36 ChatGPT prompts for small business owners to build a 90-day content calendar

I built my first content calendar in 2011 on a white board in a copy shop back room. It had eight squares, three colors of marker, and zero chance of surviving past Tuesday. The problem wasn’t motivation. The problem was that I had no system for turning “we should post more” into a specific Monday-morning checklist.

A 90-day content calendar fixes that. And in 2026, you can draft the whole thing in ChatGPT in an afternoon if you ask the right way. That’s what this guide is - 36 ready-to-use ChatGPT prompts for 90-day content calendar planning, designed specifically for small business owners who don’t have a content team. I use them in my own consulting work, and I’ve tested them with coffee shops, SaaS founders, and a mom-and-pop HVAC crew.

A 90-day content calendar is a planning document that maps every blog post, email, social post, video, and newsletter you’ll publish over the next 13 weeks, anchored to content pillars and a publishing cadence. Think of it as your content operating plan. Not a wishlist.

Here’s the deal, though. ChatGPT won’t build a calendar that works if you prompt it like a generic chatbot. You need prompts built for the actual job: research, structuring, drafting, repurposing, and measuring. That’s what follows. Each prompt includes the full text, an example output, and a pro tip you can use today.

Quick answer (TL;DR): You can build a 90-day content calendar in ChatGPT by running 36 structured prompts across six stages - pillars, keywords, cadence, drafting, repurposing, and measurement. Most small business owners finish in 4–6 focused hours and ship 30–60 pieces in the quarter. The 2026 data backs this up: 89% of small business owners now use AI for content marketing and 80% of marketers use AI for content creation, with 79% reporting higher content quality from AI assistance.

Why most content calendars die by day 12

Every small business owner I know has started a content calendar. Almost all of them abandon it inside two weeks. The reason is almost always the same: the calendar is a content list, not a content operating plan. It says “post 3x a week” but not which topics, which formats, which channel, or which buyer. It has no pillar logic, no keyword anchors, and no measurement loop.

The 2026 numbers are brutal. Only 21% of content marketers in Orbit Media’s 12th annual blogger survey reported “strong results” from their content program, and the gap is widening between the marketers who plan rigorously and the ones who wing it. Meanwhile, 48% of content marketers say scaling production is their #1 challenge and 60% of those using generative AI worry it will harm their brand’s reputation.

A 90-day calendar is a forcing function. It forces you to pick pillars before you pick topics. It forces you to commit to a cadence before you chase shiny formats. It forces you to measure what shipped. None of this is fun. All of it works.

“The marketers seeing strong results are the ones leaning into the ‘harder’ plays: long-form content, visuals, excellent editing, influencer collaborations, original research.” - Ann Handley, MarketingProfs (quoted in Orbit Media’s 12th Annual Blogger Survey, August 2025)

The good news is that in 2026 you can build this entire plan with ChatGPT. You don’t need a $5,000/mo agency. You need 36 well-built prompts, a single quiet morning, and a willingness to edit.

The 4-pillar content operating model

Before I drop the 36 prompts, you need a frame. Otherwise ChatGPT will return generic fluff and you’ll quit by prompt four.

The 4-pillar model is what I teach every small business client. It maps cleanly to ChatGPT, it survives any industry, and it’s the same backbone the big publishers use (just simpler). Each pillar has a job.

1. Pillars & topic universe - 3–5 themes that mirror how your customers actually think. Answer the question they’re already asking.
2. Keywords & SEO anchors - the specific search phrases inside each pillar. This is where Google and answer engines meet you.
3. Cadence & calendar structure - what ships, on what channel, on what day. Cadence is a constraint, not a goal.
4. Drafting, repurposing & distribution - the production engine that turns one idea into eight assets. This is where AI saves you hours.

The 36 prompts below are organized exactly along these four pillars plus a measurement loop. Run them in order. Edit ruthlessly. Ship the calendar in one sitting if you can.

One quick housekeeping note: I write these prompts to work across ChatGPT, Claude, and Gemini, and to handle the messy realities of running content for a small business. I assume you’re a one-person or two-person team with limited time and no agency. If that’s you, keep going.

Section 1: Pillar & topic prompts (Prompts 1–6)

A content pillar is a thematic category your business owns in the minds of your customers - usually 3–5 of them, each with a clear “jobs to be done” behind it. Pillars are how you stop posting random stuff and start posting recognizable stuff.

The prompts in this section take you from “I run a business” to “here are my 3–5 pillars, each with 10 sub-topics.” Run them in order. They build on each other.

Prompt 1 - Business context dump

Purpose: Capture everything ChatGPT needs to know about your business in one place so you can reuse it as context for every later prompt. This is the most important prompt in the entire library. Skip it and every later prompt will sound generic.

The prompt:

You are my content strategist. I'm going to paste raw context about my business
below. Your job is to read it, then return a clean 1-page "Business Context Brief"
with these sections, and nothing else:

1. Business name & one-line offer
2. Ideal customer (1–2 sentences - who they are, what job they hire us to do)
3. Top 3 problems we solve (in the customer's words, not ours)
4. Top 3 objections that kill the sale
5. Channels we already use (Instagram, email, blog, podcast, etc.)
6. Brand voice adjectives (pick 3 - e.g., blunt, nerdy, warm)
7. Voice anti-patterns (pick 3 things we never sound like - e.g., corporate,
   salesy, jargon-heavy)
8. Goals for the next 90 days (1–3 measurable goals)
9. Constraints (team size, budget, time per week, topics to avoid)
10. A "voice sample paragraph" written in our actual tone, ~80 words, that I
    could publish tomorrow as a LinkedIn post

Don't ask follow-up questions. If something is missing, write "[MISSING]" and
move on. Keep the whole brief under 400 words.

[PASTE YOUR RAW NOTES HERE]

Example output (truncated):

Business Context Brief 1. Business: SuperFresh Coffee Roasters - small-batch coffee for home brewers. 2. Ideal customer: Home coffee enthusiasts, 28–55, who own a $300+ setup and want café-quality results without a barista course. 3. Top 3 problems: Bad espresso shots, stale beans, gear overwhelm. 4. Top 3 objections: Too many roasters look the same, prices are higher than grocery store beans, no real “education” beyond the bag. 5. Channels: Instagram, email newsletter, blog. 6. Voice adjectives: nerdy, friendly, opinionated. 7. Anti-patterns: corporate, preachy, “wellness-coded.”

Pro tip: Save this brief as a .txt file called business-context.txt. Paste it into every later prompt as a prefix. You’ll get dramatically better outputs, and you’ll cut your editing time in half.

Prompt 2 - Pillar brainstorm

Purpose: Generate 4–6 content pillars grounded in how your customers think, not how your org chart is drawn. Most small businesses default to product-feature pillars. This prompt forces problem-and-emotion pillars.

The prompt:

Here is my business context brief:
[PASTE BUSINESS CONTEXT BRIEF]

Based on this, generate 5 content pillars for the next 90 days. For each pillar,
return exactly:

- Pillar name (3–5 words)
- The "customer problem" the pillar solves (1 sentence)
- 3 sub-topics that could each become a blog post
- 1 contrarian take the brand could own (something most competitors won't say)
- An example search query the ideal customer would type that fits this pillar

Pillars should be customer-problem-driven, not product-feature-driven. We want
each pillar to map to a job-to-be-done, not a SKU.

Don't include any pillar that's generic enough to be true for any business.
Reject anything that could apply to a competitor with no changes.

Example output (excerpt):

Pillar 1: “Pull the perfect shot at home” Customer problem: How do I stop my home espresso from tasting sour or bitter? Sub-topics: grind size for light roast espresso, why your ratio is wrong, the 30-second shot check. Contrarian take: “Stop weighing your dose. Time your shot first.” Example query: “espresso sour taste even with new grinder”

Pro tip: Pick 4 pillars, not 5. Three to four pillars is the sweet spot for a one-person team. More than that and your calendar fragments. Less than three and you sound like a brochure.

