StratApex
Operating Intelligence Suite
An Interchange Global Advisors tool · Make Your Connection
Lens B · Operating Intelligence
Quality of Controlling
Edits saved locally
Hero banner — wide image
Click · drop a file · paste a screenshot. Renders edge-to-edge above the hero.
add image
Two Questions, One Tool.
How close are the books to true accrual? Is the controller the cause of any gap? Score detail and the contextual rules below.
Quality of Controlling, not Quality of Earnings. Fourteen weighted dimensions. Seven named kill tests. A fairness layer for what sits outside the controller's reach. The default verdict is to retain the controller; only an explicit threshold crossing changes that. Operating leverage on a hiring decision, not an audit. Drop your files, run the analyzer, fill the questionnaire, see the score, then read the kill test.
Load the Sample Run first. It populates the dashboard with a worked example so you can walk every tab before deciding whether this fits your situation.
Northwind Manufacturing — a $42M mid-Atlantic manufacturer, 18 months on QBO. Synthetic data, illustrative only.
Composite & Posture
Weighted average across 14 dimensions, adjusted for the factors captured on the Context tab. Posture is a description, not a grade.
v4 · coming with commercial tier
Workbook will carry: Summary · Dimensions · Findings · Kill Tests · Context · Audit (session-keyed).
6.4/10
Hybrid
Between modified cash and modified accrual.
7/7 pass
H₀ holds. Default verdict: retain.
0 flagged
No external or organizational headwinds documented.
Dimension Snapshot
Revenue recognition timing (ASC 606)
12%
7.0
Expense matching
10%
6.0
Cut-off testing modifier-eligible
10%
5.0
Reversing JE hygiene modifier-eligible
8%
5.5
Prepaid amortization modifier-eligible
7%
6.5
Accrued liabilities completeness modifier-eligible
8%
6.0
Deferred revenue release
6%
7.5
Depreciation & amortization
5%
8.0
AR allowance / reserve
5%
5.5
BS reconciliations modifier-eligible
8%
5.0
Intercompany cut-off & elimination
5%
7.0
Period-close discipline modifier-eligible
8%
4.5
JE documentation modifier-eligible
8%
5.5
Approvals & SoD weight-fixed
0% (Q-only)
6.0
File Intake
Trial balance, full-year GL detail, AR and AP aging at month-end, payroll register, prepaid schedule, accrual schedule, fixed-asset register. Drop them in any order.
Browse files or drag them here
.xlsx · .xls · .csv up to 50MB each. Files stay on the local machine; nothing is uploaded externally.
Footnote: the analyzer evaluates whatever data is actually provided. Dimensions that cannot be scored from the supplied files are reported as N/A and their weight redistributes across the dimensions that ran. Sufficient data is required for a confident score and recommendation; the engagement note will say so when the denominator is thin.
Score detail — 14 dimensions of accrual-basis bookkeeping quality
These are the fourteen. Each row scores one aspect of accrual-basis quality on a 1-to-10 scale, carries the evidence behind the score, shows the default weight in the composite, and flags whether Context-tab factors can move that weight. Sample scores below reflect a Northwind-shape mid-Atlantic manufacturer; your scores populate after you upload or load the sample run.
Each row carries the 1-to-10 raw score, the file or questionnaire field that produced it, the GAAP / tax / cash lens framing on the Methodology tab, and a pill that says whether Context-tab factors can move its weight. modifier = external or organizational conditions can reduce the weight. fixed = anchored, doesn't bend to context. default = no special rule; sits at its baseline weight regardless of Context.
Admin only: live-app drawer screenshot
For your documentation. Drop a screenshot of what the evidence drawer looks like in the live app.
Questionnaire — 10 sections
Principal and controller answer independently. Disagreement is signal, not noise. Optional third pass for COO or CEO.
Section 11 — Context & Fairness
External and organizational factors that sit outside the controller's direct control. Captured before any kill test is read.
