEN 1 ELSE 0 END) AS within_terms,
  -- bucket columns: 0-7, 8-14, 15-21, 22-30, 31-45, 45+ via CASE WHEN ... BETWEEN
  SUM(CASE WHEN TRUNC(l.nextdate) - TRUNC(l.previousdate) <= 7 THEN 1 ELSE 0 END) AS b_0_7
  /* ... remaining buckets elided for brevity — same pattern ... */
FROM nexttransactionlinelink l
JOIN transaction inv ON l.previousdoc = inv.id
JOIN customer c ON inv.entity = c.id
LEFT JOIN term tm ON c.terms = tm.id
WHERE l.linktype = 'Payment' AND l.previoustype = 'CustInvc'
  AND l.nexttype = 'CustPymt' AND l.foreignamount > 0

Q3 · Monthly days-to-pay trend (Section 3 chart)

SELECT
  TO_CHAR(l.nextdate, 'YYYY-MM') AS pay_month,
  COUNT(*) AS applications,
  ROUND(SUM(l.foreignamount), 2) AS dollars,
  ROUND(AVG(TRUNC(l.nextdate) - TRUNC(l.previousdate)), 1) AS avg_days,
  ROUND(SUM((TRUNC(l.nextdate) - TRUNC(l.previousdate)) * l.foreignamount)
        / NULLIF(SUM(l.foreignamount),0), 1) AS wtd_days
FROM nexttransactionlinelink l
JOIN transaction inv ON l.previousdoc = inv.id
WHERE l.linktype = 'Payment' AND l.previoustype = 'CustInvc'
  AND l.nexttype = 'CustPymt' AND l.foreignamount > 0
GROUP BY TO_CHAR(l.nextdate, 'YYYY-MM')
ORDER BY TO_CHAR(l.nextdate, 'YYYY-MM')

Q4 · Restaurant Wholesale Inc drill-down (Section 5)

SELECT
  inv.tranid, TO_CHAR(inv.trandate,'YYYY-MM-DD') AS invoice_date,
  TO_CHAR(inv.duedate,'YYYY-MM-DD') AS due_date, inv.foreigntotal,
  pay.tranid AS payment_num, TO_CHAR(pay.trandate,'YYYY-MM-DD') AS payment_date,
  l.foreignamount AS applied,
  TRUNC(l.nextdate) - TRUNC(l.previousdate) AS days_to_pay
FROM nexttransactionlinelink l
JOIN transaction inv ON l.previousdoc = inv.id
JOIN transaction pay ON l.nextdoc = pay.id
WHERE l.linktype = 'Payment' AND l.previoustype = 'CustInvc'
  AND l.nexttype = 'CustPymt' AND inv.entity = 483
ORDER BY days_to_pay DESC

Q5 · Date-inversion anomaly sweep (Finding B)

SELECT
  COALESCE(c.companyname, c.entityid) AS customer,
  inv.tranid, TO_CHAR(inv.trandate,'YYYY-MM-DD') AS invoice_date,
  pay.tranid AS payment_num, TO_CHAR(pay.trandate,'YYYY-MM-DD') AS payment_date,
  l.foreignamount, TRUNC(l.nextdate) - TRUNC(l.previousdate) AS days
FROM nexttransactionlinelink l
JOIN transaction inv ON l.previousdoc = inv.id
JOIN transaction pay ON l.nextdoc = pay.id
JOIN customer c ON inv.entity = c.id
WHERE l.linktype = 'Payment' AND l.previoustype = 'CustInvc'
  AND l.nexttype = 'CustPymt'
  AND TRUNC(l.nextdate) < TRUNC(l.previousdate)
ORDER BY days

Schema notes for reproducibility

Payment-to-invoice applications in this account live in nexttransactionlinelink with linktype='Payment' (verified: nexttransactionlink and previoustransactionlink carry no CustInvc→CustPymt rows here). Terms come from the term table (daysuntilnetdue); all terms in use are simple day-driven (no datedriven='T' terms encountered, so no end-of-month due-date math was needed). Amounts use foreignamount per the account's SuiteQL exposure. Account contains 1,865 CustInvc, 1,472 CustPymt, 49 CustCred.