The batching workflow that turned LinkedIn from a time sink into a system. Voice prompt, 5 content types, one Monday session.
90 minutes. 5 posts. One sitting. Monday morning: pick 5 content types from a rotating roster, run each through a voice-trained prompt, edit for 8 minutes per post, schedule in Buffer. Done by lunch. Posted Tuesday through Saturday.
Five LinkedIn posts a week sounds like a full-time content job. It isn't. It's a 90-minute Monday morning routine once you have two things in place: a voice system and a content type roster.
I've posted 5 times per week for the last 14 weeks using this exact workflow. Average time: 87 minutes. Longest session: 112 minutes (that was a week with two long-form posts). Shortest: 68 minutes.
Here's the full system.
This workflow does not work without these. Skip them and you'll spend 3 hours writing 5 mediocre posts instead of 90 minutes writing 5 good ones.
Prerequisite 1: A voice system. A documented prompt that makes ChatGPT or Claude write in your voice. Not "write like a professional" or "write casually." Your actual sentence patterns, vocabulary, opinion style and structural habits. The full build process takes 2 to 3 hours. After that, every post starts 80% done.
Prerequisite 2: A content type roster. You need 7 to 10 content types you rotate through. Not topics. Types. A type is a structural format: the opinion post, the how-to breakdown, the metric share, the mistake post, the tool recommendation. Topics change weekly. Types stay fixed.
With both in place, writing a LinkedIn post goes from "stare at blank screen for 25 minutes" to "pick a type, pick a topic, run the prompt, edit for 8 minutes."
Every week I pick 5 from this roster of 7. The rotation keeps the feed varied without requiring me to invent new formats.
Type 1: The opinion post. State a position. Support it with one specific example. Close with the implication. 150 to 200 words. Example: "Most prompt packs are useless. Here's what works instead."
Type 2: The how-to breakdown. Pick one task. Break it into 3 to 5 numbered steps. Each step gets one sentence of explanation. 200 to 300 words.
Type 3: The metric share. Share one real number from your business. Explain what caused it. Say what you'd do differently. 100 to 150 words. This is the highest-engagement type consistently.
Type 4: The mistake post. Name a specific mistake. Explain what happened. Say what you do now instead. 150 to 200 words.
Type 5: The tool or resource recommendation. Name one tool. Say what it replaced. Give the specific use case and one concrete result. 100 to 150 words.
Type 6: The before/after. Show the state before a change and after. Works for processes, metrics, mindset shifts. 150 to 200 words.
Type 7: The question post. Ask your audience one specific question. Add 2 to 3 sentences of context about why you're asking. Shortest type. 60 to 100 words. Best for engagement rate but weakest for driving traffic.
Here's exactly what I do, with time stamps from a real session.
Open the roster. Pick 5 types for the week (I try not to repeat last week's exact combo). Assign a topic to each from my running topic list. This takes under 5 minutes because the decisions are constrained.
Open ChatGPT. Paste the voice system prompt. Ask for post type 1 on topic 1. While it generates, review and edit. Then ask for post 2. Two posts drafted and edited in 10 minutes.
Same process. Three more posts. The voice system means each draft needs 3 to 5 edits, not a full rewrite. I fix awkward transitions, swap generic words for specific ones, and trim anything over the word target.
Read all 5 posts as a set. Check for repetitive hooks (the same opening pattern across multiple posts is the most common issue). Check that no two posts make the same point. Swap one if needed. This is the quality gate.
Each post gets a hook rewrite. The AI-generated hook is usually functional but not sharp enough. I rewrite the first line of each post manually. Then I write a first comment for each (a follow-up question or additional context). First comments boost early engagement.
Drop all 5 posts into Buffer. Schedule: Tuesday 7:45am, Wednesday 8:00am, Thursday 7:30am, Friday 8:15am, Saturday 9:00am. Slightly varied times because LinkedIn's algorithm rewards non-robotic posting patterns. Add the first comments as reminders on my calendar.