Prompt 3 - Pillar scoring

Purpose: Use ChatGPT to pressure-test your pillars so you don’t waste a quarter on a bad one. I run this prompt after every brainstorm. It kills weak pillars fast.

The prompt:

Here are my 5 draft content pillars:
[PASTE PILLARS]

Score each on a 1–5 scale across these 4 dimensions, then give a final
weighted score (1.0–5.0):

- Customer demand: Is this a real, recurring problem the customer is actively
  trying to solve? (1=rare, 5=daily)
- Competitive whitespace: How crowded is the SERP/social for this topic?
  (1=saturated, 5=open lane)
- Brand fit: Does this pillar match what we actually sell and the brand
  voice? (1=off-brand, 5=on-brand)
- 90-day producibility: Can we make 8–10 strong pieces on this pillar in
  the next 90 days with our current team? (1=no, 5=easily)

Final weighted score = average of the four.

Then, in 3 sentences, recommend which 3 or 4 pillars to keep, and which to
cut or merge. Be specific. If two pillars are 80% the same, say so.

Example output (excerpt):

Pillar 1: 4.75 - keep Pillar 2: 4.50 - keep Pillar 3: 2.25 - cut, merge into Pillar 2 (it’s the same customer problem in different words) Pillar 4: 4.00 - keep Pillar 5: 3.25 - keep

Pro tip: Don’t let a low-scoring pillar die in silence. If a pillar scores low on competitive whitespace, that’s not always a reason to cut it - sometimes the opposite is true (the topic is so crowded that being there matters). Look at the combined score, not any single dimension.

Prompt 4 - Pillar-to-format mapping

Purpose: Decide which content formats each pillar lives in. The trap is treating every pillar as a blog post. That kills distribution.

The prompt:

Here are my final 3–4 content pillars:
[PASTE FINAL PILLARS]

For each pillar, recommend the 3 best content formats for the next 90 days,
given that we're a small team. Use this exact format:

PILLAR NAME:
- Format 1: [e.g., long-form blog post] - why it fits - estimated production time
- Format 2: [e.g., short-form video] - why it fits - estimated production time
- Format 3: [e.g., newsletter issue] - why it fits - estimated production time

Pick formats that are different in *production reality*, not just in label. A
"long-form blog post" and a "short blog post" are not two formats - they're
one format with a slider.

For each pillar, also recommend the ONE format that should be the "anchor
asset" - the deep, substantial piece that the other formats get pulled from.

Example output (excerpt):

Pillar 1: “Pull the perfect shot at home”

• Anchor: 1,800-word guide with embedded video
• Repurposed: 6 Instagram reels pulling moments from the guide
• Repurposed: 1 newsletter issue featuring the contrarian take

Pro tip: Anchor assets are your moat. One strong anchor piece becomes 8–10 derivative pieces. Orbit Media’s 2025 blogger survey found marketers who publish 2,000+ word articles are nearly twice as likely to report strong results (39% vs. 21% benchmark).

Prompt 5 - Topic universe (40 ideas)

Purpose: Generate a 40-topic universe across all your pillars. This is the master list you’ll pull from for 90 days. If you only run 6 prompts total, make this one of them.

The prompt:

Here are my 3–4 final pillars and their formats:
[PASTE PILLARS + FORMATS]

Generate 40 specific content topics distributed across the pillars, with
roughly even weight (give or take 2 per pillar). For each topic, return:

- Pillar tag
- Working title (specific, not generic - e.g., "Why your light roast espresso
  tastes sour" not "Espresso tips")
- The exact Google search query someone would type to find this (for SEO)
- The customer stage it serves (TOFU = "I have a problem", MOFU = "I'm
  comparing options", BOFU = "I'm ready to buy")
- One sentence on what the reader will walk away knowing or doing

Reject any title that could be published by a competitor with no changes. We
want voice, not generic SEO.

Example output (excerpt):

Pillar 1 | “Why your light roast espresso tastes sour (and the 30-second fix)” | “espresso sour taste fix” | TOFU | Reader will know how to diagnose sourness and adjust grind in 30 seconds.

Pro tip: Don’t run this prompt once. Run it three times with the same input and merge the best ideas. ChatGPT will give you 40 different ideas each time. Across three runs, you’ll see patterns - those patterns are your real pillar.

Prompt 6 - Customer question harvest

Purpose: Mine your actual customers’ questions for content. Most small businesses guess at topics when the answers are sitting in their inbox, DM’s, and sales calls. This prompt turns that into pillar fuel.

The prompt:

Here are my 3–4 pillars:
[PASTE PILLARS]

Below I'm pasting a raw dump of customer questions from emails, DMs, sales
calls, and reviews:
[PASTE RAW QUESTIONS - 30 to 100 is ideal]

Do 4 things:

1. Group the questions into 5–8 clusters. Name each cluster.
2. Map each cluster to the pillar it best fits (or flag it as a NEW pillar
   candidate if it doesn't fit any).
3. For each cluster, write 1 specific content idea that would answer the
   top question in that cluster, and rank the cluster by frequency.
4. Surface 3 "hidden" questions - the ones customers ask *in private* that
   almost never appear in public content. These are gold.

Return everything as a clean table, not prose.

Example output (excerpt):

Cluster	Frequency	Pillar	Hidden question?
Grind size for light roast	14	Pillar 1	No
Shipping damage complaints	9	NEW: Logistics	No
”I gave up, what’s the easiest way”	6	Pillar 2	Yes

Pro tip: The “hidden question” column is your real gold. Most competitors are answering the loud, public questions. You can win by answering the quiet ones - the ones customers ask their friend at a dinner party but never type into Google.

Section 2: Keyword & SEO prompts (Prompts 7–12)

Keyword research for a 90-day content calendar is the practice of finding the specific search phrases your customers type, then mapping each one to a pillar, a stage, and a content asset. You don’t need 500 keywords. You need 30–60, mapped cleanly.

These prompts take you from “I have topics” to “each topic is anchored to a real search demand.” The output is a spreadsheet-ready table.

Prompt 7 - Seed keyword extraction

Purpose: Turn your 40 topics into seed keyword candidates fast. Most beginners get stuck because they conflate topics and keywords. This prompt separates them.

The prompt:

Here is my topic universe:
[PASTE 40 TOPICS]

For each topic, generate 3 keyword candidates:

- A "head term" (1–3 words, high volume, high competition) - e.g.,
  "espresso machine"
- A "mid-tail" (3–5 words, moderate volume) - e.g., "best espresso
  machine for home"
- A "long-tail question" (6+ words, low volume, very specific) - e.g.,
  "why does my home espresso taste sour with light roast"

For each keyword, give a confidence estimate (low/med/high) on whether this
is something the customer is actually searching. Avoid keyword-stuffed
jargon. Think about how a real person would ask the question out loud.

Return as a table: topic | head | mid-tail | long-tail question.

Pro tip: You don’t need a paid SEO tool for this step. Free options that work fine for 90-day planning: Google Trends, AnswerThePublic, the “People Also Ask” boxes in Google, and Google’s “Related searches” at the bottom of the SERP.

Prompt 8 - Search intent classification

Purpose: Make sure each keyword maps to the right format. Mismatched intent is the #1 reason small business content doesn’t rank or convert.

The prompt:

Here is my keyword table:
[PASTE KEYWORD TABLE]

For each keyword, classify the search intent as one of:

- Informational: they want to learn something
- Comparison: they're comparing options
- Transactional: they want to buy / sign up / book
- Navigational: they're looking for a specific brand or page

Then, in 5 words or fewer, recommend the best content format for that intent
(blog post, comparison page, product page, video, etc.). If the same keyword
already has a format recommendation from a prior step, flag any conflict.

