Distribution: Getting Your Music on Spotify, Apple Music, and Every Platform

The Button That Isn't There

Jaylen Cole has a folder on his desktop called GLASS HOURS — MASTERS, and tonight, for the first time in fourteen months, there is nothing left to do to the music inside it.

Five tracks, printed through the chain he built in Chapter 32. Loudness measured and settled in Chapter 33 — integrated LUFS logged, true peaks parked under -1.0 dBTP, the slammed versions retired after they lost the blind test. Translation verified past stereo in Chapter 34: the binaural check, the mono fold-down, the phone, the car. The car. "Static Bloom" — the single, the track this book has been building since its session skeleton in Chapter 6 — plays in his cousin's car now and the 808 blooms, fourteen months after the night it disappeared and started this whole education.

The music is finished. So Jaylen does what every finished musician on earth does next.

He googles "how to put a song on Spotify."

He's expecting a button. He has spent his whole creative life on platforms with a button — the beat sites, the video sites, the place he uploaded loops that got 40 plays. Somewhere on Spotify, he assumes, there's an Upload tab, a drag-and-drop box, a progress bar, done by midnight.

There is no button.

What there is instead is an industry. The search results are a wall of companies he's never heard of, every one of them offering to put his music on Spotify, Apple Music, and "150+ stores worldwide," every one of them with a different price: some want a flat fee every year, some want a cut of his royalties forever, some want nothing up front and sell him add-ons at checkout. Half the articles ranking them are written by the companies themselves. It's 11 p.m. and Jaylen — who can tell you the attack time on his drum-bus compressor from memory — realizes he knows absolutely nothing about how music physically arrives at the place where people hear it.

Here's the part that almost went wrong.

Jaylen picks a distributor (it doesn't matter which — by the end of this chapter you'll know how to pick yours), creates an account, and starts the upload form. The form asks for a release date. The master is done. Waiting feels unbearable, the way it always does. He picks the soonest Friday the calendar will let him: eight days out. His thumb is one tap from confirming when he does the thing that has saved him a dozen times in this book — he stops and reads the fine print first.

The fine print is a quiet avalanche. The distributor "recommends" three to four weeks of lead time. The pitch portal inside the artist dashboard — the one that submits your unreleased track to the platform's own playlist editors — wants the track delivered and pitched before release day, with a stated minimum around a week and an unstated reality that editors work much further ahead than that. The pre-save campaign every artist he follows ran for their last single? Can't exist without a gap between upload and release. The new-release feed that notifies your followers needs the release sitting in the system early enough to be seeded. And the distributor's own review — actual humans and systems checking his files and metadata — takes days by itself.

At 2 a.m., phone notes open, Jaylen writes the sentence this chapter is built on:

"The master is finished. The release isn't."

He moves the date four weeks and a day. Then he starts learning the plumbing — the pipes between his desktop folder and every play button on earth. This chapter is that plumbing. I'm going to make you love plumbing, because plumbing is the difference between your record existing and your record being available — and because every hour you spend understanding it is an hour your music doesn't lose to a preventable mistake.

🏃 Fast Track: Comfortable with what a distributor is and itching to schedule? Read "The Gate" and the pricing-model tradeoff table, then go straight to "The Metadata That Matters" — read every word of it, it's where releases break — then the four-week timeline diagram and the release-day checklist. Do exercise C1 (the metadata sheet) and C4 (the calendar) tonight. Everything else can be reference material on upload day.

🔬 Deep Dive: The business mechanics here connect outward in every direction: the money side of this chapter is finished in Chapter 36 (the composition copyright and the royalty taxonomy), the audience side in Chapter 37, and the full industry context lives in the Further Reading — Ari Herstand's book for the artist's-eye view, the annual industry reports for the data. One honesty note before you dive: everything in this chapter that describes platform policy is true as of this writing and guaranteed to drift. The structures are durable; the details are weather. Appendix I keeps the directory of current names so this chapter doesn't have to.

The Invisible Supply Chain Behind Every Play Button

You've used the result of this chapter thousands of times without seeing it.

You tap play on a track. Within half a second, sound. It feels like the most frictionless act in music — closer to turning on a faucet than buying a record. But trace what had to be true for that half-second to work, and a supply chain unfolds that's every bit as physical as the one that used to move vinyl on trucks.

Somebody finished a master — a WAV file on a desk, exactly like yours. That file, plus a payload of facts about it (who made it, what it's called, who wrote it, who owns it, where it's allowed to be sold), traveled to a distributor — a company whose entire job is moving music and metadata into stores. The distributor validated it, formatted it to each store's delivery spec, and pushed it to dozens of platforms in dozens of countries. Each platform transcoded the audio into its own streaming formats — the lossy codecs from Chapter 2, the reason your -1 dBTP ceiling from Chapter 32 exists — indexed the metadata, attached the artwork, and wired the whole package to a play button.

Then the chain ran backward. Your tap got logged. At month's end it was counted, aggregated with a few billion of its closest friends, converted into money by the platform's accounting, reported to the distributor, matched — by an identifying code stamped into the metadata — to the exact recording you played, and paid through to whoever owns it. Months after your half-second of friction-free listening, a fraction of a cent completed the round trip.

The record store didn't die. It dematerialized. The pressing plant became an encoder, the truck became a delivery feed, the rack became a search index, the cash register became a royalty report — and the label's release department became you.

That last part is the point of Part VIII. For most of recording history, the chapters you've already finished — capture, production, mixing, mastering — were maybe half of what a label did. The other half was this: manufacturing, distribution, metadata (they called it "label copy"), rights administration, marketing. When the bedroom era handed you the first half, it quietly handed you the second half too. Nobody mentions that at the FL Studio download screen.

