Content strategy and production pipeline

Status: shaping. Central working document for this work as of 2026-09-13. Refine the SCQ and job stories, gather missing evidence, then develop an issue tree before selecting solutions. Hypotheses are intentionally not settled.

Purpose

Grow a relevant, engaged audience through high-quality, on-brand content that is consistently published and marketed. Improve the efficiency of that whole pipeline through appropriate human roles, AI, other tools, and connections between tools. Make better use of existing material as well as new production.

DAM, CRM, catalogues, and task tracking are supporting capabilities. Their value is whether they help achieve these outcomes. Comms owns the strategy, pipeline design, and next-work queue; implementation repositories own supporting work once its need is established.

Situation

Life Itself and its publishing initiatives have ongoing podcasts, recordings, writing, and other media, alongside a substantial backlog. Existing strategy articulates audience and brand ideas, but its current operational meaning needs reconciliation rather than automatic adoption.

Recording, editing, design, storage, and publication happen in different places. Rufus reports that podcasts are now recorded and edited in Riverside. With Zoom, material had to be moved out for editing elsewhere. Drive has already been tried as an archive and is okay. Canva is another working location. Finding the material, current edit, outputs, and relevant decisions across those locations remains a concern.

The supplied production diagram describes Ideas → Pre-production → Production → Post-production → Marketing content → Distribution, with quality checks around handoffs. Marketing content has its own production work: hooks/scripts, thumbnails, descriptions, reels and shorts. The diagram marks marketing as the major constraint at the time it was drawn; its cadence notes are not confirmed current targets.

Complication

  • Existing material does not reliably become published and well-marketed work. Rufus reports queues, particularly toward the marketing end. The exact limiting step—editorial choice, editing, packaging, review, coordination, distribution, or something else—is not yet established from actual cases.
  • The backlog contains unrealized value. We want to recover useful pieces and mine hooks for social media, but this must feed a pipeline capable of finishing and promoting them.
  • Strategy needs an operational brief. Which audience, message, next action, and growth outcome should guide the next cycle? Historical strategy and solution-focused issues do not answer this by themselves.
  • Quality and efficiency must improve together. AI or automation could reduce effort, but could also create extra drafts, rework, or a larger queue. Brand consistency, source fidelity, and editorial judgment matter.
  • Information is fragmented across working locations. Asset hunting and missing context may contribute to delay, but a new DAM or catalogue is not yet demonstrated to be the highest-value intervention.

Question

How should we organize our content production and marketing pipeline—and use AI, other tools, and connections between them—to turn new and existing material into high-quality, on-brand published and promoted work that supports our growth goals efficiently?

Job stories, grouped by responsibility

These are concrete draft stories to validate against real work. The examples do not imply that any tool or workflow has been selected.

1. Strategy and editorial direction — decide why and what

JobStory
Give the next cycle a strategic briefWhen planning the next releases, I want to know whom we are trying to reach, what we want to communicate, and the next step we invite them to take, so selection and promotion serve an explicit growth goal.
Choose work worth finishingWhen choosing between another recording and already-recorded interviews or other backlog material, I want to compare relevance, readiness, and remaining effort, so limited capacity goes to the most useful releases.

2. Production — take a selected piece to an approved main output

JobStory
Register material where work happensWhen a podcast is recorded in Riverside or footage arrives in Drive, I want its location and essential context captured with little effort, so someone can continue the work without requiring an archive copy first.
Resume and hand off the right workWhen an editor or reviewer picks up an episode, I want them to find the current edit, source material, prep/cut notes, and next action, so they do not reconstruct the history or edit the wrong version.
Complete and review the main pieceWhen a recording needs editing, I want a clear brief, quality bar, and actionable review, so it becomes an approved release without repeated ambiguous revisions.

3. Marketing and distribution — turn an output into audience engagement

JobStory
Find faithful, compelling hooksWhen promoting an interview or revisiting older material, I want promising excerpts and angles with source context and timestamps where applicable, so we can create strong social content without misrepresenting the piece.
Produce a complete marketing packageWhen preparing a release, I want the agreed title, thumbnail, description, clips/posts, and invitation produced and reviewed, with editable designs such as Canva projects easy to find, so the main piece does not stall awaiting promotion.
Publish and complete promotionWhen outputs are ready, I want clear channel responsibilities, release actions, and confirmed publication links, so uploading the main episode is not mistaken for completing its marketing.
Connect attention to a next stepWhen someone encounters a clip or post, I want a relevant route to the full piece, subscription, or other agreed action, so attention can become deeper engagement.

