in royalties audited with discrepancies
And it's still climbing — we find more every time a statement comes in.
Where music money leaks
Nothing on a royalty statement is technically false. The damage lives in what happens between gross income and your royalty base — and in what never appears at all. Rate miscalculations are the single largest source of underpayment we find. Wrongful deductions are the second.
LEAK / 01 masters
Your contract says 28% on streaming. A modifier buried in the rate preamble — "one-third of the otherwise applicable rate" — quietly makes it 9.33%. Most artists never multiply it out.
Test: effective rate vs. headline rate
LEAK / 02 masters
Fees, costs, and "distribution charges" subtracted from gross before your royalty is calculated — none of them on the contract's permitted-deductions list, some paid to entities the label itself owns.
Test: is the deduction list exhaustive?
LEAK / 03 masters
Reserves exist because CDs get returned. Streams can't be returned — yet labels keep withholding "breakage" reserves against streaming income under clauses written for shipping pallets.
Test: does Art. 7 restrict reserves to physical?
LEAK / 04 masters
Negative line items appear against your streams — adjustments, true-ups, clawbacks — with no contractual mechanism authorizing a single one of them.
Test: any basis for negative entries?
LEAK / 05 publishing & rights
Sync placements you've seen on TV. PRO performance income. Neighboring rights collected by societies in 40 territories. If an income type never appears on the statement, it can't be underpaid — only unpaid.
Test: statement vs. complete income map
LEAK / 06 recoupment
The advance from your first record deal is still being recouped — against your publishing, your second album, your live recording. One debt, silently attached to everything you sign.
Test: which agreements may offset which?
One dollar, followed
One hundred cells. One dollar of gross streaming receipts arriving at your label, at current US rates of roughly a third of a cent per stream. The label keeps $0.65. Of the $0.35 contractually due to the artist, watch a shifting 10–30% — the range statement reviews across the industry consistently uncover — never arrive.
Withheld from the artist's end: 10–30% — the range industry reviews consistently find
The red cells are the claim. Music royalty audits consistently find 10–30% of artist royalties unpaid — this catalog sat near the top of the range: $48,212.66 across 18 months and 137.7 million streams, plus documented discrepancies.
The pipeline
Every finding survives the same discipline: if the contract is ambiguous, we flag it as ambiguous. If the evidence is thin, it never enters the claim. A wrong finding in a document is worse than no finding.
PASS 01 — PARSE
Record deal, publishing agreement, distribution terms — rates, permitted deductions, reserve restrictions, audit rights, dispute windows, extracted with verbatim citations. Absent fields marked null, never guessed.
→ Structured contract record, red flags surfaced
PASS 02 — ANALYZE
We rebuild the math on each statement row — every track, every source, every period: deduction ratios, rate deltas, negative entries, reserve behavior, with mean and deviation across the full catalog.
→ Statistical report, flagged rows by number
PASS 03 — CROSS-EXAMINE
Each variance meets the clause that governs it. Then it meets the strongest argument the label's counsel could make. Findings that survive are confirmed; the rest are downgraded.
→ Confirmed / uncertain / rejected, on the record
PASS 04 — ENFORCE
A document built to be read by the label's senior counsel tomorrow morning: row-level citations, quoted clauses, statutory interest, and a cure demand with a deadline.
→ Notice of Underpayment, 30-day clock started
Statistical proof over assertion
An accidental error is noisy. A programmed deduction is not. When every one of 1,142 statement rows — across three accounting periods and an entire catalog — loses 9.98% between gross income and the royalty base with a standard deviation under a thousandth of a percent, the pattern isn't an opinion. It's arithmetic, and it's repeatable by anyone the label hires to check it.
standard deviation across 1,142 flagged rows
Systematic deduction confirmedOutput, matched to evidence
We never overstate what the data supports. The output type follows the strength of the evidence — and each one moves your position forward.
The full claim: every finding with its clause quoted, its rows cited, its adversarial test documented. Interest documented, correction requested, 30-day response window.
Findings confirmed with full statistical and contractual basis.
When the statement is a wall: PDF-only, no gross income disclosed. We demand the gross figures, the itemized deductions, and the identity of every affiliated entity taking a fee — on a 30-day deadline.
Statement format is opaque; no quantified finding yet possible.
A quieter instrument. Built on the contract's own "documented costs" language, it asks for the substantiation the agreement already requires — routine in tone, hard to refuse in substance.
Statement exists but discloses too little to verify accuracy.
Who we work for
Pricing · per statement
$29 / statement
One-off. Full findings, citations, formal notice.
$39 / mo
2 statements monthly. Best value.
* Subscriber overage: $20 per additional statement past your monthly allowance.
Our earliest audits ran on our founder's own royalty statements — a music royalty audit of his former label — findings submitted, matter ongoing. Inverse exists because we needed it first.
Data security
Your statements are privileged. We treat them that way.
SOC 2 Type II compliant infrastructure, TLS in transit and AES at rest, row-level security per account, never used to train third-party models, GDPR & CCPA with DPA on request. Security & Trust →
Cancel anytime
One statement — a label, publisher, PRO, or neighboring rights accounting. You'll know exactly what you're owed — to the row, with interest — or we'll say so plainly.
Send us one statement →Or try a free audit →Privileged & confidential · Findings delivered within 24 hours