Finally, sort the table by intent, with transactional at the top.

Example output (excerpt):

Keyword	Intent	Format
”espresso machine for beginners”	Comparison	Comparison guide
”why espresso sour”	Informational	How-to blog post
”buy SuperFresh espresso beans”	Transactional	Product page

Pro tip: Transactional and comparison keywords are where revenue lives. Make sure 30–40% of your 90-day calendar is built around those keywords, not just informational. Most small businesses underweight them.

Prompt 9 - Answer-engine snippet drafting

Purpose: In 2026, 50% of consumers use AI-powered search tools as their primary research method and 80% of searches now end without a click. You need content that gets cited by AI, not just ranked by Google. AEO is the new SEO.

The prompt:

Here is one of my target blog topics:
[PASTE TOPIC + KEYWORD]

Draft the first 100 words of a blog post that would most likely be cited
by an AI answer engine (ChatGPT, Perplexity, Gemini, Google AI Overviews).

Constraints:
- Lead with a 1-sentence direct answer to the implied question.
- Use a definition-style sentence (e.g., "X is Y when Z") in the first
  3 sentences.
- Include 2 specific, citable facts or numbers in the first 100 words.
- Match the brand voice from my context brief.
- End the snippet with a smooth handoff into the longer article.

Then, list 3 "snippet extraction" signals in your draft - the specific
phrases or structures that make it likely an AI would pull it as a direct
answer.

Example output (excerpt):

“Sour espresso is almost always a grind size problem, not a bean problem. When espresso tastes sour, it means the extraction is too short - water passed through the puck too fast, and didn’t dissolve enough of the acids and sugars. The fix is grinding finer, in 2-grind-step increments, until shot time hits 28–32 seconds. Here’s why that works, and the three other variables to check first.”

Pro tip: AI answer engines reward content that has a clear, extractable definition in the first 2–3 sentences. The “X is Y when Z” template is brutal and effective. If your first paragraph can be lifted as a definition, you’re in.

Prompt 10 - FAQ block generator

Purpose: Build a PAA-style FAQ for each pillar page. These FAQs are the single most underrated SEO/AEO asset in 2026. They get cited by AI Overviews disproportionately because they match how people actually ask.

The prompt:

Here is my pillar and a few of its topics:
[PASTE PILLAR + 3–5 TOPICS]

Generate 8 "People Also Ask" questions a real customer would type about
this pillar. Constraints:

- 6 of the 8 must be questions (start with how, why, what, when, can, is).
- 2 of the 8 can be comparison questions ("X vs Y").
- 1 of the 8 should be a contrarian / "myth-busting" question that
  challenges a common assumption.
- Each question should be 6–12 words long, not longer.
- Order them from most-likely-asked to least-likely-asked.

For each question, write a 2-sentence direct answer (no intro fluff,
no "Great question"). The answer should be self-contained - it should
make sense even if the reader sees only the answer, not the article.

Example output (excerpt):

Q: Why does my home espresso taste sour? A: Sour espresso is almost always under-extraction - your grind is too coarse, so water passes through too fast. Adjust grind finer in 2-step increments until shot time hits 28–32 seconds. Q: Is light roast espresso harder to dial in? A: Yes. Light roasts are denser, so they need a finer grind and a longer shot time to extract the same flavor. Most home setups are calibrated for medium-dark, which is why light roast often tastes sour at first.

Pro tip: Lift these Q&As directly into a FAQPage schema block. Google will sometimes show them as rich results, and AI answer engines read schema natively. It’s the cheapest SEO win in 2026.

Prompt 11 - Internal link map

Purpose: Build the internal linking structure across your 90-day content. Internal links are how you tell Google (and AI crawlers) which pages are your pillars. Most small businesses skip this and wonder why nothing ranks.

The prompt:

Here is my 40-topic content universe, organized by pillar:
[PASTE TOPICS]

Build an internal linking map for the next 90 days. Return a table with
these columns:

- Source topic
- Target topic
- Anchor text suggestion (the exact phrase I'd use as the hyperlink)
- Reason for the link (1 sentence: "this gives the reader X")

Rules:
- Every topic should link to at least 1 other topic in the same pillar,
  and 1 topic in a different pillar.
- Pillar pages (the broad ones) should be the hub that everything else
  links up to.
- Don't suggest links that would be weird for a human reader. Internal
  links are for humans first, search engines second.

Then, in 2 paragraphs, name the 3 "hub" pages - the pillar pages that
should exist by day 30, and that everything else should point to.

Pro tip: Internal links decay over time. Set a calendar reminder for the last Friday of every month to add 2–3 new internal links to recently published posts. It’s a 15-minute task that compounds.

Prompt 12 - Competitor content gap

Purpose: Find the topics your competitors have missed - these are your fastest path to ranking. Most small businesses copy what competitors publish. Winners publish what competitors are too lazy to publish.

The prompt:

Here are my 3–4 pillars and my 40-topic universe:
[PASTE PILLARS + TOPICS]

Here are 3 direct competitors in my space:
[PASTE COMPETITOR NAMES OR WEBSITES]

Do 4 things:

1. For each competitor, list 3 topics they have published that overlap with
   my pillars. Just the topic - I don't need links.
2. For each competitor, list 3 topics they have *missed* that I could
   publish first, that fit my pillars. Be specific.
3. Identify 2 "white space" topics - the kind of thing no competitor is
   publishing, that my ideal customer would find genuinely useful.
4. Suggest 1 original research angle I could pursue that would be hard for
   a competitor to copy in under 90 days.

Be specific. "Write better content" is not a recommendation.

Pro tip: If you don’t know your competitors’ blog URLs, paste 3 of your own blog post titles and ask ChatGPT to identify the implicit competitors in those topics. It works surprisingly well as a starting point.

Section 3: Calendar structure & cadence prompts (Prompts 13–18)

Content cadence is the rhythm at which you publish - not just how often, but on what days, in what formats, with what mix. A calendar without cadence is a wishlist. These prompts lock in the rhythm.

A 90-day calendar has 13 weeks. I typically recommend 4–5 publishing days per week for most small businesses, with a 60/30/10 split across pillars. These prompts will get you there.

Prompt 13 - 13-week calendar grid

Purpose: Build the literal 13-week grid with date stamps. This is the most “boring” prompt in the library, but it’s the one that turns the abstract into a specific file you can actually open on Monday.

The prompt:

Today is [INSERT TODAY'S DATE]. Build me a 13-week editorial calendar
grid, starting on [INSERT START DATE, ideally a Monday].

Return a markdown table with these columns:

- Week #
- Date (Monday of that week)
- 1–2 sentence "weekly theme" (a small narrative arc, not just a topic)
- Pillar focus (which pillar this week emphasizes)
- Anchor asset (the long-form piece due that week)
- 3–4 derivative assets (reels, emails, social posts derived from the anchor)
- Owner (write "me" unless we discuss otherwise)

The first 4 weeks should be heavy on "introduction" pillars and on content
that proves the brand knows what it's doing. Weeks 5–9 should mix in
comparison and "X vs Y" pieces. Weeks 10–13 should lean into
testimonials, case studies, and "ready to buy" content.

Don't fill in topic titles yet - that comes next. The first column should
be "topic TBD" or a placeholder.

Pro tip: Don’t try to plan all 13 weeks at perfect detail. Plan weeks 1–4 in full, weeks 5–8 at theme level, and weeks 9–13 as goalposts. Detail compounds in the wrong direction if you try to ship the whole quarter pre-built.