Here's the good news, and it's genuinely good: the second half is easier than the first. You spent thirty-four chapters training your ears because mixing is a craft that takes years. Distribution is a system that takes a chapter. It rewards exactly the skills you already have — the file hygiene from Chapter 2, the naming discipline from Chapter 19, the checklist habit from Chapter 30. You don't need new ears for this. You need the map.

  THE ROUND TRIP OF ONE STREAM

  YOU                  THE GATE                THE STORES                THE LISTENER
 ┌─────────┐   WAV +  ┌─────────────┐  feeds  ┌─────────────────┐       ┌──────────┐
 │ Master  │──meta───▶│ DISTRIBUTOR │────────▶│ DSPs: Spotify,  │──────▶│  ▶ play  │
 │ + alts  │  data    │ (aggregator)│         │ Apple Music,    │ audio │  button  │
 │ + art   │          │ validates,  │         │ YouTube Music,  │       └────┬─────┘
 └─────────┘          │ formats,    │         │ Amazon, Tidal,  │            │
      ▲               │ delivers    │         │ Deezer, et al.  │            │ play
      │               └──────┬──────┘         └────────┬────────┘            │ logged
      │                      ▲                         │                     │
      │   payout             │   monthly royalty       │  streams counted,   │
      └──── (minus the ──────┴──── reports, matched ───┴── pooled, valued ◀──┘
            distributor's          by ISRC
            cut or fee)

One file goes right; months later, money comes left — matched to your recording by the codes you'll learn in this chapter.

🔄 Check Your Understanding

1. Why can't you upload directly to the major streaming platforms?
2. Name three "label department" jobs that became yours the day you decided to self-release.
3. In the round-trip diagram, what travels left — and what does it travel along?

Verify

1. The platforms don't maintain supplier relationships with millions of individual artists. They take deliveries from distributors — companies that validate, format, and feed music plus metadata to every store, and handle the money coming back. (One major platform briefly tested direct artist uploads in 2018–2019, then shut the program down and pointed everyone back to distributors.)
2. Any three of: distribution operations, metadata/label copy, rights administration, royalty accounting, marketing/promotion, manufacturing/deliverables.
3. Money — royalty reports and payments — travels backward along the same chain: platform to distributor to you, matched to your specific recording by its ISRC.

The Plumbing: How a Finished Master Becomes a Play Button

The Pipeline, Named

Let's lay the pipe sections end to end before we inspect each one. A modern release passes through five stations:

1. Deliverables. The final master, its alternate versions (instrumental, clean), and artwork — printed and verified. You did the audio half of this in Chapters 31–33; Chapter 32 even had you print the alts through the identical chain while it was loaded. This chapter finishes the packaging half.
2. Metadata. Every fact the stores and the money need: names, titles, credits, codes, flags, tags. This is where most independent releases pick up their scars. It gets the longest section in this chapter because it deserves it.
3. The distributor. The gate. You choose one, you upload audio plus metadata, they review and deliver to the platforms.
4. The platforms — DSPs. The stores themselves. Your music goes live, your profiles display it, their systems recommend it (or don't).
5. The return path. Plays become reports become payments — on a lag, through pipes worth understanding before the first statement confuses you.

A release strategy — a term this chapter owns — is the plan that coordinates all five stations against a calendar: what you release (single? EP?), when, in what order, with how much runway, and what has to be true at each gate before you proceed. The strategy section of this chapter builds yours. The plumbing comes first, because strategy built on pipes you don't understand is a vision board.

The Gate: Distributors and DSPs

Two terms, owned here, used for the rest of the book:

A DSP — digital service provider — is industry shorthand for the platforms that deliver music to listeners: Spotify, Apple Music, YouTube Music, Amazon Music, Tidal, Deezer, and a long tail of regional and specialty stores. "Store," "platform," and "DSP" mean the same thing in this chapter; the trade press says DSP, so you should recognize it.

A distributor (you'll also hear aggregator — same role) is the company that delivers your music to the DSPs and routes the money back. For an independent artist, the distributor is the single business relationship that makes everything else possible.

Why does the gate exist at all? Because the DSPs want a handful of professional suppliers, not millions of amateur ones. A distributor speaks the platforms' delivery language (there's a real industry standard underneath this — DDEX, a metadata-and-delivery specification — which is why your release looks identical in structure to a major label's when it arrives), warrants that you actually own what you're uploading, screens for fraud and copyright junk, assigns identifying codes, and consolidates a million artists' royalties into accounting the platforms can process. In exchange for sitting in the middle, the distributor takes its toll — and the shape of that toll is the first real decision of your release.

Choosing Your Gate: The Three Pricing Models

I'm not going to recommend a distributor by name, and you should be suspicious of anyone who does without disclosing what they're paid for it — most "best distributor 2026" articles are affiliate marketing wearing a lab coat. Companies in this space rebrand, change terms, get acquired, and occasionally implode. The categories, though, have been stable for years, and the tradeoffs between them are honest math. (Appendix I keeps a current directory of names in each category.)

As of this writing, virtually every distributor prices one of three ways:

Three ways to pay the same toll — the right one depends on your catalog, cadence, and confidence. Terms drift; the structure doesn't.

This is Theme 3 in a business suit: every model is a tradeoff, and the right answer depends on facts about you, not facts about the distributor. A producer planning a single every six weeks amortizes a flat fee beautifully. A band releasing one EP and going quiet for two years might be better off paying a percentage of nothing. Run your own numbers: expected releases per year, honest revenue expectations (this chapter's sidebar will calibrate those), and tolerance for terms changing under you.

