Data Room Lies

Every catalog acquisition starts in the same place. A data room opens, and inside sits the familiar furniture: a stack of royalty statements, some contract summaries, perhaps a seller-commissioned earnings report with a decay curve that always seems to flatten out at exactly the right moment (funny that).

And at the centre of it all, the schedule of works- the document that defines what is actually for sale.

Photo by Gabriel

Here is the uncomfortable truth that experienced operators know, and newer entrants often learn expensively: the schedule of works is not a statement of fact. It is a claim. It is an export from a seller's system or 3rd party, reflecting years of assumptions, part-completed migrations, inherited registrations and un-actioned paperwork. It describes what the seller believes they own, or in some cases simply what their system happens to say this month.

I’ve often seen that in deal after deal that schedule is treated as ground truth- the model is built on it, the multiple is applied to it and the purchase price is wired against it.

But… the data room lies. Not maliciously- but often structurally. Buyers who don't test those claims before pricing a deal are choosing to discover the truth after completion, when claims become significantly more fraught to resolve.

A Claim is Not a Fact

Why is the schedule so unreliable? Because in music publishing, ownership and the record of ownership are two different things- with revenue attempting to follow the record.

A comprehensive schedule would show titles, registrations, identifiers, copyright numbers (critical for ownership), writers and shares. What it rarely shows is whether those shares are actually registered that way at the societies doing the collecting. Whether the ISWCs listed are preferred or archived. Whether the recordings driving the income are properly linked to the works at all. Whether the splits across all claimed parties actually sum to 100% (you would be amazed how often they don't. Though, if you've worked in publishing for any length of time, you wouldn't).

As we explored in The Metadata Iceberg, the visible layer of catalog data sits on top of much deeper questions about works, contributors and registrations. A data room presents only the visible layer- and only often 1 perspective on it at that. If the buyer is lucky it may have been polished and flattened into a spreadsheet, but it almost certainly will have been stripped of every caveat the seller's own catalog team would happily volunteer over a coffee.

The royalty statements are the most honest documents in the room, because money doesn't lie about where it came from. But even they only tell you what was collected- never what wasn't. The leakage, by definition, doesn't appear on the statement.

Lies, Damned Lies, and Schedules of Works

So what should a buyer actually test before pricing a deal? In my experience, five reconciliations do most of the heavy lifting.

1. The schedule vs. the statements

Do the works on the schedule actually appear in the income? A meaningful proportion of most schedules earns nothing at all- fine, that's the long tail. More interesting is income on the statements attributable to works not on the schedule (what exactly is being sold?), or schedule works whose earnings pattern doesn't match the claimed share. If the seller claims 50% of a work but the statement income implies 25%, someone's data is wrong- and you are about to pay a multiple on the difference.

2. The schedule vs. the registrations

Take the top earners- the 10 or 20 works that drive the vast majority of income- and check them independently at the societies, the MLC, and key licensing hubs. Are they registered? At the claimed shares? Free of duplicate, conflict and counterclaim? A work in dispute at a major society is not the same asset as a work with clean, uncontested registrations, and it should not be priced as one.

3. The identifiers

Are ISWCs and IPIs present, valid and consistent? Remember that identifiers are a currency, and their value depends entirely on confidence. A schedule with patchy or conflicting identifiers isn't just an admin nuisance, but a leading indicator of how much post-completion archaeology awaits, and how much income is currently flowing through the wrong pipe.

4. The recordings

Publishing income lives and dies on recording-work linking in the digital era. Are the ISRCs that actually generate the streaming income matched to the works being sold- at the hubs and societies where matching determines payment? A catalog can have immaculate work registrations and still leak heavily at the recording layer.

5. The dates and the documents.

Copyright dates, grant dates, reversion clauses, termination windows. Rights that look like long-term ownership can still have a scheduled departure time- and the notice may already be in the post. If the data room can't produce the chain of title behind a top earner, that absence is the finding.

None of this is exotic. It is basic verification- checking claims against the systems that actually pay. The remarkable thing is how rarely it is done systematically before a price is agreed- more typically being discovered in fragments during the years afterwards.

Discrepancies are Pricing Information

Here is the reframe that matters for acquirers: the point of testing the data room is not to kill deals, but to price them properly.

Every discrepancy found in diligence is information. Sometimes it's downside- overstated shares, disputed works, terminable grants, income that will not survive the transfer. That should flow directly into price, structure, warranties and holdbacks. A seller warranty is worth considerably more when you can attach it to a specific, evidenced list of exceptions rather than a general hope.

But just as often it's upside. Unregistered works. Unlinked recordings. Unclaimed income sitting in matching pools with the buyer's future name on it. A catalog that has been under-administered is, for the right operator, a catalog being sold below its recoverable value. It is Operational Alpha. The buyers who can quantify that gap during diligence can bid with confidence others can't- they know precisely what the asset earns today, and what it could earn once the data is fixed.

That is the real asymmetry in this market. Two bidders look at the same data room- one sees a spreadsheet and applies a multiple. The other sees a set of testable claims, validates them, and prices both the risk and the recovery. Same asset- materially different information positions.

Trust, but Verify (... then Verify Again)

The catalog market has become impressively sophisticated about the financial layer- discount rates, decay curves, vintage analysis, structured payouts. That sophistication now needs to reach a layer down, to the data the whole model stands on. A DCF built on an unverified schedule of works is precision engineering on unexamined foundations.

The good news is that this verification is more achievable than it has ever been. External registration data is more accessible, matching at scale is more economical, and the tooling exists to reconcile schedules against the outside world in days rather than weeks. What it still requires is the judgement to know which discrepancies matter, and the discipline to make verification part of the deal workflow rather than a post-completion clean-up exercise (by which point the price has been paid- and every finding is a cost rather than a negotiating position).

For anyone deploying serious capital into music rights, the question to ask of any acquisition is simple: has anyone independently confirmed that the assets in the model exist, with the correct shares claimed, in the systems that pay? If the answer is no, the model isn't a valuation, but a hope with a spreadsheet attached.

The data room will keep lying- the buyers who prosper will be the ones who check.