One Recording, Many Outputs

Creating content from scratch for every platform is how people burn out. The alternative is a pipeline: record one substantial piece, then use AI to turn it into everything else — clips, audio, posts, a newsletter, slides.

In the content module you repurposed text. This lesson does it across formats, using the voice, video and presentation tools from this module.

---

1. Pillar to periphery

The pillar is one substantial, original piece: a 20–30 minute talk, interview, workshop recording or podcast episode. It carries your real thinking.

The periphery is everything derived from it:

                     ┌─ 3–5 short vertical clips
                     ├─ audio podcast episode
30-minute recording ─┼─ blog post / article
                     ├─ newsletter section
                     ├─ 10 social posts and a carousel
                     └─ slide deck or lead-magnet PDF

You invest effort once, in the part that requires you. AI handles the transformation. You handle the quality check.

2. The repurpose ratio

Set a target so you know when you're under-using a recording. A good starting ratio for one hour of source material:

1 → 10 → 3 → 5

• 1 hour of pillar content becomes
• 10 social posts,
• 3 long-form pieces (article, newsletter, script), and
• 5 engagement assets (short clips, a carousel, an FAQ, an email, a quote graphic)

Track your actual ratio and time spent each month. If you're only getting one post from an hour of recording, the pipeline is broken, not the content.

3. The pipeline, step by step

1. Record the pillar with clean audio (see the voice lesson). Keep a version without music.
2. Transcribe and clean. Use a text-based editor; fix names and terms in the transcript once — every derived output inherits them.
3. Find the moments. Use a clipping tool (e.g. Opus Clip, Descript, or CapCut's auto-clips) or ask an assistant:

Here is a transcript with timestamps. Find the 5 moments most
likely to work as standalone 30-60 second clips. For each give:
start and end timestamps, why it works on its own, and a hook
line for the caption. Only use what's actually said.

[TRANSCRIPT]

1. Cut the clips vertically, with captions, in your editor.
2. Export audio as a podcast episode; clean with studio-sound, add intro music you have rights to.
3. Generate the text outputs from the transcript:

From this transcript, draft:
1. A 900-word article with subheadings
2. A newsletter section under 300 words
3. 10 social posts, each built on a different point, under
   150 words each
4. A 6-slide carousel outline

Write in my voice (Digital Twin attached). Every claim must come
from the transcript. Mark anything needing a source as [VERIFY].

1. Build a visual asset — feed the article into a presentation tool for a slide deck or lead-magnet PDF.
2. Edit everything. Read each piece as your audience would.
3. Schedule across two to four weeks, not all at once.

4. Consistency across formats

When the same idea appears in six formats, inconsistency is noticeable. Keep a one-page content style guide:

• Voice: your Digital Twin's writing rules.
• Visuals: brand colours, fonts, caption style, thumbnail layout.
• Audio: the same intro sound; the same voice (your real or cloned voice, never a mix that confuses listeners).
• Claims: the approved numbers and wording for anything factual.

Consistency is what makes an audience recognise you before they see your name.

5. Respect platform norms

The same idea needs a different shape for each place:

Platform	Shape
Short-video apps (Reels, Shorts)	Vertical, hook in the first second, captions on
LinkedIn	A specific professional insight, readable without clicking
YouTube	Searchable title, chapters, longer form
Newsletter	Personal, one idea, a clear link
Podcast	Audio that makes sense without visuals ("as you can see" doesn't work)

---

⚠️ Common mistakes

• Stopping at one format. Most of the value of a recording is in its derivatives.
• Inconsistent branding across outputs.
• Ignoring platform norms — a horizontal 10-minute video posted as a short.
• Publishing AI drafts unedited. The pillar was yours; don't let the periphery sound like a stranger.
• Invented details creeping into derived posts. Everything must trace back to the source.

---

What's next: the next module moves from creating content to building systems that work on their own — branching workflows and agents.