11.A — External factors
Litigation or regulatory action consuming finance bandwidth
Active dispute, examination, or subpoena cycle affecting close staff time.
Documented since
M&A activity, divestiture, or carve-out in the scoring window
Deal close, integration, or separation activity in the last 12 months.
Documented since
ERP / GL migration in the last 12 months
Platform change affecting chart of accounts, posting logic, or report shapes.
Documented since
Bank or payroll-provider change in the scoring window
Account migration, file-format change, or reconciliation gap.
Documented since
Material customer or vendor bankruptcy / dispute
Bad-debt write-down, vendor non-performance, or contract dispute affecting balances.
Documented since
Macro event affecting close cadence
Catastrophe, fraud incident, cyber event, or other one-off operational shock.
Documented since
External audit or tax exam in progress
Active third-party review pulling controller time away from close.
Documented since
11.B — Organizational factors
CFO or COO directives overriding standard treatment
Written or de-facto instruction to deviate from standard accounting policy or approval flow.
Documented since
Resource constraints — headcount, tooling, budget
No AP / AR clerk, no staff accountant, no requested tool funding.
Documented since
Reporting-line ambiguity or recent restructuring
Dotted lines, role changes, or unclear escalation paths in the last 12 months.
Documented since
Controller tenure under six months
Learning curve window. Coachability (K7) measurement still applies; weight on close discipline softens.
Documented since
Material scope expansion without role re-leveling
New entities, geographies, or chart-of-accounts complexity added without compensation, title, or staffing change.
Documented since
Software stack imposed by upstream
No authority to evaluate, choose, or replace ledger or close tools.
Documented since
Prior consultant or controller left undocumented state
Inherited config, undocumented JE patterns, or institutional knowledge gap.
Documented since
Board or investor reporting demands above standard cadence
Weekly KPI packs, ad-hoc investor packages, or off-cycle requests crowding the close.
Documented since
Attribution chips on every finding
In the Findings tab, every line item carries a chip: Controller · Upstream · External · Shared · Unattributed. Default is Unattributed; analyst sets it; controller can dispute in writing. The dispute log persists with the score.
Findings
Each finding carries evidence, an attribution chip, the affected dimension, and a remediation owner.
Working-day-7 slip in three of six months
Close package delivered WD 9, 8, 10, 7, 9, 8 vs WD 5 target. No remediation plan filed.
Revenue JE clustering on day 1 of following period
38% of revenue JEs post in the first two days of the next period. Service-date evidence missing on 22%.
Wire dual-control bypassed once in Q1
CFO authorized single-approval release of $84k vendor payment. Documented in email; controller objected in writing.
Memo field substantive-content rate at 41%
Target is 80%. Adjusting JEs over $5k missing supporting memos in 17 of 41 sampled rows.
Eisenhower Action Grid — what to do about it
Findings sorted by urgency and importance. Click any quadrant in the live app to open the evidence drawer.
Urgent
Not Urgent
Important
DO · this week
- Cut-off retraining
- Wire dual-control reinstated
- JE posting lockdown for non-controller users
- Vendor-add review (single-approver pattern)
SCHEDULE · 30 / 60 / 90
- Close-checklist rebuild
- Reversing-JE pack
- Prepaid automation
- Source-of-record canonical map
Not Important
DELEGATE
- 1099 cleanup → tax preparer
- Credit-card reconciliations → AP clerk
- Vendor W-9 chase → AP automation
DROP
- Manual COA re-map sheet
- Duplicate weekly cash report
Verdict + Kill test
H₀: retain the operator. The verdict is graduated, not binary — composite score plus kill-test outcomes map to one of five bands. No K-test trip is valid if an unaddressed external or organizational factor on the Context tab is the more-plausible cause.
Verdict scale
- Retain — composite ≥ 8.5, no kill test tripped. Top quartile.
- Retain + minor coaching — composite 7.0–8.5. Solid posture with two or three correctable gaps. Coach in place; re-score in 60 days.
