Almost every bad ghostwriting experience has the same root cause, and it is not the writer’s talent.
It is a brief that described the format and not the thinking. “Write four posts a week for a SaaS founder about growth” is a complete description of how to produce content that sounds like every other SaaS founder on the platform. The writer does exactly what was asked, and the client concludes ghostwriting does not work.
Here is what to hand over instead.
The five inputs
1. The objective, in commercial terms
Not “build my presence”. Something a writer can make decisions with:
- “I want Series A B2B founders to see me as someone who understands content pipeline pressure, because that is who hires me.”
- “I need recruiters and boards to read my thinking on category strategy before they shortlist anyone.”
- “We are hiring senior engineers and I want candidates to see how the company actually makes decisions.”
Every subsequent judgement (which post to write, which angle to cut, how to close) resolves against that sentence.
2. The audience, made concrete
One specific paragraph, plus two or three real people who fit it. The real names matter more than the paragraph: a writer who can picture a specific person will make far better choices than one optimising for a demographic.
Include what they already believe. If your reader already thinks content is a numbers game, the piece argues against that. If they think it is working fine, the piece has to start somewhere else entirely.
3. The positions, not the topics
Topics are interchangeable. Positions are not.
Topic: “LinkedIn content strategy” Position: “Most B2B content fails because it answers the question buyers have last. Build for the moment they are comparing vendors.”
Give the writer three to five positions you hold, each with the reasoning behind it. This is the single highest-value input in the entire brief, because it is the one they cannot invent without making something up.
4. Raw material, on a schedule
This is the input clients most often skip, and its absence is what forces a writer into generic territory.
Sources worth feeding a ghostwriter:
- A monthly recording, 45 to 60 minutes, unstructured conversation about what you are working on. Transcripts are a goldmine: people say things in speech they would never write.
- Recent decisions, what you chose, what you rejected, why.
- Customer conversations, the exact objection that came up, the phrasing they used.
- Things you changed your mind about. The most credible content available to you, because almost nobody publishes it.
- What you are reading and disagreeing with. Disagreement is the raw material of a point of view.
Without this, the writer has only your job title. With it, they have a decade of material.
5. The voice guide
Two kinds of evidence work here: examples and constraints.
Examples, two or three pieces of writing that genuinely sound like you. Ideally something you wrote yourself: an email you sent, a document, a post you are proud of. Not famous people’s posts you admire; the writer will copy the wrong surface features.
Constraints, be specific, and be honest about your own habits:
- Sentence rhythm: short and clipped, or long and qualified?
- Vocabulary: technical or plain? Do you use contractions?
- Emoji: none, sparingly, or as punctuation?
- Banned phrases: for most executives this includes “thrilled to announce”, “delve”, “game-changer”, ”🚀”, “I’m humbled by”, “let that sink in”
- Claims you will not make: specific numbers, results, client names, predictions
- The formats you refuse: confessionals, listicles with no argument, motivational close
A good voice guide is 400–600 words. Longer than that and nobody uses it; shorter and it does not discriminate between drafts.
What to leave out
Do not brief the hook. If you specify the opening line, you have written the post. Let the writer structure it; your job is the substance.
Do not ask for keyword-stuffed drafts. “Include ‘B2B SaaS growth’ three times” produces text that reads like it was written for a machine, and readers notice.
Do not brief a week at a time. Briefing in monthly batches lets the writer see the arc, which themes have appeared, where repetition is creeping in, and what needs saying next. Week-by-week briefing produces unconnected posts.
Do not ask for a first draft you will not read. Approval is not a formality. If you approve everything without comment, the writer has no signal, and the voice guide never improves.
The approval rule that fixes most problems
When you reject a draft, say which line broke and why.
Not “this doesn’t sound like me”. That is true but unusable, and it sends the writer guessing. Better:
- “Too polished. I would never claim that much certainty.”
- “The example is right but the framing is salesy.”
- “This is true of my clients in fintech but not in logistics, narrow it.”
- “Line 3 is a borrowed take from a podcast. Cut it or add what I actually think.”
Each of those is a specific, learnable correction. Three rounds of that and the drafts start coming back right the first time, which is the entire economics of the arrangement.
A one-page brief template
OBJECTIVE
What commercially changes if this works. One sentence.
AUDIENCE
One specific paragraph + two real people who fit it.
What they already believe.
POSITIONS (3–5)
Each: the claim, the reasoning, the evidence.
RAW MATERIAL THIS MONTH
Recording attached / decisions / objections / reversals.
VOICE
Two examples that sound like me.
Sentence rhythm, vocabulary, banned phrases, formats I refuse.
CLAIMS I WILL NOT MAKE
Specific list.
CADENCE
X posts per week. Days. Approval deadline.
REVIEW RULE
I will flag the specific line and the reason.When to conclude it is not working
After two months of specific feedback, if drafts still read as generic, the problem is upstream of writing: either the positions are not real, or the raw material is not flowing, or the writer is not right for this voice.
The first two are fixable in-house. The third is not: change writers rather than lowering expectations. But diagnose before you switch, because replacing a writer to avoid feeding the process is the most common way good programmes die.
Where to go next
- LinkedIn ghostwriting, the interview-led model this article describes.
- The structure of a LinkedIn post that earns a conversation, what a well-briefed draft should ultimately produce.