Prompt 14 - Weekly cadence lock

Purpose: Lock the actual publishing rhythm. Most small businesses over-publish and burn out in week 3. This prompt forces a sustainable cadence.

The prompt:

Based on this 13-week grid:
[PASTE GRID]

Recommend a sustainable weekly publishing cadence for a 1-person content
team. Constraints:

- 2–3 long-form blog posts per month (1,500–2,500 words each)
- 1 newsletter per week
- 4–6 social posts per week (mix of formats)
- 1 video per month (can be a repurposed clip)
- 1 podcast or guest appearance per month (optional)

Show me what that looks like as a single weekly template (e.g.:

Mon - long-form blog publishes
Tue - newsletter + 1 social post
Wed - 2 social posts
Thu - 1 social post + work on next blog
Fri - 2 social posts + 1 short video
Sat/Sun - off

Then, in 3 sentences, name the 3 things I should *not* do if I'm a
1-person team trying to keep this up for 90 days.

Example output:

Sustainable weekly cadence: Mon - anchor blog publishes + newsletter Tue - 2 social posts (one a quote pull, one a clip) Wed - 1 social post + work on next piece Thu - 1 social post + community engagement Fri - 1 short video or repurposed clip Sat/Sun - off, batch next week’s content

Pro tip: Put the rest day in writing. Rest days are not optional. HubSpot’s 2026 State of Marketing report found that 60% of marketers using AI worry about brand damage, and the best protection against AI fatigue is a sustainable rhythm.

Prompt 15 - Pillar-to-week distribution

Purpose: Make sure your pillars don’t get lopsided. The 60/30/10 rule is a good starting point: 60% of content serves your main pillar, 30% supports secondary pillars, 10% is experimental. This prompt checks that you’re not accidentally publishing all “espresso tips” for 90 days.

The prompt:

Here is my 40-topic universe and my 13-week grid:
[PASTE BOTH]

Audit the distribution. For each week, list the pillar tags of the planned
content. Then, across all 13 weeks, give me a percentage breakdown of
how much each pillar is represented.

Apply the 60/30/10 rule as a sanity check (60% of content on the strongest
pillar, 30% on secondary pillars, 10% experimental). If I'm outside that
range, recommend 2 specific swaps to fix it.

If my distribution is too lopsided toward one pillar, that's not always
bad - sometimes the brand genuinely is the dominant pillar. Tell me which
case I'm in.

Pro tip: Don’t enforce 60/30/10 rigidly. Some quarters, a single pillar is the right call (e.g., a product launch quarter). Treat the rule as a diagnostic, not a law.

Prompt 16 - Pillar mix checker (5-3-2 rule)

Purpose: Apply the 5-3-2 rule of social content: for every 10 posts, 5 should be curated or reshared, 3 should be original but not promotional, and 2 should be about you/your product. This rule keeps small business content from sliding into “buy my thing” territory.

The prompt:

Here is my 13-week calendar of social posts:
[PASTE CALENDAR]

Audit it against the 5-3-2 rule. For every 10 social posts, there should be
roughly:

- 5 reshares or curated content (industry news, customer UGC, partner content)
- 3 original posts that are NOT about us (tips, opinions, observations)
- 2 posts that are about us (product, offer, behind-the-scenes)

Count my current ratio. If it's off, recommend specific swaps - name the
exact post slot to change and what to put there.

Then, give me 5 example "reshare" templates I can use without sounding
like a content aggregator. Each template should be 1–2 sentences, easy
to fill in.

Example output (excerpt):

Current ratio: 60% about us, 30% original, 10% reshares. Off by a lot. Swap these 3 slots in week 4 with these reshare templates: […]

Pro tip: Reshares aren’t a sign of weakness. Citing smart work in your industry positions you as a curator, not just a producer. The 5-3-2 rule forces a posture that audiences reward.

Prompt 17 - Seasonal & event hook

Purpose: Layer real-world events, seasons, and trending moments onto your calendar. A calendar that ignores seasonality ships generic content. A calendar that bakes it in ships contextually relevant content.

The prompt:

Here is my 13-week calendar and my pillars:
[PASTE CALENDAR + PILLARS]

I'm in the [INSERT INDUSTRY] space. My audience is [INSERT AUDIENCE].
My geographic focus is [INSERT REGION/COUNTRY].

Identify 6–10 seasonal moments, holidays, or industry events that fall
within these 13 weeks. For each, return:

- The date or date range
- Why it matters to my audience (1 sentence)
- 1 specific content idea tied to my pillars
- 1 contrarian take I could own for that moment

Then, suggest 2 "evergreen" content pieces I should publish in week 1 to
offset any seasonal gaps.

Pro tip: Don’t chase every holiday. Pick 2–3 that are genuinely relevant to your customer. Forced holiday content is the easiest way to lose audience trust.

Prompt 18 - Calendar stress test

Purpose: Run your calendar through a “what could kill this” check. Pre-mortems catch the things that will actually make you quit by day 12. The HubSpot 2026 data is clear: 60% of marketers using AI worry about brand damage, and 31% of marketers are unsure how to use AI in their content strategies. Don’t be that marketer.

The prompt:

Here is my final 13-week content calendar:
[PASTE FULL CALENDAR]

Run a "pre-mortem" - assume this calendar failed by week 6. List the
5 most likely reasons it failed. For each reason, recommend a specific
mitigation I can build into the calendar today.

Then, identify 2 weeks where my topic mix is the weakest, and recommend
a specific swap for each.

Finally, in 1 paragraph, name the single thing about this calendar that
I'm most likely to be overconfident about. Be honest.

Pro tip: Do this prompt on a Friday, after you’ve written the calendar. Your brain will be tired. That’s when it catches the things your Monday-morning optimism missed.

Section 4: Drafting & production prompts (Prompts 19–24)

Content drafting is where the calendar hits reality. A calendar that ships no drafts is a fantasy document. These prompts turn your topics into actual posts, emails, and videos, fast.

HubSpot’s 2026 State of Marketing report found 80% of marketers use AI for content creation and 75% use it for media production, but 93% still review AI output before publishing. These prompts are designed to make that review fast, not painful.

Prompt 19 - Long-form blog post draft

Purpose: Draft a 1,500–2,500 word blog post that doesn’t read like AI. The trick is in the prompt: forcing specifics, voice, and structure up front.

The prompt:

Draft a 1,800-word blog post on the following topic, in my brand voice:

Topic: [PASTE TOPIC]
Target keyword: [PASTE KEYWORD]
Pillar: [PASTE PILLAR]
Customer stage: [TOFU / MOFU / BOFU]
Business context: [PASTE BRIEF]

Constraints:
- Open with a 1-sentence direct answer (extractable for AI Overviews).
- Use H2s every 250–350 words. Each H2 should be a question a real person
  would type.
- Include 2 specific, citable stats or examples in the first 500 words.
- Include 1 contrarian take by the 600-word mark.
- End with a single, clear CTA (not "contact us for more info").
- Use the brand voice adjectives and avoid all anti-patterns.
- Total length: 1,750–1,850 words.

Do not use these phrases: "delve", "in today's fast-paced world",
"it's important to note", "moreover", "furthermore", "in conclusion",
"tapestry", "navigate the landscape", "leverage", "unlock the power of".

Mark sections that need a human edit with [HUMAN EDIT] inline.

Example output: (paragraph excerpt)

Why your light roast espresso tastes sour (and the 30-second fix) Sour espresso is almost always a grind size problem, not a bean problem. When your shot pulls in 18 seconds instead of 28, the water passed through the puck too fast and never extracted the sugars - only the acids. The fix is finer grind, in 2-step increments, until you hit 28–32 seconds. Here’s why that works, and the three other variables to check first.

Pro tip: Always run this prompt twice with the same input. Pick the draft with the better first paragraph. The first paragraph is 80% of the work. Everything else is execution.

