Your Writer Isn't the Problem. Your Brief Is.

Here's what a typical content brief looks like when a founder delegates for the first time:

"Write a 1,000-word blog post about [topic]. Here's our website for reference. Our audience is early-stage tech founders. Keep it professional."

That brief will produce generic content from a good writer. It will produce worse content from an average one. And it will produce content that sounds nothing like you from an AI tool that otherwise generates perfectly serviceable prose.

None of that is the writer's fault.

The brief is almost always the problem. Not the writer, not the agency, not the AI tool. The brief. And after fifteen years in professional communications, I can tell you that the pattern is the same whether you're briefing a freelancer, an agency, or a language model. The quality of the output is constrained by the quality of the instructions.

What the Brief Is Actually Asking For

When you send "here's our website for reference," you're asking the writer to extract your brand voice from whatever version of your positioning is currently live on your site.

If:

• your site copy was written three redesigns ago by a contractor who referenced your pitch deck...

• The about page still uses language from your Series A narrative...

• The blog has been produced by four different people over two years with no controlling document...

  
  

Your website isn't a voice reference; it's an archive of accumulated drift.

The writer didn't fail to capture your voice. They captured exactly what you gave them: a generic version of it.

I saw this pattern repeat at EPIC Charter Schools during our pandemic growth phase, when we were scaling from 30,000 to over 63,000 students while the team grew from roughly 1,000 to 2,200 employees. Someone on the communications team would brief a new contractor or team member with "write like us, here's some examples." The examples were a mix of early materials when we had a distinct voice, later materials that had softened, and one piece that was outlier quality. The contractor averaged them. The output was indistinguishable from content produced by any comparable organization in Oklahoma.

The problem wasn't the writer. The problem was the brief treated "look at what we've done" as a substitute for "here's what our voice is."

The Three Things a Brief Needs to Preserve Voice

A content brief that actually works has three components that most briefs skip entirely.

A voice anchor

Three to five sentences pulled from your best existing content, the ones that if you read them cold, you'd know were yours. Not bullet points describing your voice ("conversational but professional"). Actual sentences. Real excerpts. The writer can't reverse-engineer your voice from adjectives. They can pattern-match from examples.

If you can't identify three to five sentences from your existing catalog that you'd hold up as examples of your voice at its best, that's diagnostic. Your voice has probably drifted enough that no clear examples exist. That's a different problem to fix first, but the brief test will surface it.

The one claim this piece makes

Not a topic. A claim. Not "this post is about content briefs" but "this post argues that the brief, not the writer, is responsible for generic brand content." One sentence. Falsifiable. If a competitor could make the same claim with equal credibility, it's not specific enough.

A brief built around a topic produces a post that covers the territory. A brief built around a claim produces a post that takes a position, and the distinction is everything. A topic is "how to pitch journalists." A claim is "your pitch is answering a question no journalist was asking." One invites a survey. The other invites a reader.

The guardrails

What you'd never say in this piece. Two or three lines that define the edge. Not just "avoid jargon," that's too vague to be useful. More like: "We don't use the word 'leverage.' We don't cite competitors by name. We don't make ROI promises we can't back with data." Guardrails exist because writers fill silence with defaults, and your defaults and theirs are different.

Some founders add a fourth element: the specific reader

Not "early-stage tech founders," that's an audience description, not a reader. More like: "A VP of Marketing at a Series B company who just hired their first PR agency and is wondering why the coverage isn't compounding." One specific person. The writing changes when you're writing for someone instead of a demographic.

Why This Isn't a Writer Problem

Founders tend to cycle through writers looking for someone who "gets it" when the brief itself doesn't give them anything to get. The reason the brief is structural and not a talent issue is worth being precise about.

A writer who produces generic content from a generic brief is doing their job correctly. They wrote something professional and appropriate for the category. That's what they were asked to produce. The failure isn't craft. The brief contained no mechanism for preserving anything distinctive.

This is the same structural failure that shows up when companies brief PR agencies. The intake produces five key messages -> the messages are accurate but generic -> the agency pitches with them -> coverage comes back with positioning that could belong to any company in the category -> the company is frustrated -> the agency reports successful placements.

Nobody asked the brief to do what it needed to do: preserve the specific, opinionated voice that makes coverage compound instead of just accumulate.

The brief is the infrastructure layer that most companies skip. They invest in the writer, the agency, the AI tool, the content calendar, and then send instructions that gut the distinctiveness before the first word is written.

How to Fix It Without Rebuilding from Scratch

If you're already delegating content and seeing the drift, you don't need to stop everything and rebuild your message framework before the next post goes out.

You can fix the brief incrementally.

Start with the voice anchor. Go through everything published under your name in the last twelve months. Find three sentences you'd point to as distinctly yours, specific claim, your register, your experience grounding it. Copy them verbatim into a note and paste them into every brief going forward.

Then add the claim. For every piece, before you brief it: finish this sentence. "This post argues that ___." If you can't finish it, you don't have a post yet. You have a topic. Develop the claim first.

The guardrails can come last. One session where you list ten things you'd never say in public under your brand. Not because they're wrong, but because they're not yours. This list is more useful than a style guide.

This version of the brief takes ten minutes to build the first time and two minutes to apply afterward. It won't catch everything (a full message framework does more), but it will stop the brief from actively destroying what you've built.

The Brief Is the Bottleneck

When I was building communications infrastructure at EPIC during that compressed growth, the limiting factor wasn't how many writers we had or how fast they could produce. It was the quality of the instructions they received. Good briefs produced content that held together under volume. Bad briefs produced drift that accumulated faster than we could catch it.

The same constraint applies at any scale. You can hire the best freelance writer in your category. You can use the most sophisticated AI content tool on the market. You can build the most ambitious publishing calendar in your space.

If the brief doesn't preserve your voice, the output won't either.

The writer, whether human or AI, is the execution layer. The brief is the standard they execute to. Most founders spend their energy on the execution layer and give no thought to the standard.

The standard is what compounds.