
Generic prompts produce generic writing. These techniques give you control over voice, structure, and quality so AI-assisted writing sounds intentional, not templated.
TLDR
Specify your audience, purpose, tone, and one stylistic constraint. Show a sample of existing writing you want it to match. Get a first draft, then use targeted follow-up prompts to refine specific elements.
Define audience and purpose before the content
Before describing what to write, tell the model who will read it and what you want them to think, feel, or do after reading. "Write for busy marketing directors who need to justify budget" produces completely different writing than "write for junior marketers learning the basics."
Specify tone with examples, not adjectives
"Conversational but authoritative" is vague. Instead say "tone of a knowledgeable friend, not a consultant: direct sentences, no jargon, occasional humor." Or paste two sentences you like and say "match this tone."
Set format and length constraints
Tell the model structure and length: "three paragraphs, no headers," "under 150 words," "open with a provocative question, then evidence, then a clear takeaway." Constraints improve writing quality, not just length.
Get the draft first, then refine
Request the full draft before editing anything. Then refine one element at a time: "the opening is weak, rewrite just the first paragraph," "the third point is unclear, give me three alternative ways to phrase it."
Give feedback on the why, not just the what
Instead of "rewrite this," say "this section sounds too formal for our audience and buries the key point in the third sentence. Revise to fix both." Specific, reasoned feedback produces dramatically better revisions.
Example prompt
LinkedIn milestone post: audience, tone, structure, and what to avoid all specified upfront
Write a 250-word LinkedIn post announcing that our company just crossed $1M in ARR. Audience: founders and operators who follow growth stories. Tone: honest and grounded, proud but not braggy. The story has real challenges we overcame. Structure: open with the number and something counterintuitive we learned, then two or three specific lessons, close with one sentence that is shareable. No cliches about "journey" or "humbling experience." No emoji.
First drafts on tight deadlines
AI is fastest at getting from blank page to rough draft. Use it to generate a full first draft, then edit it yourself. Editing is always faster than writing from scratch.
Matching an established brand voice
Paste three or four examples of your best existing content and ask the model to match the voice. It will extract patterns in sentence length, vocabulary, and tone and apply them to new content.
Generating structural variations
Ask for three different approaches to the same piece: "give me three different ways to open this article and three different structural outlines." Seeing options is faster than committing before you start.
Asking AI to write "in your voice" without examples
AI cannot guess your voice without context. Show it three to five examples of your existing writing, then ask it to match that style.
Accepting the first draft without refining
First drafts are starting points. The real value is in the refinement loop: get a draft, identify what is off, ask for specific changes. This produces better output than asking for a perfect first draft.
Prompting for "good" writing without defining what good means
"Make it better" is not actionable. "Make the opening sentence more specific" or "cut the second paragraph, it repeats point one" are.
Yes, if you give it enough examples. Paste five to ten samples of that person's writing and ask it to match the style in sentence structure, vocabulary, and rhythm. It will not be perfect but it is a strong starting point.
Always edit. AI writing is a first draft, not a final product. The goal is to spend time editing rather than writing from scratch, which is faster and often produces better output.
Specificity kills generic writing. Add concrete details: numbers, proper nouns, real examples. Ask it to "replace any generic example with a specific one from [your industry]." Generic prompts produce generic output.
Bottom line
The more specific your constraints, the better the writing. Define audience, purpose, tone, and format before you start. Use the first draft as a starting point and refine with targeted, specific feedback.
Prompt packages that apply this technique directly.
Copywriting
Most copy fails because it talks about the product instead of the reader's problem, buries the benefit in generic claims, and never gives the reader a clear reason to act now.
See promptsTechnical Blog Writing
Many technical blogs suffer from either excessive jargon or lack of clarity, leaving readers confused or disengaged.
See promptsSEO Content Writing
Many businesses struggle to create content that ranks well on search engines.
See prompts