Beyond price, four questions matter more than most reviews admit. Ask them of any distributor you're considering:

One: can you leave? Catalogs outlive distributor relationships. Moving a release to a new distributor without losing its play counts and playlist placements depends on keeping the same ISRCs (next section) — so confirm you can export your codes and that the distributor doesn't claim them. The smoothest migrations re-deliver the same recordings with the same ISRCs and let the platforms match them; support for this varies, so ask before you're trapped, not after.

Two: what's the payout plumbing? Minimum balance before they pay, payout methods and their fees, and reporting granularity (can you see streams per store, per country, per track?). You're choosing an accounting department; act like it.

Three: what happens when something breaks? A wrong artist mapping, a takedown dispute, a stuck release — these are when, not if, and the difference between distributors is the difference between a support human who can file a ticket with a DSP and a chatbot that can't. Search the company's name plus "support" and read what current customers say.

Four: what's bundled, and is the bundle a leash? Some distributors bolt on publishing administration, video Content ID, sync pitching, promo tools. Some of those are useful. Some quietly take rights or percentages you'd rather assign deliberately — the publishing-admin checkbox in particular overlaps with decisions Chapter 36 will teach you to make on purpose. Until then: read what each add-on takes, and when in doubt, leave the box unchecked. You can opt in later; un-assigning rights is harder.

🔄 Check Your Understanding

1. A flat-fee distributor charges yearly; a percentage distributor takes a cut forever. What single fact about an artist most determines which is cheaper?
2. Why is "can I export my ISRCs?" a question about leaving a distributor?
3. Your distributor offers a publishing-administration add-on at checkout. What's this book's holding position, and which chapter finishes the decision?

Verify

1. Their honest ratio of releases (and earnings) to time. Frequent releases with some income amortize a flat fee; sparse releases with little income favor paying a percentage of not-much.
2. ISRCs are how platforms recognize a recording as the same recording. Re-delivering your catalog through a new distributor with the same ISRCs preserves play counts and placements; losing control of the codes means starting your numbers over.
3. Leave the box unchecked until you understand what rights and percentages it takes. Chapter 36 covers publishing — the decision belongs there, made deliberately, not at a checkout screen.

🧩 Productive Struggle

Before the next section teaches you the form, take five minutes and write the form yourself. Open your notes app and list every fact a streaming platform needs to know about your single before it can (a) display it correctly, (b) recommend it to the right listeners, and (c) pay the right people. Don't read ahead. Most producers confidently list five or six facts. The real form has more than twenty, and three of them — which you'll probably miss — are the ones that quietly decide whether your money and your credit arrive at all. Write your list, then read on and grade yourself.

The Metadata That Matters

Metadata — a term this chapter owns — is every fact attached to your audio: titles, names, credits, codes, flags, dates, tags. It's tempting to treat the upload form as paperwork between you and the play button. Wrong frame. Metadata is the release. The audio is one field among many. Stores display metadata, search indexes it, recommendation systems steer by it, and — the part that should straighten your spine — the money follows it. Every routing error in the payment system starts life as a typo in a form.

Liner notes died when physical packaging did. Metadata is the new liner notes, except now the liner notes are also the shipping label, the barcode, and the bank routing number.

Here's the anatomy of one track's metadata, annotated. This card is the chapter's central artifact — exercise C1 has you fill one out for your own single, and you'll keep it under version control like a session file:

 ┌──────────────────────────────────────────────────────────────────┐
 │  TRACK METADATA CARD — "Static Bloom"                            │
 ├──────────────────────────────────────────────────────────────────┤
 │  Title            Static Bloom            ← title ONLY. No       │
 │  Version          (blank — original)        "(feat...)", no      │
 │                                             "[prod. by...]",     │
 │                                             no "(Official)"      │
 │  Primary artist   Jaylen Cole             ← must match the       │
 │  Featured artist  Demi Wren                 artist page EXACTLY,  │
 │                                             release after release│
 │  Writers          J. Cole / D. Wren /     ← LEGAL names,         │
 │                   T. Park                   every writer, → Ch36 │
 │  Producer         Jaylen Cole             ← credit field,        │
 │                                             not the title        │
 │  ISRC             (assigned at upload)    ← one per VERSION      │
 │  Genre            Electronic / Pop        ← steering data,       │
 │  Language         English                   not identity         │
 │  Explicit         No                      ← accuracy required    │
 │  ℗ line           ℗ 2026 Jaylen Cole      ← recording owner      │
 │  © line           © 2026 Jaylen Cole      ← artwork/packaging    │
 │  Label            (self — your imprint)                          │
 │  Release date     Friday, 4+ weeks out    ← the runway           │
 ├──────────────────────────────────────────────────────────────────┤
 │  RELEASE level:   UPC (one per product) · artwork 3000×3000,     │
 │                   no logos/handles/URLs · territory: worldwide   │
 └──────────────────────────────────────────────────────────────────┘

The metadata anatomy card: the audio is one attachment; this is the release.

Now the fields that break releases, in order of how often they break them.

Artist name: the consistency doctrine. Platforms build artist pages by matching the artist-name string on incoming releases. The match has to be exact — "Jaylen Cole," "Jaylen Cole Beats," and "JaylenCole" are three different artists to an ingestion system, and your second release can land on a stranger's page or strand itself on a fresh empty one. The doctrine: decide your artist name once, write it into your project documentation (the Chapter 19 session-hygiene habit, promoted to career level), and reproduce it character-for-character forever. Before you commit to a name at all, search every major DSP for it. If somebody already releases under your name — and at the scale of modern catalogs, somebody might — you're choosing between distinguishing yourself and a career of mis-mapped releases and split search results. Distributor forms typically let you link a new release to your existing platform artist page (or request a brand-new page when a name collides with a stranger's); using that linkage correctly is the single cheapest insurance in this chapter. Case study 2 watches Jaylen nearly learn this the hard way.