Prompt 20 - Newsletter issue

Purpose: Turn the anchor blog into a tight 400–600 word email. Newsletters are the highest-ROI content format for most small businesses. 41% of marketers call email their most effective channel and it costs 62% less than outbound.

The prompt:

Draft a 500-word newsletter issue for [INSERT WEEK] in my brand voice.

Anchor blog post this week: [PASTE TITLE + LINK]
Business context: [PASTE BRIEF]

Structure:
- Subject line: under 50 characters, ideally a question or a number
- Preview text: 1 short sentence, complements (not repeats) the subject
- Opening line: 1 short personal or story-driven sentence (max 18 words)
- Body: 3 short sections, each with a 1-sentence bolded takeaway
- CTA: 1 specific action the reader should take, with a reason it matters
  this week

Tone rules:
- Write in first person. Use "I" not "we" unless it's about a team.
- No corporate opener. Never start with "I hope this email finds you well."
- One emoji max in the entire email. Zero is fine.
- End with a P.S. line - P.S. lines get read 70%+ of the time.

Do not echo my blog post. Treat the email as the *curator's recommendation*
of the blog, with one piece of original commentary.

Example output (excerpt):

Subject: The 30-second shot check Preview: You’ll taste the difference this week. I pulled 14 shots this morning. Three were great, two were undrinkable, and the rest taught me something. The one I’m writing about today is the one I almost ignored - a tiny adjustment that fixed a flavor problem I’d been blaming on the beans.

Pro tip: The P.S. is the most-read part of any email. Use it for the one thing you most want the reader to do. Don’t waste it on a generic sign-off.

Prompt 21 - Short-form video script

Purpose: Pull a 60–90 second video script from a blog post. Most small businesses underuse short-form video because scripting feels hard. This prompt removes that friction.

The prompt:

Pull a 75-second short-form video script from the following blog post:

[PASTE BLOG POST]

Constraints:
- Hook in the first 3 seconds (one short, bold sentence - under 8 words).
- Speak in first person, present tense, casual.
- 3 quick points, each under 12 seconds when read aloud.
- End with a single sentence CTA ("Save this", "Comment X", "DM me 'Y'").
- Total word count: 160–190 words (this lands at 70–80 seconds).
- Include a "b-roll note" line under each section, suggesting what to film.
- Don't read the blog post. Pull the *point* of the blog, in your voice.

Voice: [PASTE VOICE ADJECTIVES + ANTI-PATTERNS]

Don't start with a question unless it can be answered in 1 word.

Pro tip: Film all your short-form videos in one batch. Pick a single Tuesday morning every two weeks, set up your phone, and shoot 4–6 at once. Batching is the difference between “I’ll do it later” and shipping.

Prompt 22 - Social post variants

Purpose: Pull 6–8 social post variants from a single anchor piece. Each variant has a different hook so the same content can travel across platforms without sounding duplicated.

The prompt:

Pull 8 social post variants from this blog post:
[PASTE BLOG POST]

Variants must be in these formats, 1 of each:

1. A contrarian hot take (1–2 sentences, polarizing opinion)
2. A numbered list pulled from the post (3–5 items, 1 per line)
3. A "myth vs truth" callout (2 lines: myth → truth)
4. A question to the audience (1 sentence, ends with a real question mark)
5. A customer story (fictional but realistic, 3–4 sentences, end with a
   takeaway)
6. A "what I'm working on this week" personal update (ties back to the post)
7. A pull-quote from the post, formatted as a graphic (give the exact 12–18
   word quote)
8. A "save this for later" tip post (1 actionable tip, 3 sentences)

Voice: [PASTE VOICE ADJECTIVES + ANTI-PATTERNS]

Each post must be under 280 characters for X / Threads. They can be longer
for LinkedIn, but keep them under 600 characters.

Include a 1-line image/visual idea for each.

Pro tip: The 8 variants will get you through 2 weeks of social posting. You’re not creating new content; you’re just reframing the same idea 8 different ways. That’s the engine.

Prompt 23 - Email subject line A/B test

Purpose: Generate 5 subject line variants per newsletter issue. Most small business emails have a 25–30% open rate. With strong subject lines, you can hit 40%+. This prompt is the cheapest lever you have.

The prompt:

For this newsletter issue:
[PASTE NEWSLETTER]

Generate 5 subject line variants. For each, return:

- The subject line (under 50 characters)
- The psychological hook (curiosity, urgency, identity, fear of missing out,
  specificity, etc.)
- A "predicted open rate lift" estimate vs. a generic subject line
  (low/med/high)
- A 1-sentence A/B test plan: who gets A, who gets B, and what you're
  measuring

Constraints:
- 1 of the 5 must be a question.
- 1 of the 5 must contain a number.
- 1 of the 5 must be under 30 characters (curiosity hook).
- None of the 5 may use all-caps or excessive punctuation.
- None of the 5 may use spam-trigger words (free, urgent, act now, etc.).

Don't write clickbait. The subject line has to match what the email is
actually about - open rate isn't worth a trust loss.

Pro tip: A/B test only one variable at a time. Most ESPs let you test subject lines cleanly. Test #1 vs. #2, pick the winner, test that winner against #3 the next send. Don’t test all 5 at once - the data gets muddy.

Prompt 24 - Production batch checklist

Purpose: Build a one-page checklist for content production so you never forget a step. Most small businesses have no QA process, and that shows up in the published work.

The prompt:

Build a "production batch checklist" for a 1-person content team publishing
this volume per week:

- 1 long-form blog post (1,500–2,500 words)
- 1 newsletter issue
- 4–6 social posts
- 1 short-form video

Return the checklist in 4 phases:

1. PRE-WRITE (the day before writing)
2. DRAFT (writing day)
3. EDIT (post-write, before publish)
4. PUBLISH + DISTRIBUTE (publish day + 24 hours after)

For each step, list the actual action in 5 words or fewer. Don't write
paragraphs. Make it printable on one page.

Include these 5 non-negotiables somewhere in the checklist:
- Read the post out loud before publishing
- Check every internal link
- Check every stat has a source link
- Add alt text to every image
- Verify mobile formatting (open it on your phone, not just desktop)

Pro tip: Print this. Put it on the wall above your desk. The printout is the artifact that actually changes behavior. The digital version you saved in Notion won’t.

Section 5: Repurposing & distribution prompts (Prompts 25–30)

Repurposing is the practice of taking one piece of anchor content and reformatting it for new channels - usually 4 to 10 derivative assets per anchor. Marketers who repurpose systematically are far more likely to report strong results - it’s not even close.

These prompts are the engine that turns a 90-day calendar of 12–13 anchor pieces into 60–100+ total assets.

Prompt 25 - Anchor-to-10-assets map

Purpose: Force-multiply one anchor piece into 10 derivative assets. This is the highest-leverage prompt in the library.

The prompt:

Take this blog post:
[PASTE BLOG POST]

Generate 10 derivative content assets from it. For each, return:

- Asset type (reel, carousel, email, podcast clip, story, etc.)
- Working title
- 1-sentence summary
- The "mom" in the original post this asset is pulled from (which section
  or quote)
- The platform it should ship on
- Estimated production time

The 10 must be a mix: 2 short-form videos, 2 social carousels, 2 newsletter
angles, 1 podcast episode topic, 1 community post, 1 long-form repurpose
(e.g., Twitter thread or LinkedIn carousel), and 1 "lead magnet" angle
(e.g., checklist, template, or worksheet that pulls from the post).

Don't repeat the same idea in different formats. Each asset should have a
distinct angle.

Pro tip: Most small businesses stop at 3–4 derivatives because they think each one is a “new” piece. It’s not. A carousel is the same post, scrolled. A short-form video is the same post, narrated. A newsletter angle is the same post, summarized. Stop treating derivatives as new work.

