MMichael Z Consulting
all cases
case 05 · Marketing

BlogFanout

n8n workflow: new blog post → channel-adapted posts across LinkedIn, Twitter, Instagram, Facebook, with a review queue.

n8nClaude APIRSSSocial platform APIs
[demo video coming soon]
/ the problem

A content team publishes a blog post and then spends 60-90 minutes rewriting it for four social channels — different tones, different lengths, different image sizes. The cross-posting is mechanical but the channel-voice adaptation isn't, and copy-paste accidents (LinkedIn tone on Twitter) are common.

/ the approach

n8n workflow triggered by RSS or webhook from the CMS. Claude rewrites the post per channel — LinkedIn long-form, Twitter thread, Instagram caption, Facebook short — anchored by a brand-voice system prompt + few-shot exemplars per channel. Hero image is reused with per-channel resize where needed. Variants land in a review queue; approved items publish, others sit until a human edits or drops them.

/ the outcome

Per-post adaptation drops from ~75 min to ~10 min of review. Brand voice stays consistent because the few-shot exemplars are checked into the workflow, not in someone's head.

/ engineering notes

Brand voice anchored in exemplars, not adjectives

Telling the model 'be conversational but professional' is fragile. Showing it five real prior posts that hit the voice is robust — and when the voice drifts, you update exemplars rather than wordsmithing the prompt.

Review queue, not auto-publish, for low-volume content

At this team's volume (~2 posts/week), the marginal cost of human review is small and the cost of a bad auto-post is large. The workflow optimizes for catch-rate, not throughput.

Per-channel publish modules behind a uniform interface

Each social platform is a swappable n8n node. Adding TikTok or Threads is a new node + a new prompt — no changes to the orchestration.

Got a workflow that looks like this?

Run the free 5-minute diagnostic — I'll tell you whether it's a $0 weekend project or a 2-week build.