- Soft Keep + Coach OR Replace — composite 5.0–7.0. Real gaps. Define a 60-day coaching plan with measurable cures, or replace if the gap exceeds coaching budget or timeline.
- Replace recommended — composite 3.5–5.0. Posture failing across multiple dimensions. Replace unless the Context tab carries a documented upstream cause that explains it.
- Terminate — composite < 3.5, or any K1–K6 fail. Hard override regardless of composite.
K7 (Coachability) fail inside the Soft-Keep band escalates to Replace. K7 measures whether the operator cured 30-day feedback items at the 60-day mark.
K1
Integrity
PASS
Threshold: documented misstatement directed or executed by the controller, in writing, ≥ 1. Weight class: fixed — no Context-tab modifier applies.
K2
Concealment
PASS
Threshold: known control failure not disclosed for > 30 days, surfaced by another party.
K3
Materially inaccurate reporting
PASS
Threshold: restatement > 5% on revenue, EBITDA, or cash in 3 of 3 months, attributable to the controller rather than to an upstream directive. Attribution gate consults the Context tab and Findings attribution chips.
K4
Cash / bank variance
PASS
Threshold: unreconciled cash variance > 1% of revenue for 60+ days with no remediation plan.
K5
Close-cycle drift
WATCH
Threshold: 4+ consecutive months WD slip ≥ 5 with no remediation owned by the controller. Current: 3 of 6 months over WD 7; no remediation plan filed. One more month of slip without a plan trips this.
K6
Refusal to document or sign
PASS
Threshold: documented refusal to sign a rep letter or document a judgment. Weight class: fixed.
K7
Coachability
PASS
Threshold: < 30% of written 30-day feedback items cured by the 60-day mark. Measured at the re-score date, not at first run.
Web research
Curated sources per dimension: FASB, AICPA, COSO, IRS publications, CFA Institute.
Encoder posture: any free-text the user pastes into the search box is scanned for entity names matching the loaded data set; matches are replaced with synthetic tokens before the query goes outbound. Replacement table stays local. The user sees their original text; the network sees only the encoded query.
Methodology
Plain-English, GAAP lens, tax lens, cash lens, and the source field that produced each score. Written so a CPA, attorney, or controller can audit the read.
Weight-modifier rules — fairness layer
Context-tab factors move the weight of modifier-eligible dimensions and leave weight-fixed dimensions alone. The raw 1-10 score for every dimension always shows. The composite reflects the adjusted weights.
Modifier-eligible 7 dimensions
- Period-close discipline
- BS reconciliations
- JE documentation
- Accrued liabilities completeness
- Prepaid amortization
- Cut-off testing
- Reversing JE hygiene
Weight-fixed 3 anchors
- Approvals & segregation of duties
- Integrity (K1)
- Refusal to document or sign (K6)
1
Revenue recognition timing (ASC 606)
Source: GL revenue distribution + deferred-revenue presence + contract review. GAAP: five-step model. Tax: §451(b)/(c) book-tax conformity. Cash: not applicable.
2
Expense matching
Source: GL expense distribution vs invoice or service dates. GAAP: matching principle. Tax: §461 all-events test. Cash: ignored.
3
Cut-off testing
Source: day-of-month clustering of revenue and expense JEs around period boundaries. GAAP: period assignment. Tax: economic performance. Cash: not applicable.
5
Prepaid amortization
Source: prepaid schedule vs expected straight-line curve. GAAP: systematic and rational allocation. Tax: 12-month rule. Cash: expensed at pay.
12
Period-close discipline
Source: GL effective-date to posting-date lag, average working-day lag per period. GAAP: timely reporting. Tax: not applicable. Cash: not applicable.
How To
Run order, what each tab does, and the rules the tool enforces. Read this before your first scoring pass.
Recommended run order
- Pull files into one folder. TB, full-year GL detail (with posting_date if your GL exposes it), AR/AP aging at month-end, payroll register, prepaid schedule, accrual schedule, fixed-asset register.