Prompt 26 - Carousel slide deck

Purpose: Build a 6–10 slide carousel from a blog post. Carousels are the highest organic-reach content format on LinkedIn and Instagram in 2026, and most small businesses have no idea how to script them.

The prompt:

Build a 7-slide Instagram / LinkedIn carousel from this blog post:
[PASTE BLOG POST]

For each slide, return:
- Slide # (label as "Cover" for slide 1)
- Headline (5–8 words, max)
- Body text (1 short sentence, max 12 words)
- Visual direction (1 sentence: what the slide should *look* like)
- B-roll / icon suggestions (if applicable)

Constraints:
- Slide 1 (Cover): bold claim or question, max 6 words.
- Slide 2: name the problem the post solves.
- Slide 3–5: 3 key insights, 1 per slide.
- Slide 6: a "what to do" slide (actionable).
- Slide 7 (CTA): one specific next step, not "follow for more."

The carousel should be readable in 45 seconds. If you can't read it in
45 seconds, it's too long.

Don't use exclamation points. Don't use marketing-speak. Write in
[PASTE VOICE].

Pro tip: Carousels in 2026 are not “screenshot your blog post.” They’re a different format with a different rhythm. A 1,500-word post becomes a 7-slide visual narrative. If your carousel looks like a blog screenshot, you’ve missed the format.

Prompt 27 - Twitter / X thread

Purpose: Pull a 6–10 tweet thread from a blog post. Threads still drive serious traffic and AI citations when done well.

The prompt:

Pull a 7-tweet thread from this blog post:
[PASTE BLOG POST]

Constraints:
- Tweet 1 (the hook): under 240 characters, no link, must stand alone. It
  should make the reader want tweet 2.
- Tweets 2–6: 1 idea per tweet, each under 270 characters.
- Tweet 7 (the closer): 1 sentence + the link to the blog post.
- Tweet 1 must be a contrarian / surprising statement, not a summary.
- Tweets 2–6 should each pull a different specific point from the post -
  no overlap.
- Voice: [PASTE VOICE]
- Do not use hashtags in the body of any tweet. One hashtag at the end of
  tweet 7 is fine.

For each tweet, give a 1-line "why this lands" note. Keep it tight.

Pro tip: Threads are a first-draft format. The first tweet does 80% of the work. If tweet 1 doesn’t earn the click, nothing else matters. Spend an extra 10 minutes on tweet 1 alone.

Prompt 28 - LinkedIn post

Purpose: Turn a blog post into a single, well-crafted LinkedIn post. LinkedIn is the highest-trust platform for B2B in 2026, and most small businesses post generic announcements there. This prompt makes you sound like a human.

The prompt:

Write a 220-word LinkedIn post based on this blog post:
[PASTE BLOG POST]

Constraints:
- Open with a 1-line personal hook (story, observation, or "I used to
  think X").
- No external links in the body of the post. Put the link in the first
  comment instead.
- 3 short paragraphs max. Each paragraph 1–3 sentences.
- No hashtag walls. 3 hashtags at the bottom, max.
- End with a single question to the reader. Not "what do you think?" -
  a *specific* question.
- Use line breaks between paragraphs (the LinkedIn algorithm rewards
  them).
- Voice: [PASTE VOICE]
- Don't start with "I'm excited to announce." Don't use "delve." Don't
  use "game-changer."

Then give me a 1-line draft of the first-comment link post that includes
the URL and a 1-sentence reason to click.

Pro tip: LinkedIn rewards first-person narrative posts over link posts. The trick is to put the link in the first comment - the algorithm still tracks it, and you get more reach on the post itself. This is a small but real edge.

Prompt 29 - Podcast episode outline

Purpose: Turn a blog post into a 20–30 minute podcast episode outline. Most small businesses will never start a podcast, but every small business should be a guest on one. This prompt makes that easy.

The prompt:

Build a 25-minute podcast episode outline from this blog post:
[PASTE BLOG POST]

Return:
- Episode title (10 words or fewer, ideally a question)
- Episode subtitle (one-sentence, sets the stakes)
- Cold open (first 30 seconds - a story, a question, or a provocative
  statement)
- 3 main segments (each 6–8 minutes), with a sub-outline of 3–4 talking
  points per segment
- 1 "hot take" segment (2–3 minutes) - the contrarian angle the guest
  takes
- Closing (last 2 minutes) - a single question for the listener

Constraints:
- This is a SOLO episode, not an interview. The host is the only speaker.
- The hot take segment should be the most-quotable 2 minutes - the bit
  people clip and reshare.
- Include 2 specific "anecdote slots" - places in the outline where
  the host tells a personal story to make the point land.
- The episode should have 1 "if you take away one thing" line. Give me
  the exact line, 12–18 words.

Pro tip: You don’t need to launch a podcast. You need to be a guest on 1–2 per quarter. This outline works as a guest pitch - send it to a host and say “I’d love to come on and cover this.” Most small business owners dramatically underestimate how easy guesting is.

Prompt 30 - Distribution sequence

Purpose: Build a 14-day distribution plan for one anchor piece. A blog post that ships and then sits there is a wasted asset. This prompt turns every anchor into a 14-day campaign.

The prompt:

Build a 14-day distribution sequence for this anchor blog post:
[PASTE BLOG POST + URL]

Day-by-day, return:
- Day #
- Channel (e.g., LinkedIn, newsletter, IG, podcast, community)
- The exact asset (reel, carousel, thread, email, etc.)
- The hook or first line
- A 1-line note on *why* this asset is on this day

Constraints:
- Day 1: blog publishes + newsletter announcement.
- Day 2–3: 2 social posts pulling different angles from the post.
- Day 4: 1 short-form video.
- Day 5: community post (Slack, Reddit, Facebook group, etc.).
- Day 6: email to a partner for cross-promo.
- Day 7: 1 more social post, reframed for a different audience segment.
- Day 8: first carousel or thread.
- Day 9–10: re-share the original newsletter to a non-openers segment.
- Day 11: short-form video, different angle.
- Day 12: community post in a different community.
- Day 13: pull a quote and turn it into a graphic.
- Day 14: repurpose the most-clicked social post as a paid boost (optional).

Don't include a step that doesn't have a clear purpose. Every touchpoint
should serve one of: reach, depth, conversion, or community.

Pro tip: The “Day 10: re-share to non-openers” step is the most-skipped and the highest-ROI step. Most ESPs let you resend to non-openers with a new subject line. It’s a 5-minute job that doubles your open rate over the life of the email.

Section 6: Measurement & iteration prompts (Prompts 31–36)

Content measurement is the loop that makes the next 90 days better than this one. Orbit Media’s 2025 data is brutal: only ~32% of marketers who always check analytics report strong results, vs. 13% for those who rarely do. Most small businesses publish and pray. The prompts below turn prayer into a process.

Prompt 31 - KPI tree

Purpose: Pick the 5–7 metrics that actually matter for a 90-day calendar. Most small businesses measure too many things and learn nothing. This prompt forces focus.

The prompt:

For a 1-person content team running the calendar we built, recommend the
5–7 metrics that matter most. For each, return:

- Metric name
- Why it matters (1 sentence)
- The tool to measure it (be specific: Google Analytics 4, the ESP, etc.)
- A "good" 90-day benchmark for a small business in [INSERT INDUSTRY]
- A "great" 90-day benchmark
- The single most common misinterpretation of this metric (so I don't
  fool myself)

Constraints:
- Include at least 1 awareness metric (e.g., impressions, reach)
- Include at least 1 engagement metric (e.g., time on page, scroll depth)
- Include at least 1 conversion metric (e.g., email signups, demos, sales)
- Include at least 1 SEO/AEO metric (e.g., AI Overview citations,
  organic CTR)
- Reject any metric that is a vanity metric (e.g., follower count, raw
  page views) unless you can defend it in 1 sentence.