One more name rule: featured artists go in the featured artist field, not hand-typed into the title. Every DSP publishes a metadata style guide, and the guides disagree on rendering — some stores display "Title (feat. X)," others list the feature in the artist line — so the distributor formats it per store from the role field. Type it into the title yourself and you've broken the formatting on half the stores and possibly broken Demi's artist-page mapping in the bargain. Same logic, harder rule, for producer tags: "[prod. by …]" in a title field violates basically every style guide on earth. Producer credit goes in the producer credit field. It still counts. It's still yours.

Writer credits: required, and not optional-required. The composition — the song under the recording — has writers, and the stores' style guides increasingly require writer names on delivery, in legal names, not artist names or aliases. This isn't bureaucratic fussiness; it's the thread the other half of the money hangs on. Here's the handshake with the next chapter, and it matters enough to land hard: the writer field in your upload form is a public credit, not a contract. It tells the stores who wrote the song. It does not establish what percentage each writer owns, and it pays nobody by itself. The document that does that is a split sheet, and Chapter 36 opens with Jaylen, Demi, and Theo at a kitchen table writing one for this exact track — before it earns a dollar, which is the only sane time to do it. For now, the rule is accuracy: every actual writer, listed, in legal names. Omitting a co-writer from the form doesn't make their share yours; it makes your metadata false and your future conversation worse.

ISRC and UPC: the license plate and the shipping container. Chapter 31 introduced the ISRC — International Standard Recording Code — as the identifier for a specific recording, per version. Now the full picture, because these two codes are the skeleton of the entire accounting system:

        ISRC  ·  one per RECORDING (per version)

        QZ - AB1 - 26 - 00007
        ─┬─  ─┬──  ─┬─  ──┬──
         │    │     │     └─ designation: 5 digits, unique
         │    │     │        within registrant + year
         │    │     └─ year of reference (2 digits)
         │    └─ registrant code (3 characters):
         │       who issued the code
         └─ country / allocator prefix (2 characters)

        UPC  ·  one per PRODUCT (release)
        a 12- or 13-digit barcode identifying the
        single/EP/album as a sellable unit

Two codes, two jobs: the ISRC follows the recording everywhere forever; the UPC identifies the release that packages it.

The division of labor: the ISRC identifies a recording — it's the license plate every play, report, chart tally, and royalty payment travels on. The UPC identifies a product — the release as a unit of commerce, a direct descendant of the barcode on a CD longbox. One recording, one ISRC, for life. But the same recording can ride inside multiple products: "Static Bloom" the single (its own UPC) and Glass Hours the EP (another UPC) both contain the identical recording with the identical ISRC — and because the ISRC matches, the platforms recognize it as the same track, and its play counts and playlist placements carry over rather than resetting. Hold that thought; it's the load-bearing wall under the waterfall strategy in the next section.

The per-version rule from Chapter 31, now operational: every distinct audio gets its own ISRC. Main version, instrumental, clean edit, future remix, re-recorded version — separate codes, every one. A metadata correction (fixed typo, added credit) does not need a new ISRC, because the recording didn't change. A remaster is a judgment call with a default answer of "new code if the audio meaningfully differs." When in doubt: did the sound change? New sound, new code.

Mechanics, honestly: your distributor will assign ISRCs and a UPC free at upload, and for your first releases that's the right answer — take the free codes, record them in a ledger (a plain spreadsheet: version, ISRC, UPC, date, distributor), and treat that ledger like a session backup. The control path — registering your own registrant code with your country's ISRC agency (the system is administered internationally by IFPI through national agencies; in the US there's a dedicated ISRC agency, and fees are modest as of this writing) — buys you codes that are yours before any distributor touches them, which makes catalog moves cleaner. Either path works. The ledger is non-negotiable on both.

Explicit flags: accuracy, both directions. Each track carries an explicit-content flag, and the only rule is the truth. Mark explicit content explicit — stores enforce this, listeners filter by it, and mislabeled tracks get corrected, filtered, or pulled. But don't over-flag either: a "clean" label on a clean edit is its own marker (set in the version field — "Static Bloom (Clean)" exists as a separate recording, separate ISRC, flagged clean), and flagging a genuinely clean original "explicit" buries it behind parental filters for no reason. One famous gotcha: the flag is per track, and an explicit lyric anywhere in the track makes the track explicit — there's no "barely" setting.

Genre tags: steering data, not identity. The genre field feels like a self-expression box. It isn't. It's routing information: it suggests which editorial teams see your pitch, which browse categories shelve you, and it seasons the cold-start guesses recommendation systems make before real listening data exists. So answer it like a router, not like a poet — tag where your listeners actually live, which is an ears-first call you're equipped to make: it's Chapter 18's genre-audit skill pointed at commerce. The honest test is Theme 4 wearing a business suit: pull up your Chapter 4 reference playlist — the three tracks your single honestly belongs between — and look up how they're categorized. That's your tag. Jaylen's instinct, after years of beat-store uploads, is to tag everything hip-hop; "Static Bloom," with Demi's sung vocal over an electronic-pop arrangement, would be mis-routed there — pitched to the wrong editors, shelved in the wrong browse rows, seeded to the wrong listeners, all of whom skip it, which the system duly notes. Wrong tag, real cost.