- Open File Intake. Browse-and-add every file. Hit Run Analyzer.
- Fill the Context tab. External and organizational factors come first. Each one reduces the weight of dimensions it can plausibly excuse, and gates any kill-test trip whose more-likely cause is upstream.
- Fill the Questionnaire. Principal first, save draft, send the link to the controller (or print and have them fill independently). Disagreement is signal.
- Re-run scoring. Read Findings, work the Eisenhower 2×2, open the Kill Test.
- Document a 30/60/90. Pull actions from the Do and Schedule quadrants. Re-score at 60 days. Coachability (K7) is measured here.
What each tab does
- Dashboard. Composite, posture, kill-test verdict at a glance. Click any KPI to drill in.
- File Intake. Browse + add files. Click Run Analyzer to score.
- Score Detail. All 14 dimensions of accrual-basis quality with their raw 1-to-10 scores and evidence. The pill on each row marks whether Context-tab factors can move that dimension's weight.
- Questionnaire. Principal and controller answer independently. Optional third pass for COO/CEO.
- Context · Fairness. External and organizational factors. Activated factors reduce dimension weights and gate kill-test trips.
- Findings. Each finding carries evidence, an attribution chip (Controller · Upstream · External · Shared · Unattributed), and a remediation owner.
- Eisenhower 2×2. Do / Schedule / Delegate / Drop quadrants sourced from Findings.
- Kill Test. H₀: retain the operator. Default verdict regardless of composite. Only an explicit threshold crossing changes that.
- Web Research. Curated FASB / AICPA / COSO / IRS / CFA Institute lookups. Demo and Internal tiers route every query through a name-encoder.
- Methodology. Per-dimension plain-English, GAAP lens, tax lens, cash lens, source field. Designed so a CPA, attorney, or controller can audit the read.
Rules the tool enforces
- Composite is posture, not a grade. Low composite plus upstream-attributed findings equals coaching plan, not termination.
- The kill test is independent of the composite. A 4.0 composite with all seven kill tests passing means retain the operator and fix the structure.
- Context-tab factors don't hide raw scores. They move weights, never numbers. The Methodology tab documents every modification.
- No K-test trip is valid if an unaddressed external or organizational factor is the more-plausible cause. Predicate is (threshold crossed) AND (no upstream cause documented in last 90 days).
- No data leaves your machine in the Demo or Internal tier. Flask binds to 127.0.0.1 locally. Web research routes through a name-encoder so vendor, customer, and employee names never leave in clear text.
- This is Quality of Controlling, not Quality of Earnings. Operating leverage on a hiring decision, not an audit.
Editing this artifact
- Every visible text element is editable. Click any heading, label, paragraph, or list item and start typing. Changes save automatically to your browser's local storage about half a second after you stop.
- Plain-text paste: Cmd+Shift+V (Mac) or Ctrl+Shift+V (Windows) when pasting from a styled source.
- Image blocks accept files three ways: click to browse, drag-and-drop, or paste from clipboard. Replace swaps. Remove clears.
- Reset wipes every edit and every image back to defaults. It asks for confirmation first.
- Edits are scoped to this device. If you share the .html file, the recipient starts fresh. Your edits don't travel with the file.
Deploy profiles
- Demo. Synthetic Northwind data, no upload accepted, no web egress. Public; no auth.
- Internal. Real uploads, name-encoder on all web queries, ephemeral session storage. Gated by Cloudflare Access — Brandon allowlist only at first.
- Commercial (v4). Magic-link auth, charge wall, persisted score history, 60-day re-score for K7. Different surface, different domain.
- Profile is set by the
QOC_PROFILEenvironment variable. The same codebase ships every tier.
Take it further with Interchange Global Advisors.
For follow-on advisory, licensing, partnership, or reseller conversations on the Quality of Controlling Evaluator and the broader StratApex Suite.
Get in touch