Direct answer: start with one approved source asset and ask five questions in order: is it still permitted and accurate, does it still fit the creator's current promise, has the intended audience already received the same experience, can a meaningful new use be made with bounded effort, and is there evidence for that use? The branches end in reuse, recut, refresh, hold or retire. Record the path so a decision can be reviewed instead of guessed.
This decision tree chooses the next use for one asset. It does not design an end-to-end repurposing workflow, choose every distribution channel, organise storage, schedule posts or approve automatic reposting. Before deciding, the asset must already have passed the creator's current approval process.
Use the content-library coverage calculator to find structural gaps across a collection. This page begins later, when one source asset has been selected and needs a defensible next action.
Define the decision unit before judging it
Create an Asset ID and record the source version, creation date, approval status, people represented, intended experience, current locations and last meaningful use. One row should represent one approved source item or set. Do not mix a full shoot, a finished clip and six derivatives in the same decision.
Write the proposed next use in one sentence: audience, surface, format and purpose. “Use this again” is too vague. “Recut approved source V3 into a short orientation clip for new subscribers” gives the reviewer something testable.
Freeze the review date and evidence window. Later activity must go in a new record. This prevents a weak decision from being rescued with information that was unavailable when it was made.
Gate one: confirm the source is still usable
Check the current approval record, participant scope, promised boundaries, factual accuracy, technical quality and any creator-imposed restrictions. If the required evidence is missing or unresolved, the branch is Hold. Do not assume an old approval automatically covers a new context.
If the asset is no longer accurate or no longer represents the creator's current promise, decide whether a bounded refresh can correct it. Refresh means updating a limited element while preserving the source experience. If correction would require rebuilding the piece, treat that as new production rather than repurposing.
OnlyFans publishes current platform terms at OnlyFans Terms. Platform rules can change, so verify the live source at review time instead of copying a permanent rule summary into the worksheet.
Gate two: test promise and audience fit
Compare the asset with the creator's current content promise. Record fit as strong, partial or absent, with one sentence of evidence. A popular old asset can still be a poor fit after a deliberate change in voice, theme or subscriber expectation.
Name the intended audience state: prospect, new subscriber, active subscriber, returning subscriber or another creator-defined group. Then write what that person should understand or experience after seeing the version. The same source can serve different groups only when the adaptation is genuinely useful to each.
If fit is absent, Retire is usually clearer than forcing a derivative. Retirement means no new use under the current promise; it does not require deleting the historical record.
Gate three: detect experience duplication
Check whether the target audience has already received substantially the same reveal, sequence, message and payoff. File differences do not prove experience differences. A crop, new filename or altered thumbnail can still deliver an obvious duplicate.
Record the last comparable use, audience segment, delivery window and meaningful difference. If the audience overlap is high and the new value is weak, choose Hold or Retire. If the audience is genuinely new or the purpose changes clearly, continue to the effort gate.
Use a conservative Unknown status when audience history cannot be verified. Unknown is not permission to treat everyone as unseen.
Gate four: choose reuse, recut or refresh
Reuse keeps the approved experience and format substantially intact for a valid new audience or context. Recut changes structure, duration or sequence to create a different experience from the same source. Refresh updates a bounded factual, visual or contextual element. The choice describes the work, not the expected outcome.
Estimate creator minutes for selection, editing, review, captioning and publishing handoff. Compare that estimate with a creator-selected effort ceiling. If the derivative would consume more capacity than a new item and has no stronger evidence, retire or hold it rather than invoking “repurposing” as a reason to continue.
Quality still needs a separate review. Send any proposed derivative through the content quality checklist before it becomes publishable.
Gate five: require a testable reason
Write the evidence behind the proposed use: a documented library gap, repeated fan question, incomplete audience journey, prior direct response or a planned variation test. “It has not been posted lately” is a date observation, not evidence of usefulness.
State one observation that would support the decision and one that would challenge it. Keep the measurement proportional. A small recut may need a simple completion or response note; it does not need an elaborate attribution model.
