DDEX MEAD is the delivery message that carries mood, genre, instrumentation and focus-track data to streaming services, and most independent releases never send one.
That gap is the difference between a catalogue a recommendation engine can read and a catalogue it has to guess at.
The message that sits next to your release
DDEX means Digital Data Exchange, the standards body the industry uses to move release data between labels, distributors and streaming services.
ERN means Electronic Release Notification. It is the message that delivers label copy: track titles, ISRCs, credits, territory rights, release date. We covered it in our ERN explainer.
MEAD means Media Enrichment and Description. It is a separate message that carries everything ERN does not.
DDEX’s own documentation describes more than 30 descriptive mechanisms in MEAD: genre, mood, theme, instrumentation, tempo, key, vocal register, chorus timestamps, recording location, awards, chart history, pronunciation aids and focus-track flags.
The standard is modular. DDEX notes that useful implementations typically populate 5 to 20 data points rather than the full set.
MEAD 1.1 is the current version and is still a Candidate Standard. The format was shaped in 2017 sessions involving Apple, Amazon, Spotify, Pandora and the major labels, as Music Ally reported when version one landed in October 2019.
The one test with published numbers
In June 2022, DDEX published results from a joint Amazon Music and Universal Music Group implementation, presented at the Music Biz conference in Nashville.
The test compared the 90 days before and after tracks were tagged with higher-quality genre and mood data delivered through MEAD.
- Overall streams rose 4.6 percent
- Streams from algorithmic stations and playlists rose 11.9 percent
- Skip rate fell 7.8 percent
Treat those figures with the caveat they deserve. One label, one service, four years ago, and no comparable test published since.
The direction is the useful part. The lift concentrated in algorithmic surfaces, which is precisely where descriptive metadata gets consumed.
Why the pitch form is not the same thing
Artists assume the Spotify for Artists pitch form covers this. It does not.
Spotify’s own documentation is explicit: you can pitch one song at a time, and you cannot pitch another until the first one goes live. Music should be delivered at least seven days before release.
The form asks for genre, mood, instrumentation and a written pitch capped at 500 characters, per Ryan DiLello’s July 2026 pitching guide on Hypebot.
So a ten-track album gets one described song, once, at one service.
MEAD applies to every recording, at every service that accepts it, and can be sent for catalogue released years ago.
Five questions for your distributor
Ask these before your next delivery
- Do you send MEAD messages at all, or only ERN?
- How many MEAD fields do you populate, and which ones?
- Can I submit enrichment data for back catalogue after release?
- Which services do you actually deliver MEAD to?
- Do you pass my values through, or auto-fill from your own genre taxonomy?
Most self-serve distributors answer the first question with silence, because a single-page upload form has nowhere to put instrumentation or a chorus timestamp.
It is the same pattern we found in AI disclosure as a delivery field and in DDEX’s catalogue transfer standard. The standard exists, the DSPs accept it, and the middle layer decides whether you get to use it.
What the gap actually costs
The economics are unglamorous. MEAD does not raise your per-stream rate and does not unlock a royalty line. Anyone selling it as a payout mechanism is overselling.
What it changes is whether a recommendation engine can place your record next to the right neighbours, which is the last free distribution available at scale.
For catalogue outside the Anglo-American core the stakes are higher. A Piseiro record, an Amapiano single or a V-pop ballad meets a DSP genre taxonomy with no matching bucket. If you do not describe the record, the platform’s classifier will, and it will reach for the nearest large genre it recognises.
That is a distribution question before it is a marketing one. A DDEX-native pipeline treats enrichment as a field you fill, not a favour you request. Start with what metadata actually is, then audit one release against the five questions above.
