Deal Model vs.
Actual Performance

—Executive Summary

This scorecard compares the platform's actual financial trajectory — taken from the full general-ledger history — against the LBO base case, and recalculates the projected return under the trajectories the ledger will actually support. Because the deal closed recently, the honest form of the year-by-year comparison is: entry baseline vs. model, Year 1 actuals vs. plan, and a return recalculation that carries the unresolved questions from the 100-Day Baseline into the MOIC/IRR arithmetic — where their cost becomes explicit.

What the ledger supports

The growth story is real: revenue $1.90M (FY24, first operating year) → $6.42M (FY25) → $5.68M in seven months of FY26. EBITDA margin expanded from 15.1% (FY25) to 29.1% (FY26 YTD). Capex of $2.01M since FY24 is front-loaded build-out (leasehold improvements $1.24M), consistent with a platform investment phase, and FY26 capex is running well below prior years — the asset base is largely built.

What the ledger does not support

Roughly half the FY26 margin expansion sits in the unreconciled June–July costing artifact flagged in the baseline. The recalculated return is therefore not a number but a range, and the range is brutal: 0.65× MOIC / −8.2% IRR if the pre-June baseline regime is the true business, versus 2.98× / +24.4% if the June–July volume holds at the demonstrated historical margin structure, versus 5.07× / +38.4% only if the as-booked June–July economics are real — which this report does not treat as defensible until the costing reconciliation completes. No sponsor debt exists in the ERP ($10K line of credit only): either the deal was all-equity or acquisition debt sits at a holding company outside NetSuite — the debt-paydown scorecard cannot be computed from this instance either way, and that gap is itself a finding.

1Actual Trajectory — Fiscal Year Scorecard

* FY2024 is the first operating year (first postings Dec 2023; FY2023 activity $5.8K). † Simple 12/7 annualization of YTD — shown for scale only; the regime question (§3) makes any single FY26 annualization unreliable. FY2026 GL contains Aug+ activity and future-dated postings which are excluded via the Jan–Jul monthly series.

2Actual vs. Base Case — Variance Scorecard

Cumulative variance to base case

At month ~8 of Year 1, cumulative variance is dominated by a single unresolved question rather than by operational drift: the two defensible readings of the same ledger differ by ≈$2.6M of annualized revenue and ≈$1.3M of annualized EBITDA — larger than every other variance line combined. Formal cumulative tracking (plan-to-date vs. actual-to-date by quarter) should begin from the August 2026 close, the first under new ownership, with the sponsor's real base case loaded as the NetSuite budget (the baseline found budget coverage already exists for two fiscal years — the mechanism is in place).

3Recalculated Returns — MOIC / IRR Under Actual Trajectories

What moves the return, in order

1. Resolving the regime question (A vs. C): worth 2.33× of MOIC — more than any operational initiative in the 100-day plan. Adjudicated by the August close: does invoice volume hold ≈220/month, and does reconciled gross margin land in or above the 34–43% band?
2. Killing the costing artifact (C vs. D): worth knowing, not worth booking — presenting D externally before reconciliation risks a credibility-destroying restatement later.
3. The $2.89M journal AR/AP resolution: a potential one-time equity-bridge hit at exit if the receivable side writes off; quantified in the baseline, carried here as a return-adjacent risk.
4. Margin structure work (baseline 100-day plan): each sustained point of gross margin ≈$97K of EBITDA at Scenario C volume ≈$0.78M of exit value at 8× — material, but an order of magnitude below items 1–2.