Pro tip: Vanity metrics kill small business content. Follower count and raw page views feel good and change nothing. The only metrics that matter are: did the right people see it, did they care, and did they act.

Prompt 32 - 90-day retro template

Purpose: Build the 90-day retrospective you’ll run on day 91. If you don’t pre-build the retro, you won’t run it. Pre-build it now, in the same week you build the calendar.

The prompt:

Build me a 90-day content retrospective template I'll run on day 91.
The retro should be 90 minutes, solo, with 4 phases:

1. NUMBERS (20 min) - what the 5–7 KPIs actually were
2. WINS (15 min) - 3 specific things that worked, with evidence
3. MISSES (20 min) - 3 specific things that didn't, with a hypothesis for
   why
4. NEXT QUARTER (35 min) - 3 things to keep doing, 3 to stop, 3 to start

For each phase, give me 4–6 specific questions to ask myself. The questions
should force specifics, not generalities. "What worked?" is a bad question.
"Which pillar drove 60% of our email signups, and why?" is a good question.

End with a "single most important lesson" prompt that forces one
sentence of self-honesty about the quarter.

Pro tip: Schedule the retro on day 91 now. Put it on the calendar before you forget. The retro is where compounding happens - every quarter gets 5–10% better if you actually run it.

Prompt 33 - Win-loss post-mortem

Purpose: For your top 3 performing pieces and bottom 3 performing pieces, run a quick post-mortem. This is the single highest-ROI analytical exercise in content marketing.

The prompt:

Here are the top 3 and bottom 3 pieces from my 90-day calendar, ranked
by [INSERT METRIC]:

Top 3:
[PASTE TITLES + URLS]

Bottom 3:
[PASTE TITLES + URLS]

For each piece, write 3 sentences:

- What I think drove the result (success or failure)
- What the data actually says (not what I want it to say)
- The single specific change I'd make if I rewrote this piece today

Then, in 2 paragraphs, identify the 1 pattern across the top 3 (what they
have in common) and the 1 pattern across the bottom 3 (what they have in
common). Be specific.

End with 3 sentences on what the top-vs-bottom comparison tells me about
my audience that I didn't know 90 days ago.

Pro tip: Don’t trust your own narrative about why something worked. The data is usually more interesting than your story. The marketers who grow fastest are the ones who override their intuition when the data contradicts it.

Prompt 34 - Content refresh trigger

Purpose: Identify which posts from this 90 days are worth refreshing in the next 90 days. Refresh is almost always higher ROI than net-new.

The prompt:

Here is my full 90-day content list with metrics:
[PASTE LIST + METRICS]

For each piece, recommend one of 3 actions:

- REFRESH: rewrite, expand, or reformat. High value, low cost.
- RESCUE: keep the URL, replace the body. The piece has potential but
  isn't delivering.
- LEAVE: don't touch it. The piece is doing its job or not worth the
  effort.

For each REFRESH and RESCUE recommendation, give:
- 1 specific change to make
- 1 new angle to add
- The estimated effort in minutes

Apply the "80/20 of refresh": focus on the 20% of pieces that drive 80%
of the traffic. Refresh those first.

Pro tip: Most small businesses leave their old content to rot. Refreshing 5 high-traffic posts is often worth more than publishing 20 new ones. The old posts have authority. The new posts don’t yet.

Prompt 35 - Pillar health scorecard

Purpose: Score each pillar’s 90-day performance so you can decide what to keep, kill, or expand next quarter.

The prompt:

Here are my 3–4 pillars and their 90-day metrics:
[PASTE PILLARS + METRICS]

For each pillar, give a 1-page scorecard with:

- 90-day traffic (total)
- 90-day email signups or conversions attributed
- 90-day top-performing piece (and why it worked)
- 90-day under-performing piece (and why it underperformed)
- 3 specific changes I'd make next quarter
- A "keep, expand, or kill" recommendation with reasoning

End with a 1-paragraph recommendation on the 1 pillar I should *expand*
next quarter and the 1 I should *kill* or merge.

Pro tip: Kill is the hardest word for a small business owner. But merging two underperforming pillars into one stronger one is almost always a win. Be ruthless. Robert Rose at CMI keeps writing about this in 2026 - the strongest content operations are the ones that know what to stop.

Prompt 36 - Next-quarter input

Purpose: Convert everything you learned in this quarter into a clear input for the next 90-day calendar. This is the prompt that closes the loop.

The prompt:

Here is my 90-day retro, win/loss analysis, pillar scorecards, and
refresh list:
[PASTE ALL]

Based on this, draft the *first prompt* I should run when I start
planning the next 90 days. That first prompt should:

- Reference the 3 things I'm keeping from this quarter
- Reference the 3 things I'm stopping
- Include a specific hypothesis I want to test (1 sentence)
- Have a clear measurable goal for the next 90 days
- Be written in first person, in my voice, ready to paste into ChatGPT

Don't write the whole next-quarter calendar. Just the first prompt. The
first prompt is the seed for everything else.

Pro tip: Most small businesses start the next quarter from scratch. The compounding move is to start the next quarter with the last quarter’s data baked in. This prompt makes sure you do.

Comparison table: prompt categories × 90-day timeline

Here’s how the 36 prompts map to a 90-day calendar across the three months. Use this as a build order. Prompts 1–6 in week 1, prompts 7–18 in week 2, prompts 19–24 in weeks 3–6 (rolling), prompts 25–30 alongside, prompts 31–36 in week 13.

Category	Prompts	Week 1–4 (build)	Week 5–8 (ship)	Week 9–12 (scale)	Output
Pillars & topic universe	1–6	All 6 prompts run, pillars locked	Refresh pillars if new data surfaces	Refresh pillars if new data surfaces	4 pillars, 40 topics, voice brief
Keyword & SEO	7–12	All 6 prompts run, keyword table built	Use keyword table to inform drafts	Audit and refresh underperformers	Keyword table, FAQ blocks, internal link map
Calendar & cadence	13–18	Prompts 13–16 run, 13-week grid built	Prompt 17 layers in seasonality	Prompt 18 stress-tests the calendar	13-week calendar, weekly template
Drafting & production	19–24	All 6 prompts tested on 1 piece	Run rolling on the week’s anchor piece	Continue rolling; refine templates	12–13 anchor pieces, derivative drafts
Repurposing & distribution	25–30	Prompts 25–26 tested on first anchor	Run rolling on every new anchor	Run rolling; expand to guest pitching	60–100+ derivative assets, 14-day distribution per anchor
Measurement & iteration	31–36	Prompts 31–32 set up at calendar launch	Prompt 33 run mid-quarter on early data	Prompts 33–36 run on day 91 retro	KPI tree, retro, scorecards, next-quarter prompt

This is the order. If you can only do prompts 1–6, prompts 13, prompts 19, prompts 25, and prompt 32, you still have a working 90-day calendar. Don’t let the perfect 36-prompt library stop you from running the 6-prompt minimum.

People Also Ask: 90-day content calendar FAQ

These are the questions real small business owners ask when they’re about to commit to a 90-day calendar. Answer-first, no fluff.

How long does it actually take to build a 90-day content calendar with ChatGPT?

For a small business owner running all 36 prompts, budget 4–6 focused hours. The pillar and topic prompts (1–6) take ~90 minutes. The keyword and SEO prompts (7–12) take ~60 minutes. The calendar structure prompts (13–18) take ~45 minutes. Drafting and repurposing (19–30) are rolling through the quarter, not front-loaded. Measurement (31–36) takes ~30 minutes at the end. Most of my clients finish the planning phase in a single afternoon.