4. Reuse and continuity — make existing work available again

JobStory
Recover and prioritize old materialWhen reviewing old recordings, edits, and content logs, I want to know what exists, what was published, and what could serve the current brief, so valuable work can re-enter production or promotion.
Find supporting mediaWhen a page, post, or video needs imagery or footage, I want to find and compare suitable existing material without knowing its folder, so we can reuse good assets quickly.
Keep sources and outputs connectedWhen an edit, Canva design, export, clip, or publication is created, I want its relationship to the original and other versions retained, so we can find, adapt, and reuse it later.

5. Pipeline management and learning — improve how the whole system works

JobStory
Identify and relieve the constraintWhen pieces queue up, I want to see where they wait, why, and who can unblock them, so improvements increase completed, marketed releases rather than activity upstream.
Improve from results and effortWhen reviewing a release cycle, I want to compare audience response, human effort, waiting, and rework, so we improve choices and workflow while protecting quality.

AI, integration, retention, and metadata are cross-cutting means of delivering these jobs. Assess them at the particular step they would improve. CRM belongs where ongoing audience relationships require support; DAM belongs where finding, preserving, or reusing material requires support.

Inputs needed to refine the SCQ and job stories

InputWhat it should establishStarting evidence
Current strategic briefPriority audience, promise, intended action, definition of growth, brand/quality expectationsExisting strategy/ documents plus current human judgment; reconcile stale or conflicting material
Actual queue sample5–10 pieces, stage, last meaningful movement, next action, waiting reason, owner if knownExisting issues/backlog and a small number of current/older episodes; do not infer elapsed waiting from an old issue timestamp alone
End-to-end walkthroughWho does what, in which tool, with which inputs/outputs and approvalsSupplied diagram and one real episode; Riverside candidate identified in prior research
Backlog examplesWhere old material lives, how it is found, what is missing, what makes it worth revisitingclog/, Drive, episode workspaces and publication records
Quality and capacity evidenceExamples of acceptable/rejected outputs, review burden, available human capacityCurrent team practice; do not assume old cadences or ownership are current

Gather available evidence first; bundle remaining human questions with recommended options. Record unknowns explicitly. No need to interview everyone or exhaustively audit all content before drafting the issue tree.

Issue-tree work to do next

Build the tree from the refined question and evidence. Initial branches to examine are outcomes/strategy, actual constraint, production/marketing jobs and handoffs, backlog reuse, and enabling capabilities. These are candidate branches, not a finished decomposition or an implementation plan.

The tree should distinguish independently answerable research from human decisions and show which findings determine later choices. Only then propose alternative interventions and bounded pilots. Do not silently promote the earlier tracker/catalogue design to the answer.

Hypothesis status

Deferred pending clearer framing and the issue tree. Preserve these leads for later evaluation:

  • Publication/marketing is the leading constraint hypothesis, based on Rufus's account and diagram.
  • AI-assisted hooks and marketing packages may be useful, but the affected work and quality checks need definition.
  • An episode-centered coordination record may help; its home and whether an internal directory is needed remain open.
  • Raw backup policy may differ from the working-location policy; Riverside retention and export capabilities have not been verified here.

Terminology and prior research

  • Content backlog board: project #38 in existing documentation; general Comms Board: project #36. Live configurations were not audited. Earlier "comms tracker" wording conflated issues and these views.
  • Internal directory: finding published and unpublished pieces and their working assets. A public website episode directory is a separate audience-facing product.
  • Existing GitHub issues are evidence and current operational work, not requirements that dictate the solution. Do not duplicate them while designing the pipeline.

Sources and preserved work:

Working-state convention

Beads epic: comms-l53. Run bd -C /Users/rgrp/src/life-itself/comms ready --label who:ai for the current queue.

BeadOutcomeDependency
comms-l53.1Draft current marketing brief and missing strategic decisionsReady
comms-l53.2Evidence from 5–10 queued pieces and one real handoff walkthroughReady
comms-l53.3Refined SCQ and grouped job stories grounded in both inputs.1 and .2
comms-l53.4Human review of outcomes, constraint evidence, and jobs.3
comms-l53.5Needs-led issue tree with bounded leaf investigation specs.4
comms-l53.6Evidence-backed intervention/pilot options; create missing research prerequisites first.5

The next design work lives in comms Beads, with this document as its context. Existing GitHub issues continue to hold operational comms work. This introduces a scoped design queue; it does not migrate all comms tracking to Beads. Beads in DAM from the earlier solution-first decomposition are superseded, with their history retained.