Music Ledger HQ, an Instagram page that models what songs earn and who keeps the money, published an estimate on 26 October for one of the biggest records of the last three years. By its reckoning, "Espresso" by Sabrina Carpenter has generated about $42.55 million across every platform it earns on. The same model has the artist walking away with roughly $6.07 million of that, 14.3 per cent before tax. The rest went to the label, the producer, the publisher and the manager. Every number in that paragraph is an estimate built from public stream counts and standard deal terms, and Music Ledger says so itself. Nobody outside the contracts has seen the real statements. It is still the clearest worked example of a modern hit's economics that most working musicians will ever get to read, and the shape of it matters far more than the decimal places.

The estimate, and why it is worth reading carefully

Here is the ledger as Music Ledger lays it out. "Espresso" has passed 3.23 billion streams on Spotify alone. Counting every platform, and counting both the recording and the song underneath it, the model arrives at $42.55 million generated. That pot splits into two lines: an estimated $34.92 million of streaming income, which belongs to the recording, and about $7.63 million of publishing income, which belongs to the songwriting. From there the money is carved up. The label keeps $27.94 million. Producer points earn $1.40 million. Publishing and administration take $6.08 million. Management earns $1.07 million. What is left for the artist comes to roughly $6.07 million.

Music Ledger grades deals on a six tier retention scale, judged by how much of the money an artist keeps. Tier 1 is 95 per cent or more. Tier 6 is under 10 per cent. "Espresso" lands at Tier 5, which the page defines as keeping between 10 and 17 per cent. The full post is at instagram.com/musicledgerhq, and the figures above are theirs, not ours.

Two honest caveats before anything else. First, the model assumes standard terms, and an artist of Sabrina Carpenter's standing may have negotiated better than standard, so her real share could be higher. Second, the ledger covers one song's recorded and published income only. It says nothing about touring, merchandise, brand deals or the songs she wrote for other people, which is where a great deal of a pop career's money actually lives. Neither caveat changes the lesson. A record deal signed on ordinary terms sends most of the money somewhere other than the person whose name is on the sleeve.

The recording and the song are two different businesses

The single most useful thing in the Music Ledger post is the split between $34.92 million and $7.63 million. Those two lines are not two ways of describing the same income. They are two separate copyrights, owned by different people, licensed under different rules and paid through different pipes.

The recording, sometimes called the master, is the specific performance captured in the studio. In a conventional major label deal the label owns it, pays for it, and keeps the lion's share of what it earns. The song is the composition: the melody, the chords and the lyric. It belongs to the writers and is usually administered by a publisher, who collects performance and mechanical royalties from every version of the song anywhere in the world, including cover versions and sync placements the label never touches.

"Espresso" was written by four people, with Carpenter among them, and that writing credit is why the artist line on the ledger is not smaller still. Music Ledger's $6.07 million includes her writer's share of the publishing income alongside her artist royalty on the recording. An artist who does not write, singing a song the label sourced from outside writers, has no claim on that $7.63 million at all. The oldest advice in the business is still the best advice in the business: write, or co write, the songs you sing.

Why the label's share is the biggest line on the ledger

$27.94 million to the label reads like a scandal on a phone screen, and the comments under the post treat it as one. It is worth being more precise than that. The label's share pays for the thing most artists cannot fund themselves: the recording budget, the radio promotion, the playlist pitching, the video, the photographers, the press team and the global distribution that put the song in front of enough people to generate 3.23 billion plays. The label also carried the risk. For every "Espresso" there are many singles the same label funded that returned nothing, and the hits pay for the misses.

That is the trade. An artist royalty of somewhere between 15 and 25 per cent of the recording's income, after the label has recouped its advance and its agreed costs, in exchange for scale no independent could match. Whether the trade is fair is the argument the UK industry has been having in public for years. When the Musicians' Union and the Ivors Academy surveyed professional musicians in December 2020, 82 per cent reported earning less than £200 a year from streaming. Naomi Pohl, then the Musicians' Union's Deputy General Secretary, said at the time: "These statistics show that music streaming does not play its part in supporting the careers of the vast majority of creators." A Commons select committee went on to call for a complete reset of streaming economics the following summer. The ledger for one of the most streamed songs in the world shows the same structure the survey was describing at the bottom of the market. The proportions do not change much with scale. Only the zeros do.

We looked at the other end of this same question earlier this week, in the George Michael versus Sony case, where an artist at the height of his powers tried to argue his way out of exactly this kind of deal and lost.

Producer points, management and the people paid before the artist

Two smaller lines on the ledger deserve a closer look, because they are the ones a working musician is most likely to meet in their own career.

Producer points are a percentage of the recording's income paid to the producer, typically three or four points of the retail or wholesale price on a major release. Music Ledger puts the producer's take on "Espresso" at $1.40 million. Points are negotiated per record and they come out of the artist's side far more often than out of the label's, which is why an artist can find their 18 per cent quietly becoming 15 before a penny arrives. On a function band's own single, the producer may well be a band member or a friend with a studio, and the conversation about points never happens. It should. A written agreement about who owns what and who is paid what, signed before the session, is worth more than any amount of goodwill afterwards.

Management is the other line, at $1.07 million. A manager's commission runs at 15 to 20 per cent of the artist's income, not of the gross, which is why the management figure on the ledger is a fraction of the label's. The point to notice is where the commission applies. A well drafted management contract commissions the income the manager helped create during the term. A badly drafted one commissions everything the artist earns from any source, including income the artist generated alone, and keeps doing so for years after the relationship ends. The ledger shows management as a modest cost against a large income. Read the wrong agreement and it becomes a large cost against a modest one.

What a function band's ledger looks like next to this one

Most readers of this blog will never have a song on 3.23 billion streams, and the comparison still matters, because the structure is the same whatever the scale. Take a wedding band quoting a client for a Saturday in June. The fee comes in, the agency or management commission comes off, and the rest is the band's. There is no label in the chain, no producer points, no publisher. The recording and the song that the audience hears belong to somebody else, licensed to the venue through PRS, and the band's income is the performance itself. On a typical night a function band keeps the large majority of what the client pays. On Music Ledger's reckoning, Sabrina Carpenter kept 14.3 per cent of what "Espresso" made.

That is not an argument that a wedding band has the better deal. The pop star's 14.3 per cent is 14.3 per cent of a number with seven zeros, earned while she sleeps, from a recording made once. The band's share is earned again every Saturday, in person, with a van to load at 1am. It is an argument about ownership. The band owns its product and sells it directly. The artist on a conventional deal sold a large piece of hers before the first stream was counted, in exchange for the machine that counted them.

The lesson for anyone in between, which is most working musicians at some point, is to know which business you are in before you sign. Releasing your own recordings keeps the master and its $34.92 million line, at the cost of funding and promoting them yourself. Writing your own material keeps a share of the $7.63 million line whatever happens to the master. Booking your own shows, or working with a management partner who charges for a service rather than taking a piece of everything, keeps the live income whole. Each of those is a line on a ledger you control.

Music Ledger's post ends by asking which artist's ledger to run next. The more useful exercise is to run your own. Write down every pound you earned from music last year, then write next to each line who else took a share of it before it reached you, and on what terms. Most musicians have never done it. The ones who have tend to sign very different contracts afterwards.