Blog Content Prompts

Prompts for planning and drafting blog posts that answer real questions, using AI for structure and your own experience for substance.

6 promptsCopy and pasteWorks with any AI assistant

About these prompts

AI can draft a blog post on almost any topic in seconds, which is exactly why generic AI posts add so little value. Search engines have said they reward helpful, original content however it is produced, and they treat mass-produced pages made mainly to rank as spam. The useful middle ground is to let AI help with ideas, structure and editing while you supply the experience, examples and opinions only you have.

The prompts below are built that way. They ask the assistant to leave clearly marked gaps for your own knowledge rather than filling them with plausible-sounding filler.

Copy any prompt below, or build a custom one with the Prompt Builder. Highlighted [PLACEHOLDERS] are the parts to replace with your own details.

The prompts

01

Topic ideas from customer questions

When you do not know what to write about.

#
bl-ideas.txtChatGPT ↗Claude ↗
I run [BUSINESS NAME], a [TYPE OF BUSINESS] for [IDEAL CUSTOMER].​Suggest 20 blog post ideas that answer real questions my customers have at different stages: before they know they need us, while comparing options, and after buying. For each idea give a working title and the specific question it answers. Avoid generic topics that any business could write about.

TipPick topics where you have real experience to add. Those are the posts worth writing.

02

Detailed outline with gaps for my experience

Before drafting any post.

#
bl-outline.txtChatGPT ↗Claude ↗
Create an outline for a blog post titled "[WORKING TITLE]" for [AUDIENCE].​Include: the main question the post answers, a short introduction, 5 to 8 section headings with bullet points of what to cover, and a conclusion. Mark every place where my own experience, examples, photos or opinions would make the post better than a generic article with [MY EXPERIENCE: what to add]. Flag any facts or statistics I would need to check.
03

Draft from my notes

When you have notes and want a readable draft.

#
bl-draft.txtChatGPT ↗Claude ↗
Write a blog post from my outline and notes below. Keep my examples, opinions and facts exactly as written, and keep my voice: [DESCRIBE YOUR VOICE OR PASTE A SAMPLE OF YOUR WRITING].​[PASTE OUTLINE AND NOTES]​Do not add statistics, studies or quotes that are not in my notes. Where a section is thin, tell me what to add instead of padding it.
04

Edit a draft for clarity

After you have written a draft yourself.

#
bl-edit.txtChatGPT ↗Claude ↗
Edit the blog post below for clarity and readability without changing my meaning, facts or voice.​[PASTE DRAFT]​Shorten long sentences, cut repetition, and suggest clearer headings. Then list any claims that need a source and any sections where a reader might get confused.

TipAsk for a list of changes, not just the edited text, so you can accept or reject each one.

05

Turn one post into other formats

When a post is published and you want more use from it.

#
bl-repurpose.txtChatGPT ↗Claude ↗
Here is a blog post I published: [PASTE OR SUMMARIZE]​Turn it into: an email to my newsletter (under 200 words), 3 social media posts for [PLATFORM], and a short FAQ entry for my website. Keep the facts consistent and link back to the full post where it makes sense.
06

Refresh an old post

When a post is out of date.

#
bl-update.txtChatGPT ↗Claude ↗
This post was written in [YEAR]: [PASTE POST]​Identify anything that may be out of date (prices, rules, tools, statistics, advice that has changed) and list it for me to check. Then suggest new sections that would make the post more useful today. Do not state new facts as certain; flag what I need to verify.

Worked example

What a response looks like

The prompt we ran

Prompt 1 with: Ridgeway Bike Repair, a bike repair shop for commuters and weekend riders in Tacoma.

The response

A few of the 20 ideas returned:

  • "How often should a commuter bike really be serviced?" (Answers: when to book a tune-up based on miles ridden.)
  • "Is it worth repairing an old bike or should I buy new?" (Answers: the repair-or-replace decision.)
  • "What we check in a basic tune-up, and what costs extra" (Answers: what customers are actually paying for.)
  • "Five noises your bike makes and what they mean" (Answers: whether a problem is urgent.)

Sample response, produced by running this prompt in Claude (Anthropic) with the fictional business above. AI responses vary every time you run a prompt; use this to see the kind of result to expect, not as copy to reuse.

Mistakes to avoid

  • Publishing AI drafts with no original experience added.
  • Publishing many similar posts mainly to target keywords.
  • Including statistics the AI made up.
  • Writing for search engines instead of customers.

Check every fact, number and claim in AI output before you publish. See how to fact-check AI content.

Frequently asked questions

Will Google penalize AI-written blog posts?

Google has said it focuses on the quality and helpfulness of content rather than how it was produced. What it treats as spam is content produced at scale mainly to manipulate rankings. A useful, accurate post that you have checked and improved is fine; hundreds of thin auto-generated posts are not.

How often should a small business blog?

Consistency matters more than frequency. One genuinely useful post a month that answers a real customer question is better than weekly posts with nothing new to say.

Updated October 2026Written by the AI Website Prompts editors