When evidence is directional, label it directional. Do not turn engagement into a revenue claim or treat one strong response as a universal preference.
Copy the five-gate decision card
Artifact: copy this header row into a sheet: Decision ID | Asset ID | Source version | Review date | Proposed audience | Proposed surface | Proposed format | Proposed purpose | Current approval evidence | Accuracy status | Promise fit | Last comparable use | Audience overlap | Meaningful difference | Estimated minutes | Effort ceiling | Evidence for use | Counter-evidence | Decision | Reviewer | Next review.
Apply these branch rules in sequence: unresolved approval or accuracy means Hold; absent promise fit means Retire; high audience overlap with no meaningful difference means Hold or Retire; valid fit with unchanged experience means Reuse; valid fit requiring structural change means Recut; correctable outdated detail means Refresh. Record why the chosen leaf beat the nearest alternative.
A fictional example might read: approved source C17, new-subscriber orientation purpose, partial overlap, shortened sequence needed, 25-minute ceiling, repeated navigation question as evidence, decision Recut. The example demonstrates the fields, not a recommendation for duration or audience.
Hand off without silently widening scope
The decision record ends when it names the next action, owner, review requirement and deadline. It does not authorise a channel, publish date or automated rule. Add an accepted derivative to the content calendar planning guide only after quality review.
Use a new Decision ID if the audience, format, purpose or source version changes. Never overwrite Retire with Reuse after a later insight; link a new record so the history shows what changed.
Sample several decisions monthly. Look for excessive Hold states, repeated Unknown audience histories, effort estimates that consistently miss, or one branch chosen without evidence. Those patterns point to a weak input system rather than a need to change the tree's outcome labels.
Audit a sample of decision paths
Select several recent cards across every outcome leaf. For each one, ask a reviewer who did not make the original choice to retrace the gates from the attached evidence. Agreement matters more than whether the reviewer personally likes the asset. If two reviewers repeatedly split at the same question, clarify that gate's definition and examples.
Compare estimated effort with actual effort for completed Recut and Refresh actions. Keep the original estimate, actual minutes, difference and reason. Repeated underestimation can make a reasonable-looking branch unsustainable. Adjust future effort ceilings prospectively instead of rewriting old decisions.
Track how many cards end in Hold because audience history, approval evidence or source versions are missing. A large Hold share is not evidence that the tree is too strict. It may identify an upstream inventory problem that should be fixed once rather than guessed around on every asset.
Handle mixed evidence without forcing consensus
Some assets will have strong promise fit but weak audience-history evidence, or clear audience eligibility but uncertain creator effort. Keep each gate's result visible. Do not average the uncertainty into a single score because a high score can conceal one blocking condition.
When evidence conflicts, write both observations and identify which gate controls the decision. For example, direct subscriber questions can support usefulness while a missing approval version still requires Hold. The evidence is not discarded; it waits for the blocking record to be resolved.
Use a review note to record dissent. A short note such as “secondary reviewer preferred Retire because the difference is only visual” gives the next reviewer useful context and prevents the same disagreement from being rediscovered.
Measure the tree as an operating tool
Useful process measures include cards completed, decision lead time, percentage with complete evidence, estimated-versus-actual effort and number returned at quality review. These measures describe the workflow. They do not prove that repurposed content caused subscriber or revenue outcomes.
Review the process after a fixed creator-selected number of decisions, not after only the most successful derivative. Keep changes small: clarify one gate, add one missing evidence field or improve one handoff. A stable tree produces more comparable learning than frequent rewrites.
Limitations and source notes
Limitations: the tree depends on current approvals, usable audience history and honest effort estimates. It does not predict demand, sales, retention or platform distribution. Two creators can reach different valid outcomes because their promises, evidence and capacity differ.
The evidence-first mapping method adapts the UK Government Service Manual guidance to map a user's whole problem. The five gates, copyable card and outcome leaves are original SirenCY editorial tools. Verify both external sources when using the worksheet because platform and service guidance can change.