The small fields that aren't. Language (of the lyrics — it gates lyric display and regional features). The ℗ and © lines: ℗ (the "phonogram" symbol) declares the year and owner of the sound recording; © declares the artwork and packaging — and if that distinction smells like Chapter 36's two-copyrights model, your nose works; this is its first paperwork appearance. Label name: you almost certainly don't have a label, so this is your imprint — make one up once ("Glass Hours Audio"), keep it forever; it's another consistency string. Artwork: as of this writing the common spec is a 3000×3000-pixel square, and the style guides' real rules are about content — no URLs, no social handles, no pricing, no platform logos, nothing blurry or misleading. Your art will be judged at thumbnail size in a moving feed; design for 64 pixels, deliver 3000.

🔄 Check Your Understanding

1. "Static Bloom" the single and Glass Hours the EP both contain the same recording of the song. How many ISRCs and how many UPCs are involved, and what does the shared code accomplish?
2. Your collaborator sang the hook and co-wrote the melody. Name the two different fields in the upload form that involve her, and what each one does and doesn't establish.
3. Why is the genre tag a routing decision rather than an identity statement — and what's the T4 method for choosing it?

Verify

1. One ISRC (same recording in both products) and two UPCs (single and EP are separate products). The shared ISRC tells every platform it's the same track, so plays and placements accumulate instead of resetting.
2. Featured artist (a performance credit — affects display and her artist-page mapping) and writer credit (a composition credit, legal name — feeds the credit systems but does not establish ownership percentages; that's the split sheet, Chapter 36).
3. The tag steers editorial pitching, browse shelving, and cold-start recommendations — it determines who gets sent the track. Method: check how the three reference tracks your single honestly sits between are categorized, and tag there.

⚙️ Advanced Sidebar: How Streaming Payouts Actually Calculate

Somewhere this month, someone will tell you confidently that a stream "pays" some specific fraction of a cent. Here's why that sentence — any version of it — misunderstands the machine.

Most major platforms, as of this writing, run some variant of a pro-rata pool model. Plain words: in each market, for each month, for each subscription tier, the platform gathers the revenue (subscriptions, ad income), keeps its negotiated share, and puts the rest in a pool for rights holders. Then it counts every eligible stream in that market that month — all of them, every artist — and divides the recording-side pool by that total. Your payout is the pool times your fraction of total streams, routed to your distributor, minus the distributor's toll, into your balance.

Look at what that structure implies. There is no rate card. The "per-stream rate" is an output of the month's arithmetic, not an input — it moves with the month's total revenue and the month's total streams, neither of which anyone controls. It differs by country (subscription prices differ), by tier (an ad-supported stream draws from a different, usually shallower pool than a paid-subscriber stream), and by month. When you see per-stream figures quoted online — fractions of a US cent, commonly — you're seeing somebody's back-calculated average across all of that, folklore with a kernel of arithmetic. Useful for setting expectations (calibrate low), useless as a rate card, and never something to put in a business plan with three decimal places.

The model has a much-debated structural quirk: your subscription money doesn't follow your listening. It goes into the big pool and flows by aggregate share — so a subscriber who streams one obscure ambient album all month is, in effect, part-funding whatever dominated the pool. The proposed alternative — user-centric payment, where each listener's fee divides among only what they played — gets studied and piloted periodically, and platform payout structures have been visibly in motion in recent years: as of this writing at least one major platform stopped paying recording royalties on tracks below a minimum annual stream count (around a thousand, when the policy landed), redirecting that long-tail money upward into the pool. Policies will have shifted again by the time you read this. The durable lessons survive the shifting: payouts are shares of pools, not prices per play; the pools differ by market and tier; and nobody owes you the average.

One boundary line, so the sidebar doesn't overreach: everything above describes the recording side — the money that reaches you through your distributor. Every stream also generates composition-side money (the song under the recording), which flows through entirely different pipes to writers and publishers. That's Chapter 36's territory; the sidebar's job is to make sure you know the pool you're hearing about is only one of the pipes.

The Strategy: Releasing on Purpose

The plumbing carries anything you put in it, on any schedule. Strategy is choosing what and when. Two doctrines and a checklist cover most of what a first release needs.

The Release Timeline Doctrine: Four Weeks Minimum

The single most common self-release mistake — more common than every metadata error combined — is finishing a master on Tuesday and releasing it Friday. The impulse is pure and I respect it: the track is done, waiting is agony, and the calendar widget will happily accept tomorrow. But a day-of-or-close upload doesn't merely "miss optimizations." It structurally forfeits almost everything the modern release system can do for you, and it can't be repaired afterward — the system's biggest levers only exist before release day.

Count the forfeits. Distributor review and store propagation take days and arrive raggedly, so a too-tight date means your track materializes on some stores after "release day" — pre-release QA time goes to zero, and metadata errors go live instead of getting caught. The editorial pitch portals inside the artist dashboards — the front door to platform-curated playlists — accept unreleased tracks only; stated minimums hover around a week before release as of this writing, and editors plainly work further ahead than the minimum. Submit late and the front door is closed; submit day-of and it was never open. Follower-notification features (the new-release feeds that put your track in front of people who already follow you — the warmest audience you have) seed from releases sitting in the system ahead of their date; a day-of upload can sail right past that cycle. And a pre-save — a term this chapter owns: a listener pre-authorizes the platform to add your track to their library the moment it goes live, the streaming era's pre-order — is definitionally impossible without a gap between announcement and release.

