event-content-repurposing

Event Content Repurposing

You decide what happens to an edition's captured output once a master file already exists.

• Inputs, every one of them somebody else's output: a published recording, slides, a transcript, a photo archive, hallway notes.
• Decisions you own: which derivative artifacts get made, in what order, on what cadence, under which licence, and what each speaker is asked to do about the piece cut from their own talk.
• Outputs you produce: published derivative artifacts, plus a content inventory the next edition's campaign can schedule.

Where you start, and the six lines you never cross

Sibling	Owns	The line
event-production	Master recording, capture through published file	Take the master as handed over; never re-edit, re-publish, or send a speaker their own link. You begin at the derivative.
event-social-media	The event's own channels	You decide which derivative exists and when it drops; that skill owns platform, hashtag, copy, and posting.
event-speaker-experience	The consent record, post-event follow-up	Enforce the consent record; never collect or re-ask it. You decide only whether the speaker amplifies the derivative.
event-marketing-plan	The attendee-acquisition plan and its funnel targets	Repurposed content is inventory you hand up, never a target you set.
tech-podcast-youtube-channel	Any standing owned-media property	A one-off derivative is yours; a format outliving any single edition is a property and belongs there.
event-debrief	The organizers' public recap	That skill covers how the edition went; yours covers what was said in the sessions. Write one of each.

Which siblings state the same boundary back, and which do not: references/decision-rationale.md.

Every rung, axis ordering, efficiency line, default and promotion condition in the four menus below is this skill's own construction, not a measured benchmark. Each is a default, not a law: it shifts with context and with who executes it. Re-rank all four menus against what you already know:

• A volunteer who edits video for a living.
• An existing subtitling community.
• An audience that finds the event through search.
• A speaker with a large following of their own.
• A consent record that stops at "recording and publication".

Each of these overturns a default rung.

Interview

Ask one question at a time, multiple-choice where possible. Questions 4-6 exist because the menus below diverge sharply on time-to-effect, durability, and effort; those defaults cannot be picked for the user.

1. What exists to work from and in what state (published master, raw files, slides, transcript, photo archive)? Who handed it over, anything still owed?
2. Where is the consent record, what does it cover per session format, does it reach derivative works or stop at recording and publication?
3. What photo opt-out did badges carry, retrievable per photo or only for the room as a whole?
4. One-off edition, or one year of a recurring event? A compounding mandate promotes every rung whose payoff lands after this edition ends.
5. Any date anything must land by (campaign moment, sponsor report, partner calendar)? A hard date promotes rungs shipping in one push.
6. Who owns publishing after the event, how many hours a week, until what date? Decides Menu 1 more than any other answer.
7. Which sessions did people actually ask about - in the room, in feedback, in chat?
8. Does the event already run a standing owned-media property with its own audience?
9. Is search a real route into this event for people who've never heard of it?
10. Has a licence been decided, and by whom? If nobody has, you are deciding it - say so.
11. Is the audience multilingual, and did the accessibility work commit to anything that implies text derivatives?
12. Which speakers, if any, have an audience of their own larger than the event's?
13. Is this the event's first edition, and does the program include a keynote session?

What actually splits this work

One axis reorders every menu below: whether anyone owns publishing after the crew disperses. An owner with a standing weekly slot and an owner with one free afternoon are running two different plans. Every menu below assumes you have asked which one you have.

samber/dev-event-organizer-skills@event-production names the same risk in its own failure list: a pipeline with no owner after the event, where files are collected, the crew disperses, and the videos are published very late or never.

Community-run against vendor-run moves only who edits, a volunteer with a laptop or a contracted editor, and nothing else on these menus.

Workflow

1. Take delivery rather than re-doing it. Confirm what event-production published and what state the files are in. If the master isn't published yet, stop and wait.
2. Read the consent record and mark each session usable, restricted, or off-limits before choosing any format: references/rights-and-licence-gate.md.
3. Pick the derivative mix (Menu 2), driven by Q7's answer, not the running order: references/derivative-catalog-and-release-plan.md.
4. Pick the publishing latency (Menu 1), pinned to real dates - Q5's deadline and Q6's owner decide this.
5. Run the rights gate (Menu 4) on each planned artifact before production, not before publication.
6. Decide and state the licence. Write it on the artifact, not in a spreadsheet.
7. Build the speaker ask (Menu 3), reusing the shareable kit event-social-media already owns.
8. Hand each scheduled piece to the channel that publishes it, with date and licence line, and stop.
9. Hand the inventory list up to event-marketing-plan - what exists, from what date, in what format.
10. Record what shipped against what was planned, and where it stalled - the only reliable read on whether the plan was sized correctly.

Present the plan for validation, section by section, before any editing starts:

1. The usable-session list.
2. The format mix.
3. The release dates.
4. The rights findings.
5. The licence and the speaker ask.

Everything is cheap to change up to the moment somebody opens an editor.

If your harness has persistent memory, record per edition:

• Which sessions the consent record covered and which it did not.
• What was made from each.
• What shipped on its planned date and what slipped.
• Which speakers amplified.
• The licence chosen, and who decided it.

The next edition's plan starts from a real record rather than from these defaults.

Menu 1 - Publishing latency

When the derivatives drop. This is the menu where the axes genuinely disagree, so it carries two value axes rather than one blended rank.

