2.13 — BI/ERP integration stubs (CSV → Excel / Power BI / Looker / API)
Minimal CSV schema (owner/executive views).
time, unit, kpi, m, a, band, U, W, knobs_hash, build_id, stamp
# U,W optional (expose if you want verifiable rollups)
# stamp optional (enable if provenance needed)
Excel / Sheets (paste-ready pseudo-formulas).
# Band from lane (cell formula)
=IFS(a>=0.75,"A++", a>=0.50,"A+", a>=0.25,"A0", a>=0.10,"A-", TRUE,"A--")
# Hysteresis (requires a_prev and band_prev)
=LET(delta_a, a - a_prev,
cand, IFS(a>=0.75,"A++", a>=0.50,"A+", a>=0.25,"A0", a>=0.10,"A-", TRUE,"A--"),
IF( (cand>band_prev)*(delta_a>=0.05) + (cand<band_prev)*(delta_a<=-0.05),
cand, band_prev))
Power BI / Looker (semantic hints).
# Calculated column BandFromA
BandFromA =
SWITCH(TRUE(),
a >= 0.75, "A++",
a >= 0.50, "A+",
a >= 0.25, "A0",
a >= 0.10, "A-",
"A--"
)
# Measures for weekly counts (illustrative)
App_Count := COUNTROWS( FILTER(Fact, Fact[BandFromA] = "A++") )
Ap_Count := COUNTROWS( FILTER(Fact, Fact[BandFromA] = "A+") )
...
Idempotent ERP/API touchpoint (minimal).
POST /lane
Content-Type: application/json
{
"kpi_id": "AR_collected_issued",
"date": "2025-10-21",
"m": 0.9631,
"a": 0.6123,
"band": "A+",
"knobs_hash": "9d5c...e2",
"build_id": "pilotW1", # optional
"stamp": "SSMCLOCK1|..." # optional
}
# Idempotency key (recommended):
# Idempotency-Key: kpi_id|date|knobs_hash
# Server stores/replaces by (kpi_id,date) if knobs_hash matches
2.14 — CFO ROI worksheet (template starter)
Inputs (example variables).
revenue
daily_rev
days_reduced_in_DSO
refund_rate_baseline
refund_rate_after
gross_sales
MAPE_baseline
MAPE_after
k_cash
Outputs (computed).
Savings_AR := days_reduced_in_DSO * daily_rev
Savings_ref := (refund_rate_baseline - refund_rate_after) * gross_sales
Savings_fc := (MAPE_baseline - MAPE_after) * revenue * k_cash
Savings_total := Savings_AR + Savings_ref + Savings_fc
Notes.
# k_cash translates forecast accuracy into cash impact (org-specific).
# All rates as decimals; units consistent (currency/day where applicable).
# Use rolling baselines so improvements are attributable to lane-driven actions.
2.15 — Evidence pack & governance (archive per pilot)
What to keep (plain list).
Data: stamped CSVs, manifest (ASCII), knobs_hash, CI PASS/WARN/FAIL report
Plots: weekly a-trends with band shading; SDI if used
Notes: 3–5 "caught-early" examples (band slide -> action -> outcome)
Review: quarterly knobs review; update manifest; recompute conformance_checksum
One-screen audit checklist (paste-ready).
[ ] C1 Clamp safety (-1 < a < +1 after clamp)
[ ] C2 Collapse parity (phi((m,a)) = m everywhere)
[ ] C3 Band parity (band_from_a)
[ ] C4–C6 Order/ratio/rollup parity (within epsilon)
[ ] C7 Deterministic export (replay identical)
[ ] C8 Knobs fingerprint (hash flips on knob change)
[ ] C9–C10 Stamps/anchors (if enabled)
[ ] C11 Hysteresis sanity (promote/demote rules)
[ ] Privacy posture (no PII in lanes/stamps)
Navigation
Previous: SSM-Audit – Implementation Guide (2.10–2.12)
Next: SSM-Audit – Deeper Finance Modules (3.1–3.2)
Directory of Pages
SSM-Audit – Table of Contents
Frequently asked questions
SSM-Audit – Q & A