Fill out the Brand Context screen once, and every draft Versoo Content Ops generates comes back sounding like you wrote it — same angle, same tone, same habits of phrasing. Fill it out badly, and you get the same generic AI voice you were trying to escape, just with your topic slapped on top. The difference isn’t the plugin. It’s what you type into five fields. Here’s how to fill out each one so the output actually holds your voice.
In This Post
Name the audience you argue with, not the audience you tolerate
Vague: “WordPress users.” That’s not an audience; it’s a product category. The AI has nothing to grab onto, so it defaults to neutral, safe, forgettable prose aimed at everyone and no one.
Specific: “WordPress developers who ship client sites and are skeptical of AI hype but curious about practical use cases.” Now the model knows who it’s arguing with. It knows which objections to preempt and which jargon it can use without explanation. Write the audience field like you’re describing one person you’ve actually talked to — a client, a commenter, a coworker who pushed back on something you said.
Give the angle a stance, not a subject
Vague: “AI and content creation.” That’s a topic, not an angle. Every blog on the internet covers that topic. An angle takes a position on it.
Specific: “Native platform integrations beat bolt-on tools because they get access to real third-party plugins that bolt-on tools don’t.” That’s a claim someone could disagree with. It gives the AI a thesis to write toward instead of a subject to summarize. If your angle field could sit on a Wikipedia page, rewrite it. If it sounds like something you’d say in an argument, keep it.
Describe tone with behavior, not adjectives
Vague: “Professional and friendly.” Every brand on earth claims this. It tells the model nothing about sentence length, word choice, or where you’d crack a joke versus where you’d stay dead serious.
Specific: “Direct and matter-of-fact. Short sentences. Technical precision over hedging. Confidence of someone who built the thing, not someone marketing it.” Notice this describes the actions the writing takes — short sentences, no hedging — rather than the vibes the writing gives off. The AI can execute on actions. It can’t execute on vibes.
List topics as boundaries, not brainstorms
The topics field isn’t a wish list of everything you might someday write about. It’s a fence. It tells the AI what’s in scope so it doesn’t wander into generic advice territory when your title gets slightly ambiguous.
Vague: “Technology, WordPress, AI, productivity.” Broad enough to justify almost any draft, which means it won’t shape any draft.
Specific: “WordPress plugin development, AI content workflows, brand voice consistency, native API usage versus third-party wrappers.” Four topics, each one specific enough that the model can tell when a draft has drifted off them.
Let the writing sample do the voice work
The other four fields tell the AI what to say. The writing sample tells it how to say it, and it’s the field most people skip or half-fill. Don’t skip it.
Paste in 300–500 words of your actual writing — a blog post, a README, an email you sent a client. The model picks up on sentence length and rhythm, where you place emphasis, whether you use contractions, how you open paragraphs, when you use a fragment for punch. None of that transfers through the tone field’s adjectives. “Direct” doesn’t tell the AI you write four-word sentences after a long one for effect. Your sample does.
One good sample beats three vague sentences describing your style. If you only fill in one field carefully, make it this one.
Run this checklist before your first draft
Before you hit generate on your first post, check each field against this:
Audience: Could you name a real person who fits this description? Angle: does it state a position someone could argue against? Tone: does it describe sentence-level behavior, not just adjectives? Topics: Are there four or fewer, and would you notice if a draft strayed from them? Writing sample: Is it at least 300 words of your own actual prose, not a description of your prose?
If you answered yes to all five, generate the draft. If you answered no to any of them, fix that field first — the plugin will still produce something, but it’ll sound like AI wearing your name tag instead of you. Google’s own guidance on helpful content makes the same point from the ranking side: content that sounds like a specific person outperforms content that sounds acceptable to everyone.
Your Brand Context AI Setup: Now Generate a Draft That Sounds Like You
With brand context filled out this way, the title-and-intent step becomes almost boring: type a title, write one sentence about what the post covers, and let Versoo Content Ops do the rest. The draft that lands in your WordPress editor won’t need a voice pass. You’ll spend your editing time on facts and structure, not on stripping out robot phrasing — because you already told the AI, once, exactly how you sound.