Authentic Insight Content Flywheel
Turn one authentic conversation into credible content across channels
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 6
- Confidence
- 99%
The flywheel begins by capturing a genuine source of insight rather than prompting AI from an empty context window. That source can be a webinar, podcast, founder interview, resource guide, or a recording of someone thinking aloud. Its transcript provides facts, opinions, examples, vocabulary, and voice. AI then identifies reusable ideas and adapts them into channel-specific assets such as social posts, articles, or emails. Because each derivative points back to the same substantive source, the operation can increase output without fabricating expertise or losing the speaker's identity. Repeating the process with regular interviews creates an accumulating context library that improves future research, drafting, and consistency.
Origin
Barbara Jovanovic developed this approach after discovering that ChatGPT produced much stronger articles when given webinar transcripts, brand guidelines, and real expert insights. Extracted from Marketing Against The Grain.
Core principles
- 01Never ask AI to create important content from a blank page
- 02Capture genuine expertise before choosing an output format
- 03Preserve the speaker's insights, phrasing, and tone
- 04Reuse one strong source across multiple channels
- 05Improve the source material before increasing content volume
How to run it
- 1
Capture authentic expertise
Record a conversation, presentation, interview, or spoken reflection containing real knowledge and opinions.
Pro tip A founder can record a weekly monologue if no interviewer or public appearance is available.
Watch out Do not manufacture a source merely to create content volume.
- 2
Create a usable transcript
Transcribe the recording so AI can search, connect, and reuse the material.
Pro tip Retain distinctive phrasing because it helps preserve the speaker's voice.
Watch out Check important product names, figures, and claims when using local transcription.
- 3
Extract distinct insights
Ask AI to identify substantive, original ideas rather than produce a vague summary.
Pro tip Request the strongest fully developed topics and their timestamps.
Watch out Not every topic mentioned deserves a content asset.
- 4
Choose the right formats
Match each strong insight to a social post, article, email, or larger resource based on its depth.
Pro tip A cluster of related insights may justify a newsletter series or long-form guide.
Watch out Do not force every insight into every channel.
- 5
Draft from the source
Generate each asset using the transcript's facts, examples, tone, and wording as its foundation.
Pro tip Tell AI to consult the entire transcript for related remarks.
Watch out Prevent the model from adding outside information unless research is explicitly requested.
- 6
Review and feed back
Edit for accuracy and quality, then retain approved outputs as additional voice examples.
Pro tip Add work the founder loves to the long-term context library.
Watch out Human judgment remains necessary even when the source is strong.
In the wild
A software founder records one hour each month discussing product decisions, customer problems, lessons, and opinions. The agency transcribes the call, identifies several original ideas, and turns the strongest ones into social posts, an article, and an email while retaining the founder's phrasing.
→ One hour of founder time supplies grounded content for several channels.
A founder without a podcast records a spoken review of the week, including challenges, discoveries, and product progress. AI mines the transcript and develops only the insights that pass editorial review.
→ The founder creates authentic source material without hiring an interviewer.
Common mistakes
Starting from a blank prompt
A generic request to write about a topic lacks the proprietary knowledge and voice required for differentiated content.
Repurposing shallow material
A transcript is only useful when it contains genuine insights, examples, or opinions worth developing.
Optimizing for output count
Producing every possible derivative creates repetitive content and weakens the value of the original insight.
Is it for you?
Best for
Founders, experts, and lean content teams with valuable knowledge but limited writing capacity.
Not ideal for
Teams that lack credible source material or want AI to invent expertise on their behalf.
From the transcript
“We never start with just like prompting AI just by itself to do anything, even a simplest social media post.”
“But we always start with that transcript. So we have someone's actual insights and actual knowledge, and we also have like, you know, the tone…”
“You never start with a blank page.”
From the episode
I Run a 6-Figure Agency with ZERO Employees (AI Only)
Barbara Jovanovic