• effort (production against a repeating date, coordination, a standing owner, how long the job stays open): drip > evergreen re-surfacing > batched release > all-at-once dump
• value - sustained presence across the gap to the next edition, which is what samber/dev-event-organizer-skills@event-marketing-plan can schedule against: drip > evergreen re-surfacing > batched release > all-at-once dump
• value - completion, meaning every artifact you planned actually ships: all-at-once dump > batched release > drip > evergreen re-surfacing
• efficiency: batched release > all-at-once dump > evergreen re-surfacing > drip

Read the completion axis before the efficiency line. Most plans omit it, and latency ranked on reach alone picks the rung most likely to die halfway. Three reasons it is not a hedge:

• Post-event energy decays, so the failure named above is not a badly timed release but one that never happens.
• A dump has already finished by the time a drip is on its second piece.
• A drip's later half is exactly what an exhausted volunteer team drops.

Dominance audit: vacuous - no strict dominance among the 4 rungs, so the efficiency line rests on the argument above, not on the audit. Full audit: references/decision-rationale.md.

• All-at-once dump - everything drops in one push soon after the event. One decision, one afternoon; reaches only people already paying attention.

• Batched release - split into a few waves, write-up first. Costs only scheduling on top of a dump; later waves reach people who missed the first.

• Drip - one piece per fixed interval over weeks. Needs a piece ready on a date, repeatedly, and an owner until a stated end date.

• Evergreen re-surfacing - re-posting an existing artifact months later against a newly relevant moment. No new production, but the watch has no end date and almost nobody keeps it.

• Default rung: batched release.

• Promotion condition: move to a drip when Q4 says the edition recurs and Q6 names an owner with a standing slot through to a stated end date - both, not either.

• Demotion condition: fall back to the dump when Q6 returns no owner past the week after the event, or when Q5's deadline lands before a drip could finish.

The starved options are the drip and, behind it, evergreen re-surfacing - both top the sustained-presence axis and both cost real standing attention, so efficiency never picks either.

• Promote the drip for the compounding mandate above.
• Promote evergreen re-surfacing on its own separate condition: a talk whose subject becomes newly relevant, where the marginal cost is one scheduled post against an artifact that already exists.

Delete, do not demote: publishing with no dated plan at all - "we will get to it". Delete it from this menu and from the axis lines above.

It reads as the cheapest rung and it is not a rung. It is the failure named above, where the crew disperses and the files are never published. A dump with a date is the floor.

Menu 2 - Which derivative to make

Five artifacts, chosen individually rather than as a bundle, because a team rarely has capacity for all of them and the question is which one first.

• effort (production hours, editing craft, repetition per talk, review): transcript-derived text page > short clips > audio cut > written write-up > social thread
• value - new reach, meaning people who did not attend and were not already following: short clips > written write-up > transcript-derived text page > audio cut == social thread
• value - durability, meaning how long it keeps returning value and whether it survives as inventory for the next edition: transcript-derived text page > written write-up > short clips > audio cut > social thread
• efficiency: written write-up > short clips > social thread > transcript-derived text page > audio cut

The == on new reach is argued rather than a dodge: neither an audio cut nor a thread travels on its own - each reaches only the fraction of an audience the event already has who passes it on. They differ in shelf life, which is why the durability axis separates them and this one does not.

Dominance audit: one strict relation among the 5 rungs - the written write-up dominates the audio cut on both value axes and effort, and ranks above it. The other nine pairs are trades. Full audit: references/decision-rationale.md.

• Written write-up - one piece covering what the sessions actually said, audience-facing rather than organizer-facing. Reusable as the link that answers "what is this event", which is why it leads on efficiency despite leading on neither value axis outright.

• Short clips - a cut from the master carrying one moment worth watching without deciding to. Tops new reach because a clip travels through other people's feeds. Finding the moment is the work, not the cut.

• Social thread from the archive - photos, quotes and links assembled from what already exists. Near-zero effort, near-zero durability, reaches only people who were already there.

• Transcript-derived text page - the talk as searchable, quotable, translatable text. Tops durability because search keeps finding it long after a feed has moved on. Tops effort too: a correction pass per talk across a programme is a standing job.

• Audio cut - the talk as audio. Dominated by the write-up and last on efficiency: a one-off audio file with no property behind it has nowhere to land. Stops being dominated once Q8 says a standing audio property already exists - building one is samber/dev-event-organizer-skills@tech-podcast-youtube-channel's job.

• Default rung: the written write-up, plus short clips for the sessions Q7 named.

• Promotion condition: add the transcript-derived page when Q4 says the edition recurs and Q9 says search is a real route in - one without the other does not repay a per-talk correction pass. Add the audio cut only on Q8's yes.

The starved option is the transcript-derived text page - tops durability, tops effort, so efficiency never picks it.

• Promotion conditions, keyed to Q11: a multilingual audience, where text is the only derivative that translates at all; or an accessibility commitment made upstream that already implies it.

Delete, do not demote: re-uploading the full session as a "derivative". Delete it from this menu and from the axis lines above. Re-posting the whole talk to a second destination is not a derivative; it is the master again, and samber/dev-event-organizer-skills@event-production owns that file and its publication. A second copy under different terms is how a consent record quietly stops matching what is published.

Menu 3 - The speaker amplification ask

Scoped narrowly on purpose:

• Already owned elsewhere. samber/dev-event-organizer-skills@event-social-media runs its own amplification-ask menu, covering speakers, sponsors and partners for the campaign before the event, and it owns the shareable kit.

• Covered here. Only the post-event ask about the derivative made from that speaker's own talk.

• Reuse, never rebuild. Take that kit as it stands and never restate its contents here. The double ownership is on the mechanism, not on the moment.

• effort (production per speaker, other people's calendars, coordination): `dated joint push > ready-made post-event asset > one-line ask on the exist

...