Presence10 min read
A content engine in one afternoon, with Claude
Turn one idea into a LinkedIn post, a blog, a newsletter, and a landing page — in a single afternoon. Here's the exact pipeline we use for our own marketing.
By Roki HasanMarch 27, 2026
The problem with content calendars
Most content strategies collapse because they separate "making content" from "having ideas." The calendar says "Monday: blog post, Tuesday: LinkedIn, Wednesday: newsletter." You sit down Monday morning and realise you have nothing to say.
The better model is: have one idea a week. Turn it into six formats. Publish rolling.
Claude makes the "turn it into six formats" part approachable.
The pipeline
Input: one idea, written up as 400–800 words. Or a transcript of you talking about it for 20 minutes.
Outputs:
- A LinkedIn post (250 words)
- A longer blog post (1,200 words)
- A section in the week's newsletter
- A short video script (2 minutes, for a talking-head)
- A landing page section (for a future /learn tutorial)
- A set of 5 short tweets
One afternoon. Six assets. One week of content.
The prompts
We use one chain with Claude. It looks like this.
Prompt 1 — Anchor the idea
Here is my core idea: [paste 400–800 words]. Summarise the idea in one sentence. Identify the 3 most important supporting points. Identify one contrarian claim. Identify the ideal reader: who, what they care about, what they'll do with this.
Output: the idea, tightened.
Prompt 2 — LinkedIn post
Using the summary above, write a LinkedIn post in my voice (examples attached). Target 250 words, 3 short paragraphs, no emojis, no "here's what I learned." Open with the contrarian claim.
Prompt 3 — Blog post
Expand the idea into a 1,200-word blog post. Use the 3 supporting points as sections. Include at least one specific example. Tone: essay, not listicle. Target reader: someone deciding whether the idea applies to their business.
Prompt 4 — Newsletter section
Write a 200-word newsletter section based on the idea. Frame it as: "This week I was thinking about [X]. Here's the short version." End with one call to action (read the blog, reply with thoughts, come to an event).
Prompt 5 — Video script
Write a 2-minute video script for me to film. Tone: talking to one person across a coffee table. No script markers. Just spoken words, broken into 4 short paragraphs.
Prompt 6 — Tweet thread
Write 5 standalone tweets, each 240 characters or less, each covering one facet of the idea. Each should be standalone — no "1/" or "2/" numbering.
What this looks like in practice
We did this for the idea "most agencies should be boutique" last month. One afternoon. Output:
- LinkedIn post — 2,400 views, 12 saves
- Blog post — ranks on Google for "why agencies should be boutique"
- Newsletter section — 42% open rate, 14% click
- Video — 800 views on YouTube in 3 weeks, slow burn
- Tweets — one went modestly viral (11k views)
Total cost: 3 hours of editing. Claude did 4 hours of drafting in about 12 minutes of compute.
Editing discipline
Every output needs a human editor. Non-negotiable. The edits are small — tightening, one swapped sentence, adjusting voice — but they're what makes the content feel yours.
If you skip editing, readers notice. AI-native content has a tell: too-consistent paragraph lengths, no surprising word choices, overly polite transitions. A 10-minute edit per asset removes the tell.
Batching matters
Do this in one sitting. Not Monday LinkedIn, Tuesday blog, Wednesday newsletter. All at once. The idea is fresh, your voice is warmed up, the thinking is loaded.
Then publish over the week. The content feels consistent because it was written consistently.
When to not do this
When you don't have the idea. No content engine survives an ideas drought. If the core idea is weak, six formats of weak content is worse than no content.
The better use of that afternoon: have a real conversation with someone, record it, sit with it. The idea shows up.
This is how we generate most of the content you're reading on this site. Including this post — which started as a 40-minute call between two of our team last Tuesday.
Keep reading
More from the Presence series.
Want us to run this for you?
Thirty minutes, no slides. We'll tell you honestly whether this is the right first thing to fix.