The 6-step content workflow that turns raw ideas into voice-matched, published content in 90 minutes. No content team. No generic output. Just a system that runs the same way every time.
An AI content workflow has six steps: capture, brief, draft, edit, publish, distribute. With a trained voice prompt, the draft step takes 5 minutes instead of 2 hours. Total workflow time for one blog post or five LinkedIn posts: 90 minutes. The workflow is part of the content batching system we use at Syxo.
You open ChatGPT. You type "write a LinkedIn post about [topic]." You get something that sounds like a corporate press release written by a committee. You close the tab.
That's not a workflow. That's a slot machine with bad odds.
The output quality of AI content depends entirely on the inputs. A generic prompt produces generic content. A structured workflow with a trained voice prompt produces content that sounds like you wrote it. Same tool, different system around it.
Here are the six steps.
Keep a single running list of content ideas. Phone notes app, Notion page, plain text file. The format is irrelevant. What matters is that you capture ideas when they occur, not when you sit down to write.
Sources that produce the best content ideas: client questions (the ones you answer repeatedly), mistakes you made and fixed, opinions you hold that most people in your space disagree with, specific results with numbers attached.
Aim for 20+ ideas in the bank at all times. When the list drops below 10, spend 15 minutes refilling it. But this step has zero time cost during production because the capture happens throughout your week.
Pick one idea from the capture list. Write a three-line brief:
The brief prevents drift. Without it, AI drafts wander. With it, the output stays focused because the input was focused. Ten minutes here saves 30 minutes of editing later.
Feed the brief into your trained voice prompt. Not a generic ChatGPT session. A custom prompt trained on your writing style, your sentence patterns, your vocabulary, your opinions.
The voice prompt is the single biggest time-saver in this workflow. It's the difference between a draft that needs a complete rewrite and a draft that needs 15 minutes of editing. Without one, skip AI drafting entirely and write manually. Generic AI output costs more time to fix than it saves.
For a LinkedIn post: one prompt, one output, 60 seconds. For a blog post: prompt section by section (intro, each H2, conclusion), 5 minutes total.
This is where most solopreneurs cut corners. Don't. The edit pass is what separates content that builds authority from content that gets ignored.
Run three checks:
Thirty minutes for a blog post. Ten minutes for a LinkedIn post. This step is non-negotiable.
For blog posts: write the title tag (under 60 characters), meta description (under 160 characters), add 3-5 internal links to related posts, add FAQ schema if relevant, publish.
For LinkedIn posts: write the hook (first two lines), add a line break after the hook, schedule via LinkedIn's native scheduler or Typefully.
Fifteen minutes. No perfectionism. A published post at 85% quality beats an unpublished post at 100% quality. Every time.
One piece of content becomes three. This is where the repurposing system kicks in.
Blog post distribution: extract 3 LinkedIn posts from different sections, write 1 email teaser, post a summary in one relevant community.
LinkedIn post distribution: if engagement is high (above your average), expand into a blog post or newsletter issue. If engagement is low, note the topic as lower-priority and move on.
The distribution step is where most solopreneurs leave the biggest returns on the table. You already did the thinking. Reformatting takes 15 minutes. Skipping it throws away 60% of the value.
For one blog post:
Total: 75 minutes. Call it 90 with context switching.
For five LinkedIn posts (batched):
Total: 70 minutes. Under 90. The detailed batching walkthrough breaks this down further.
Step 3 of this workflow depends on a trained voice prompt. The DFY Voice System builds yours in 2-3 working days from your existing writing. One build powers every piece of content you create.
See The Voice BuildNo voice prompt. Without a trained voice prompt, Step 3 produces generic output. You'll spend Step 4 rewriting instead of editing. The workflow still works, but takes 3-4 hours instead of 90 minutes. Build the voice prompt first.
Skipping the brief. Going straight from idea to draft is like cooking without a recipe. You might get something edible. You'll probably waste ingredients. The 10-minute brief is the highest-ROI step in the entire workflow.
Editing for grammar instead of substance. AI drafts are grammatically perfect. They're also substantively empty. The edit pass catches substance problems: missing examples, generic claims, absence of opinion. Grammar isn't the bottleneck. Specificity is.
Publishing without distributing. A blog post without distribution gets zero readers for 3-6 months while Google indexes it. Distribution puts the content in front of people today. SEO puts it in front of people next quarter. You need both.
Month 1: The workflow is slow. You're learning the voice prompt, building the capture habit, finding your editing rhythm. Expect 2-3 hours per blog post.
Month 3: The workflow runs at speed. Voice prompt is calibrated. Capture list is full. Editing is instinctive. You hit the 90-minute target consistently.
Month 6: The workflow generates its own inputs. Community replies surface new content ideas. Blog posts feed LinkedIn posts. Email replies reveal what your audience actually cares about. The system starts compounding.
Don't judge the workflow by Month 1 performance. Judge it by Month 3.
90 minutes per blog post or per batch of five LinkedIn posts. That's the target with a trained voice prompt and documented workflow. Without those, expect 4-6 hours. The time difference comes from eliminating idea selection (capture system), blank-page syndrome (structured brief) and voice editing (voice prompt that produces 80% accurate first drafts).
Six steps: capture ideas as they occur, brief each piece with audience and angle, draft with a trained voice prompt, edit for substance not grammar, publish with SEO metadata, distribute to 2-3 channels. Each step has a defined input and output. The workflow runs identically whether you're writing a blog post or a LinkedIn post.
AI replaces the drafting step, not the thinking. You still choose which ideas matter, decide the angle, review for accuracy and voice. What AI eliminates is the 2-3 hours spent turning thoughts into sentences. With a voice prompt, that step takes 5 minutes. Your job shifts from writer to editor and strategist.
Three tools: an LLM with custom instructions (ChatGPT Plus, Claude Pro or Gemini Advanced at around £20/month), a notes app for idea capture (Apple Notes, Notion or a text file), and your publishing platform. Optional: a scheduling tool for social distribution. Total cost: £20/month.
The DFY Voice System builds your voice prompt, custom GPT and content workflow documentation in 2-3 working days. One build. Every piece of content benefits.
See The Voice Build