4Thesis Checkpoint — Actions

AAppendix A — Assumptions & Methodology

1. LBO model inputs are illustrative placeholders. Entry multiple (8.0×), entry equity ($16.17M), base-case projections (10% growth, 24% margin, $400K capex, $1.5M amortization), and exit assumptions (entry multiple, 5-year hold, no distributions) are not recorded in the ERP and are shaded throughout. The actual-trajectory figures and the relative scenario spread are instance-derived and survive parameter replacement.
2. Fiscal years by transaction date (TO_CHAR(trandate,'YYYY')), not posting period — the GL contains future-dated postings (baseline finding) and journal-posted history; trandate grouping is the stable basis. FY2026 figures use the Jan–Jul verified monthly series; the 12/7 annualization is labeled indicative.
3. FY2024 is a partial first operating year (operations began Dec 2023; FY2023 = $5.8K). Growth off FY2024 overstates trend; FY25→FY26 is the meaningful comparison.
4. EBITDA definition inherited from the 100-Day Baseline: revenue − COGS − opex excluding 6800-series D&A; other income/(expense) excluded; no interest or tax lines exist in the ledger.
5. Return engine mechanics: MOIC = exit equity ÷ entry equity with exit equity = scenario exit EBITDA × entry multiple; IRR = MOIC^(1/5) − 1. All-equity assumed per the ERP debt evidence; if holdco leverage exists, equity MOIC/IRR sensitivity to it is not computable from this instance (assumption 7). Scenario start-points and their derivations are from the baseline's §2.3 regime analysis.
6. Capex = gross debit additions to FixedAsset accounts, excluding accumulated-depreciation postings. The baseline's FAM reconciliation gap affects net book value, not gross additions.
7. Debt evidence: zero LongTermLiab postings in any year; a single $10K LOC draw in 2026. The holdco-debt hypothesis is stated as such — the ERP cannot confirm or deny debt outside its books.
8. Scenario C margin arithmetic uses the baseline's construction (Jun–Jul volume at ~38% GP midpoint, baseline opex structure); the "$97K per margin point" figure is 1% of Scenario C annualized revenue ($9.7M) at flow-through.
9. Consistency with companion reports: TTM EBITDA ($2,021,516), regime figures, the costing-artifact finding, and the $2.89M journal AR/AP carry over from the 100-Day Baseline unchanged; no re-adjudication.
10. Extraction date: all queries executed August 17, 2026 against the live production instance.

BAppendix B — Data Lineage: Queries

SELECT TO_CHAR(t.trandate,'YYYY') AS yr,
 ROUND(SUM(CASE WHEN a.accttype IN ('Income','OthIncome') THEN -tal.amount ELSE 0 END),0) AS revenue,
 ROUND(SUM(CASE WHEN a.accttype = 'COGS' THEN -tal.amount ELSE 0 END),0) AS cogs,
 ROUND(SUM(CASE WHEN a.accttype = 'Expense' AND a.acctnumber NOT LIKE '68%' THEN -tal.amount ELSE 0 END),0) AS opex_ex_da,
 ROUND(SUM(CASE WHEN a.acctnumber LIKE '68%' THEN -tal.amount ELSE 0 END),0) AS d_and_a,
 ROUND(SUM(CASE WHEN a.accttype = 'OthExpense' THEN -tal.amount ELSE 0 END),0) AS other_exp
FROM transactionaccountingline tal
JOIN transaction t ON tal.transaction = t.id
JOIN account a ON tal.account = a.id
WHERE t.posting = 'T'
  AND a.accttype IN ('Income','OthIncome','COGS','Expense','OthExpense')
GROUP BY TO_CHAR(t.trandate,'YYYY') ORDER BY yr

SELECT TO_CHAR(t.trandate,'YYYY') AS yr, a.acctnumber, a.fullname,
       ROUND(SUM(tal.amount),0) AS net_postings
FROM transactionaccountingline tal
JOIN transaction t ON tal.transaction = t.id
JOIN account a ON tal.account = a.id
WHERE t.posting = 'T' AND a.accttype = 'FixedAsset'
GROUP BY TO_CHAR(t.trandate,'YYYY'), a.acctnumber, a.fullname
ORDER BY yr, a.acctnumber

SELECT TO_CHAR(t.trandate,'YYYY') AS yr, a.acctnumber, a.fullname,
       ROUND(SUM(tal.amount),0) AS net_postings
FROM transactionaccountingline tal
JOIN transaction t ON tal.transaction = t.id
JOIN account a ON tal.account = a.id
WHERE t.posting = 'T' AND (a.accttype = 'LongTermLiab' OR a.acctnumber = '2410')
GROUP BY TO_CHAR(t.trandate,'YYYY'), a.acctnumber, a.fullname
ORDER BY yr, a.acctnumber