So: four weeks between a confirmed, distributor-accepted upload and release day. Not four weeks from "I finished the master" — four weeks from upload accepted. Here's the runway, drawn:

 THE FOUR-WEEK RUNWAY                                  ▼ = a gate you must pass

 WEEK -4   ▼ upload accepted by distributor (audio + metadata FINAL)
           ▼ release date locked: a Friday, 4+ weeks out
           ▼ editorial pitch submitted via artist dashboard(s)
           ▼ pre-save link generated and tested
 ──────────────────────────────────────────────────────────────────
 WEEK -3   ▼ artist profiles claimed on every major DSP (do it NOW
             — verification takes time) · bio, photos, links dressed
           · announcement + pre-save campaign begins (content: Ch 37)
 ──────────────────────────────────────────────────────────────────
 WEEK -2   · pre-save push continues · collaborators re-share
           ▼ verify release shows "scheduled" correctly in dashboard
           ▼ alt versions / credits / artwork triple-checked
 ──────────────────────────────────────────────────────────────────
 WEEK -1   ▼ confirm track visible as "upcoming" where supported
           · release-day checklist staged (links page ready to flip)
           · final social drumbeat (Ch 37's content calendar)
 ──────────────────────────────────────────────────────────────────
 DAY 0     ▼ FRIDAY: execute the release-day checklist (below)
 ──────────────────────────────────────────────────────────────────
 WEEK +1   · follower-feed Friday (new-release features cycle weekly
             as of this writing) · thank-yous · log the numbers
           · start the next song (the Ch 39 rule, previewed)

The runway, gate by gate. Everything above Week -2 is impossible to retrofit after release.

Why Friday? In 2015 the industry standardized a global release day — Friday — ending decades of regional patchwork (US records came out Tuesdays; ask an older engineer about Tuesday-midnight lines at the record store). The platforms' weekly rhythms (editorial refreshes, new-release feeds) synchronized around it, as of this writing. The counterargument is real, though, and it's a Theme 3 tradeoff: Friday is also the loudest day — every major label release lands on top of you. An off-Friday release trades the platform rhythm for quieter shelf space, and for an artist whose listeners come from their own channels rather than playlists, that trade can be sane. Default Friday; deviate on purpose.

🔍 Why Does This Work?

Why does a runway mechanically beat a surprise drop? Because the discovery systems read day-one engagement density as a quality signal. The platforms publish only sketches of how recommendation works, but the sketches and the observed behavior agree on the shape: when a track goes live and immediately accumulates saves, library adds, completions, and playlist adds from its first listeners, the system's cheapest interpretation is "the people this was made for like it" — and it widens the test audience. Pre-saves convert to library adds at midnight; followers who were warmed up all month actually listen today instead of "sometime"; the pitch (if it landed) supplies a seed audience the same week. The runway doesn't make the music better — it concentrates your existing audience's genuine response into the window where the system is paying closest attention, instead of dribbling the same response invisibly across six weeks. Same people, same love, radically different signal. That's also the honest demystification: none of this manufactures interest that isn't there. It stops you from wasting the interest that is.

Single, EP, or Album? Streaming Economics, Both Sides

You have five mastered tracks. Do you drop Glass Hours whole, or lead with "Static Bloom"?

The single-cadence argument. Streaming rewards recurring presence. Each release is a fresh pitch opportunity, a fresh follower-feed appearance, a fresh reason for the audience to come back — and a catalog that arrives as six singles over nine months generates six of those cycles, while the same songs as one album generate one. Discovery systems also test tracks individually; an unknown artist's album is, from the system's view, ten simultaneous cold starts competing with each other. The widely repeated independent-artist playbook of recent years — a single every four to eight weeks — exists because consistent cadence compounds: each release re-engages the listeners the last one earned. (If that "compounding catalog" logic sounds familiar, it's the Chapter 6 template argument at career scale.)

The album-statement argument. And yet. Albums remain the unit of artistic reputation — the format reviews engage with, the thing that earns a profile rather than a placement, the format sync supervisors and serious fans dig into. Some music is built as a sequence (you sequenced Glass Hours in Chapter 32 with gaps and level steps for a reason) and dies dismembered. And there's a fan-depth argument: singles recruit listeners; bodies of work convert them into people who care. Artists with patient audiences — and any artist whose genre culture venerates albums — can rationally choose the statement over the cadence.

The waterfall, the middle path. The release strategy that's become the independent default, as of this writing, threads the needle: lead with singles, then release the EP/album containing those singles — same recordings, same ISRCs. Because the codes match, the platforms recognize the singles inside the new product, and their accumulated plays and placements carry into it: each release inherits the last one's momentum instead of resetting it. (Some artists run it even more literally — re-delivering a growing product where each "single" is the previous release plus one track. Distributor support for the mechanics varies; what's universal is the principle: ISRC continuity is what lets a catalog accumulate instead of restart.) For Glass Hours: "Static Bloom" leads as the single — four weeks of runway, pitch, pre-save — and the EP follows behind it, carrying the single's numbers with it. Jaylen gets cadence and the statement; the only cost is patience, which is the one ingredient this chapter keeps prescribing.

The honest Theme 3 summary: cadence buys discovery cycles and costs cohesion; statements buy depth and cost frequency; the waterfall buys both and costs time. There's no rule. There are consequences, and now you know them.

🔄 Check Your Understanding

1. Name three concrete things a day-of upload structurally forfeits, and which (if any) can be recovered after release.
2. A pre-save campaign "worked": 80 people pre-saved. What mechanically happens at midnight, and why does the system care?
3. What single piece of metadata makes the waterfall strategy work, and what does it prevent?