How many pieces of content should a small business publish in 90 days?

For a 1-person team, 12–13 anchor pieces (1 per week) plus 60–100 derivative assets is realistic and sustainable. That’s roughly 5–7 published assets per day across the quarter when you factor in reposts and the derivative pipeline. If you can’t commit to a piece per week, do a piece every 2 weeks and double down on derivatives. Consistency beats volume every time.

Is ChatGPT good enough to write the actual content, or do I still need a human?

ChatGPT is good enough to draft 80% of the content. It’s not good enough to publish the content without a human pass. HubSpot’s 2026 data confirms 93% of marketers review AI output before publishing. Treat ChatGPT as a brilliant but sloppy intern: it drafts fast, but you must edit for voice, fact-check the stats, and rewrite the intro. Plan for 30–45 minutes of human editing per anchor piece.

What’s the difference between an SEO content calendar and an AEO content calendar in 2026?

An SEO calendar optimizes for ranking in Google’s traditional blue links. An AEO (Answer Engine Optimization) calendar optimizes for being cited by AI answer engines - ChatGPT, Perplexity, Gemini, Google AI Overviews. AEO rewards content with clear definitions, FAQ blocks, structured data, and quotable stats. The two aren’t opposites - they’re nested. AEO is SEO that also works for AI. In 2026, you need both.

Should I use ChatGPT, Claude, or Gemini for content planning?

They’re all strong for planning; they have different strengths for drafting. ChatGPT is the most flexible and has the largest prompt library community. Claude is stronger for long, structured outputs (multi-page articles, complex outlines). Gemini has the best native integration with Google Search and Trends data. For a 90-day calendar, I’d run prompts 1–18 in ChatGPT, then test prompts 19–24 in Claude, and pull keyword data from Gemini’s search integration. Mix tools; don’t marry one.

What’s the most common reason 90-day content calendars fail by week 3?

Cadence mismatch. The calendar promises a piece a week; reality delivers a piece every 3 weeks. The fix is to set a cadence that you can sustain on your worst week, not your best. If you can ship 1 anchor piece every 2 weeks and 3 social posts a week for 90 days, do that. The exact volume doesn’t matter. The consistency does.

How do I measure if my 90-day content calendar is actually working?

Pick 5–7 KPIs up front (use Prompt 31). The non-negotiables: email signups attributed to content, organic traffic to pillar pages, AI Overview citations, and time-on-page for anchor pieces. Most small businesses get this wrong by measuring the wrong things. Orbit Media’s 2025 blogger survey found that marketers who always check analytics are 2.5× more likely to report strong results than those who rarely do.

Can I really build this calendar without a paid SEO tool?

Yes. The free toolstack that works fine for a 90-day calendar: Google Trends, AnswerThePublic (free tier), the “People Also Ask” boxes in Google, the “Related searches” footer in Google, your own email/CRM data for customer questions, ChatGPT (free or Plus), Canva for visuals, and your existing email service provider. You can add Semrush, Surfer, or MarketMuse later, but they’re not required for the first 90 days.

A 90-day content calendar walkthrough

Here’s what the first 90 days actually look like when you run these prompts. I’m going to narrate it as if you started today.

Week 1: foundation. Run prompts 1–6 in one sitting. You now have a business context brief, 4 content pillars, a 40-topic universe, a pillar scoring sheet, a pillar-to-format map, and a customer question harvest. Save everything in a single Google Doc called Q3 Content Plan. Time spent: ~3 hours.

Week 2: structure. Run prompts 7–18. You now have a keyword table, search intent classification, an AEO snippet draft, a PAA-style FAQ block for each pillar, an internal link map, a competitor gap analysis, a 13-week grid, a weekly cadence template, a pillar distribution check, a 5-3-2 audit, a seasonal hook layer, and a stress test. Time spent: ~3 hours.

Weeks 3–6: ship the first half. Run prompts 19–24 every week for each new anchor piece. By the end of week 6, you have 4 anchor blog posts, 4 newsletter issues, 4 short-form videos, 16–24 social posts, and a 4-week distribution trail for each anchor. Time spent per week: ~6 hours total (1 piece + 1 newsletter + 1 video + socials).

Weeks 7–12: scale. Run prompts 25–30 alongside drafting. By week 12, you have 8–9 anchor pieces, 12 newsletter issues, 8 videos, 50+ social posts, plus 4 guest pitches. Run a mid-quarter win-loss check using prompt 33. Time spent per week: ~6 hours.

Week 13: retro and plan. Run prompts 31–36. You now have a KPI scorecard, a 90-day retro, a win-loss post-mortem, a refresh trigger list, a pillar health scorecard, and a first-prompt input for next quarter. Time spent: ~3 hours.

Total time over 90 days: roughly 90–100 hours. That’s about 7–8 hours a week - a single workday for a small business owner. Not a side hustle. A real commitment.

The result: a quarter of consistent, well-distributed content, mapped to pillars, anchored in real customer questions, and ready to compound into the next quarter. This is what separates businesses that post content from businesses that run a content operation. Same word count, completely different outcome.

Common mistakes to avoid

I’ve watched a lot of small businesses run 90-day calendars. These are the mistakes that show up over and over.

Mistake 1: Treating the calendar as a topic list. A topic list is what you brainstorm on a whiteboard. A calendar has pillars, cadence, formats, and owners baked in. If your calendar could be a single-column bullet list, it’s not a calendar. It’s a wishlist.

Mistake 2: Skipping the business context brief. Prompt 1 is the most-skipped. It’s also the most important. Without it, every later prompt returns generic outputs that don’t match your voice. The 10 minutes you save skipping it costs you 5 hours of editing later.

Mistake 3: Picking too many pillars. Five pillars, four pillars, six pillars. The math doesn’t work. With a 1-person team, 3–4 pillars is the right answer. More than that and your calendar fragments; less than that and you sound like a brochure.

Mistake 4: Over-indexing on production volume. Hootsuite’s 2026 trends report flags that the “fastvertising” trend is pushing brands to publish more, faster. Resist. Volume without cadence is just noise. Pick a cadence you can sustain on your worst week.

Mistake 5: Ignoring AEO. In 2026, 80% of searches end without a click and AI answer engines are eating the rest. If your content isn’t structured to be cited by AI, you’re invisible to a growing slice of your audience. Prompts 9 and 10 fix this.

Mistake 6: Not running the retro. Day 91 retro is where compounding happens. The marketers who grow 10× over 2 years are the ones who actually run the retro and feed it into the next quarter. Skip it once and you’ll keep shipping the same calendar in a loop.

Mistake 7: Trusting AI to do fact-checking. ChatGPT will confidently invent a stat and put a real-looking URL on it. Always verify every number in every draft. HubSpot’s 2026 data makes this clear: the marketers winning in 2026 are the ones combining AI speed with human judgment.

Mistake 8: Forgetting the customer. It’s easy to get lost in the calendar mechanics and forget the calendar is supposed to serve a real human with a real problem. Re-read Prompt 6 (the customer question harvest) every 30 days. It will keep you honest.

Final word

A 90-day content calendar is the single highest-leverage marketing artifact a small business can build. It’s not a deliverable. It’s an operating system for the next quarter. With ChatGPT and the 36 prompts in this guide, you can draft that operating system in a single afternoon. You can ship it for the cost of one quiet week.

The difference between a small business that intends to do content and one that does content is rarely talent or budget. It’s a calendar that survives the first 12 days. Build the calendar. Run the prompts. Edit ruthlessly. Ship. The compound effect shows up in month 4, not month 1. Most small businesses never get there. You will.

If you want a single starting point, run prompts 1–6 this week. Save the output. That’s the seed. Everything else follows. The 90 days start the moment you hit “go.”