The Ghostwriter's System for Capturing Client Voice from Interviews
How professional ghostwriters extract authentic voice from client calls. The transcript-based approach that makes writing sound like them, not you.
The Voice Problem
Your job as a ghostwriter is to sound like your client. Not like you. Not like "professional copy." Like them.
This is harder than it sounds.
- •Notes from discovery calls
- •Client-provided materials
- •Their interpretation of client voice
The result is writing that sounds generically professional. It has the client's ideas but not their personality.
Why Transcripts Change Everything
One approach that separates good ghostwriters from great ones: regularly scheduled calls where you interview clients to uncover stories, ideas, and hot takes. The calls are recorded and processed using transcription tools.
The key insight: starting from a transcript means starting in the client's real speaking voice.
If the client uses a particular metaphor or defaults to a certain word to describe something, it's visible in the transcript. This ensures everything written has a foundation in the client's natural speaking voice.
What Summaries Lose
AI summaries compress a 60-minute call into 500 words. In that compression, you lose:
Verbal tics and patterns: "You know what I mean?" "Here's the thing..." "It's basically like..."
These are voice. These are personality.
Emotional language: Summary: "Client expressed frustration with competitors." What they actually said: "Those clowns couldn't ship a product if their lives depended on it."
Which one sounds like a real person?
Tangents that reveal character: Summaries cut the "off-topic" moments. But those tangents often contain the stories, analogies, and opinions that make content distinctive.
Building Voice Profiles
Experienced ghostwriters create voice profiles documenting:
- •Preferred phrases and word choices
- •Stylistic nuances
- •Topics they get animated about
- •Language they avoid
- •Sentence rhythm and length
- •Level of formality
This profile comes from one source: listening to how they actually talk. Not reading what they've written (which may have been edited or ghostwritten). Hearing their authentic voice.
The Interview Mining Process
Step 1: Ask Better Questions
Don't ask "What do you think about X?" Ask "Tell me about a time when X went wrong."
Stories reveal voice. Opinions reveal ideas. You need both.
Step 2: Let Them Ramble
The polished, edited version of their thoughts isn't useful. You want the raw, unfiltered version.
Resist the urge to interrupt or redirect. The best material often comes after they think they've finished answering.
Step 3: Capture Everything
"If you have any recorded interviews, that's incredibly helpful. In the age of videos and podcasts, this is increasingly common. Listening to these recordings helps study the client's voice and verbal tics."
Every call. Every rambling answer. Every tangent. This is your raw material.
Step 4: Mine the Transcripts
- •Phrases that sound distinctly them
- •Metaphors and analogies they use
- •Stories they tell well
- •Strong opinions with memorable phrasing
This becomes your voice database.
Voice as Collaborative Creation
Some experienced ghostwriters argue that the premium placed on "capturing" voice is overblown.
Part of the job is to "create" that voice in collaboration with the client. Unless working with someone super well known, clients probably don't have an easily recognizable voice, so it's part of the process to figure out not just what their voice is, but what it should be.
This is true. But you can only elevate voice if you first understand the raw material. Transcripts give you that foundation.
The Workflow
For LinkedIn ghostwriting and executive content:
- •Multiple interview calls (recorded)
- •Review all transcripts
- •Build initial voice profile
- •Identify top stories and hot takes
- •Weekly or bi-weekly calls (recorded)
- •Mine transcripts for new material
- •Refine voice profile as patterns emerge
- •Extract quotes to use directly
The workflow involves voice capture and topic mining from interviews, Slack conversations, and meeting notes. Everything the client produces becomes source material.
Summaries vs. Transcripts for Voice Work
| Element | AI Summary | Full Transcript |
|---|---|---|
| Key points | ✓ Good | ✓ Complete |
| Exact phrasing | ✗ Lost | ✓ Preserved |
| Verbal patterns | ✗ Lost | ✓ Preserved |
| Emotional tone | ✗ Flattened | ✓ Authentic |
| Stories | ✗ Compressed | ✓ Full detail |
| Tangents | ✗ Removed | ✓ Available |
For idea generation, summaries work fine. For voice capture, only transcripts work.
Making It Practical
The minimum viable system: 1. Record every client call 2. Transcribe automatically 3. Store transcripts searchably by client 4. Review before writing anything new
The advanced system: 1. Everything above 2. AI-powered search across all transcripts 3. Voice profile document updated monthly 4. Quote database organized by theme
The question that changes your writing: Before drafting anything, ask: "How would they actually say this?"
Then search their transcripts. Find how they've discussed similar topics. Start from their real words, not your interpretation.
The Competitive Advantage
Ghostwriters who rely on summaries produce competent but generic content.
Ghostwriters who mine full transcripts produce content that makes clients say: "This sounds exactly like me. How did you do that?"
The second type commands higher rates and longer relationships.
The difference is one choice: what you record and how you store it.
Full transcripts. Organized by client. Searchable forever.
That's the voice capture system that actually works.
Eddie
Founder, Magnative
Never forget what a client told you
Magnative auto-records every call and files transcripts to your Google Drive client folders. So your AI assistant actually knows your client history.
Related Articles
Why Your AI Assistant Has Amnesia (And How to Fix It)
Every Claude conversation starts from zero. Here's how to give your AI a memory that actually knows your clients.
I Recorded 500 Client Calls. Here's What I Learned.
After two years of capturing every client conversation, patterns emerge. Some are obvious. Some will surprise you.