Verify

1. Editorial pitch eligibility (pitches accept unreleased tracks only), the pre-save window (definitionally pre-release), follower-feed seeding for release week, and pre-release QA time. Essentially none recover afterward — the levers exist only before release day.
2. The platform adds the track to 80 libraries at release, and those listeners get it surfaced immediately — concentrating genuine day-one saves, adds, and completions into the window where discovery systems weight engagement most heavily.
3. The shared ISRC. The platforms recognize the single inside the later EP as the same recording, so plays and placements accumulate across releases instead of resetting with each product.

The Release-Day Checklist

Release day is an execution day, not a decision day — every decision was made weeks ago, which is the point. The checklist (Appendix G carries a printable version; Chapter 39 fires it for real):

Release day in nine rows. Everything here was staged during the runway; today you flip switches.

A timing honesty note: "midnight release" typically means midnight per territory as stores roll over local time, as of this writing — your track dawns across the planet in waves, New Zealand first. Don't panic at 12:01 your time when a store hasn't flipped; check the pattern, then escalate.

Where the Money Goes: The Two Pipes (a preview)

Money earned by one stream flows back along two separate plumbing systems, and confusing them is the most expensive confusion in independent music:

                          ONE STREAM
                              │
              ┌───────────────┴────────────────┐
              ▼                                ▼
      RECORDING side                   COMPOSITION side
      (the master — the         (the song — melody/lyric,
       actual audio file)        the thing that's written)
              │                                │
        DSP → distributor →            DSP → collection
        YOUR BALANCE                   infrastructure → the
              │                        song's WRITERS and
        minus the toll,                PUBLISHERS — by routes
        on a lag of months             Chapter 36 maps fully
              ▼                                ▼
        arrives because of            arrives because of your
        your ISRC + distributor       writer credits + the
        relationship                  registrations Ch 36 teaches

Two pipes from one play. Your distributor handles exactly one of them.

The chapter-35 facts you need now: your distributor collects and pays the recording side only, on a real lag — platforms report and pay months behind the listening month, as of this writing, and your balance pays out past a minimum threshold. Read your first statement with this sidebar's pool model in mind and it'll make sense; read it expecting a per-stream rate card and it won't. The composition side — the money the song earns, for its writers — does not arrive through your distributor, and (the cruel part) much of it goes uncollected entirely unless the writers register with the right organizations. You typed writer names into a form this chapter; Chapter 36 turns those names into registrations, splits, and collected money. If you write your own material and skip Chapter 36, you're doing half this chapter's work for half the pay.

Your Profiles Are Storefronts

The artist page is where a curious stranger lands ten seconds after your music interests them, and as of this writing every major platform offers an artist-services portal to claim and run yours — Spotify for Artists and Apple Music for Artists are the big two; the others follow the same shape. Claiming is not optional, for a blunt mechanical reason: the editorial pitch tools live inside these portals. No claimed profile, no pitch. Claim during Week -3 at the latest (verification can take days), then dress the storefront: real bio with your actual story (the Chapter 5 manifesto, compressed), photos that look intentional at thumbnail size, links that work, the new release pinned where the platform allows featuring.

Then run this book's most reliable move — Theme 4, pointed at commerce. You A/B'd your mix against reference tracks for twenty chapters; now A/B your presentation against reference releases. Pick three artists one rung above you in your exact lane and audit their storefronts like you audited their mixes in Chapter 4: how they bio, how they sequence their catalog visually, what their artwork does at thumbnail scale, how their release cadence reads on the page. The gap between their storefront and yours is — say it with me — your to-do list.

International and Edge Realities

One honest paragraph, because the worldwide checkbox hides a planet: your release defaults to global, and global is almost always right — territory restrictions are usually a rights problem (a sample cleared for some regions, a license with borders — Chapter 36 territory), not a strategy. But know what worldwide means: subscription prices differ wildly by country, so the pro-rata pools differ too, and a thousand streams in one market can pay a different amount than the same thousand elsewhere — your per-country statement lines aren't errors. Language and localized-title metadata fields exist for a reason; fill them if they apply. Regional stores you've never heard of genuinely matter in their markets — another reason a distributor's "150+ stores" line isn't pure puffery. And one fraud flag while we're abroad: services selling "guaranteed streams" run bot farms through exactly these pipes; platforms and distributors detect artificial streaming, strip the numbers, withhold the money, and can take down the release or the account. Never. Chapter 37 covers the adjacent playlist-scam economy; the distribution-chapter version is one word long.

Common Mistakes and the 60-Second Fix

The diagnose-then-fix table, distribution edition — tape it inside your upload-day checklist:

Every row is someone's worst release week. The left column is what it feels like; the middle is what it is.

🔄 Check Your Understanding

1. Which two pipes carry a stream's money, who handles each for an independent artist, and which one fails silently if ignored?
2. Why does this chapter tell you to claim your artist profiles weeks before release rather than on release day?
3. Your single went live an hour ago and the featured artist's name is misspelled. Is this fixable, and what's the mechanism?

Verify

1. Recording side (DSP → distributor → you — your distributor handles it) and composition side (DSP → collection organizations → writers/publishers — handled only if writers register, per Chapter 36). The composition pipe fails silently: unregistered money goes uncollected, and nobody sends a warning.
2. The editorial pitch tools live inside the claimed profiles, verification takes time, and the storefront needs dressing while the pre-save campaign is driving strangers to it — all Week -3 work at the latest.
3. Yes — metadata corrections (unlike audio changes) can typically be delivered post-release through your distributor. File the correction immediately with the UPC/ISRC in hand; rendering updates as stores re-ingest.

Project Checkpoint: Building "Static Bloom"

Where the track stands: mastered (Chapter 32, two-stage limiting into a -1.0 dBTP ceiling), loudness-verified (Chapter 33), translation-checked past stereo (Chapter 34), with the alternate masters — instrumental and a cappella — printed through the identical chain back in Chapter 32 while everything was loaded. "Static Bloom" is clean as written, so no clean edit is needed; one line in "2 A.M." isn't, which is why that track's clean edit exists in the EP deliverables and will carry its own ISRC and a Clean version tag. This week the single became a release: deliverables verified and named per the Chapter 19 convention; the metadata card filled — title "Static Bloom," version blank, primary artist Jaylen Cole, featured artist Demi Wren, writers Cole/Wren/Park in legal names (the percentages wait for next chapter's kitchen table — deliberately); genre Electronic/Pop per the reference-playlist test; explicit: no; ℗/© 2026 lines under Jaylen's imprint. ISRCs assigned at upload — main and instrumental — and logged in a new ledger file that lives beside the session backups. UPC assigned to the single; Glass Hours will get its own and inherit the single by ISRC, waterfall-style. Distributor chosen by the tradeoff table (flat-fee made sense for a producer planning a release cadence — your math may differ). Release date: Friday, four weeks and a day out. Pitch submitted. Pre-save link live and tested. Case study 2 documents the three metadata mistakes that nearly shipped.

Your assignment — release prep:

1. Verify deliverables. Final master at -1.0 dBTP or under, plus the alt versions your track honestly needs (instrumental almost always; clean only if there's something to clean). If you skipped printing alts in Chapter 32, do it now through the identical chain.
2. Fill the metadata card from this chapter for your single — every field, including writers in legal names. This is exercise C1; it takes an evening and prevents years.
3. Choose your distributor category with the tradeoff table and your own honest numbers (releases per year, expected earnings, exit-friendliness). Appendix I has the current directory.
4. Walk the upload — for real if you're releasing, as a complete dry run if you're not — and schedule a Friday at least four weeks out. Generate the pre-save link. Submit the pitch.
5. Claim your artist profiles the moment the release appears in the system, and start the ledger: ISRCs, UPC, dates, distributor, login locations.

Genre adaptations: Hip-hop: producer credits go in credit fields, never titles — retire the beat-store "[prod.]" habit at the DSP gate; features are mapping events for both artists' pages, so coordinate name strings with every collaborator; instrumentals are non-negotiable deliverables in your economy. Rock/band: the band is the primary artist — one canonical string, decided once, in writing, because five members will type it six ways; all writers listed regardless of who sings; the "who is the artist entity" question previews next chapter's splits conversation. Electronic: version fields are your native habitat (Extended Mix, Radio Edit, Remix — each its own ISRC); the label-imprint field matters more in your scene culture, so name yours deliberately; remixes of other people's tracks need permission you don't have yet — Chapter 36 before upload, always.

Cross-Chapter Connections

Backward: Chapter 31 introduced ISRC and the deliverables mindset this chapter operationalized; Chapter 32 printed the masters and alts this chapter shipped (and sequenced the EP the waterfall will deliver); Chapter 33's normalization threshold is why your master needed no per-platform loudness surgery before upload — one honest master serves every store; Chapter 34's fold-down checks were the last audio gate before the pipe. The metadata discipline is Chapter 19's session hygiene at career scale; the clean edit is Chapter 15's editing craft with a commercial job; the lossy-encode realities trace to Chapter 2; and the genre tag is Chapter 18's audit plus Chapter 4's reference method, pointed at commerce.

Forward: Chapter 36 turns this chapter's writer-credit fields into split sheets, registrations, and the composition-side money this chapter could only diagram; Chapter 37 fills the four-week runway with an actual campaign and finishes the analytics conversation the release-day checklist opened; Chapter 39 executes everything — the checklist fires for real and the track goes live.

Spaced Review

Three from the recent past — answer before you peek.

1. (Chapter 34) Most listeners who hear an Atmos master hear it folded down to binaural on ordinary headphones. Why did that fact lead Chapter 34 to a "stereo first" doctrine for independent releases?
2. (Chapter 33) Your friend wants to master their single to maximum loudness "so it hits harder on Spotify." What does platform normalization actually do to that plan?
3. (Chapter 31) You're delivering a single plus its instrumental to a distributor. How many ISRCs, and what's the rule's name?

Verify

1. Because the dominant playback reality is headphones rendering a fold-down, your stereo craft — balance, space, clarity — remains the product most listeners receive; immersive is an additional deliverable when stakes justify it, not a replacement for getting stereo right.
2. Platforms turn playback down to a reference loudness (commonly around -14 LUFS, as of this writing), so the slammed master plays at the same loudness as a dynamic one — minus the dynamics it destroyed to get there. Loudness beyond the reference buys nothing; dynamics are a choice again.
3. Two — the per-version rule. Every distinct recording (main, instrumental, clean, remix) carries its own ISRC, because the code is how plays and payments find each specific piece of audio.

What's Next

Your release is scheduled and your metadata names three writers — and right now, nothing on paper says what percentage of the song each of them owns. Chapter 36 fixes that before the first dollar arrives: Jaylen, Demi, and Theo at a kitchen table with a split sheet, the two-copyrights model that runs the entire music business, and the registrations that turn your writer credits into collected money instead of donations. Before you go: the exercises build your metadata card and four-week calendar — the two artifacts your actual release will run on — and the quiz makes sure the plumbing stuck.