Done. 90 minutes. Five posts scheduled for the week. I don't think about LinkedIn again until next Monday.
This workflow depends on a voice prompt that makes AI write like you. The Voice Build creates yours in 5 business days. $497, one-time, no subscription.
See The Voice BuildThe biggest objection I hear: "Won't the posts feel stale if I write them all on Monday?"
No. For three reasons.
1. Your audience doesn't know when you wrote it. A post about your marketing process is equally relevant whether you wrote it 3 hours ago or 3 days ago. The only content that goes stale is news commentary, and I keep one slot flexible for that.
2. Context switching kills quality. When you write daily, you lose 10 to 15 minutes per session just getting into writing mode. Over 5 days that's an hour of wasted mental warmup. Batching means you warm up once and ride that focus for 90 minutes straight.
3. Batching lets you see the week as a set. The edit pass at minute 30 catches problems that daily posting never reveals: three posts in a row with the same tone, two posts making the same argument from different angles, a week with no how-to content. You can only fix these when you see all five together.
The Monday routine is the visible part. Behind it sits a content batching system that handles the less glamorous work: topic sourcing, content type rotation tracking, performance review.
Every Sunday I spend 10 minutes reviewing the previous week's posts. Which got the most engagement? Which fell flat? I note patterns. If metric shares consistently outperform opinion posts, I schedule more metric shares next week.
Topic sourcing runs on a simple system: any time I see a question in a LinkedIn comment, a Reddit thread or a client conversation, I add it to the topic list. By Monday morning I have 15 to 20 candidate topics. I pick 5. The surplus rolls to next week.
Some specific numbers from running this workflow since mid-January:
Weeks 1 to 4: Average impressions per post: 180. Engagement rate: 2.1%. Follower growth: 12 per week. Normal starting range for a small account.
Weeks 5 to 8: Average impressions per post: 340. Engagement rate: 3.4%. Follower growth: 28 per week. The algorithm starts rewarding consistency around week 5.
Weeks 9 to 14: Average impressions per post: 520. Engagement rate: 4.1%. Follower growth: 45 per week. Compounding effect. Earlier posts still generate profile visits weeks later.
None of these numbers are viral. That's the point. Viral is unpredictable. Consistent output at 500 impressions per post is predictable and stackable.
Each LinkedIn post feeds two other channels automatically. The full repurposing strategy is documented separately, but the short version:
LinkedIn post → X thread. Take the 3 key points from a LinkedIn post, rewrite each as a standalone tweet, thread them. 5 minutes per repurpose.
LinkedIn post → Newsletter section. The week's best-performing LinkedIn post becomes a section in Friday's email newsletter. Already written. Just reformat.
LinkedIn post → Blog seed. Any LinkedIn post that gets 3x average engagement gets expanded into a full blog post. The LinkedIn version becomes the outline. This is how roughly one third of our blog posts start.
Skipping the voice system. Without it, each post takes 25 to 30 minutes of heavy editing instead of 8 minutes of light editing. The 90-minute window becomes 3 hours. Most people quit after 2 weeks.
Writing all 5 posts about the same topic. Your audience sees all five across the week. Five posts about AI content systems in a row is a content flood, not a content strategy. Use the type roster to force variety.
Editing during generation. Don't stop the AI mid-draft to fix things. Let it finish the full post, then edit. Interrupting the flow adds 20% more time for no quality improvement.
Posting without first comments. LinkedIn's algorithm gives posts an initial distribution boost based on early engagement. A first comment from you (posted 1 to 2 minutes after the main post) gives your connections something to reply to. It's a small thing that consistently adds 15 to 20% more impressions.
The Voice Build gives you the prompt that makes this 90-minute workflow possible. One-time $497. Yours forever.
See The Voice BuildThree to five. Below three, the algorithm forgets you exist and your audience loses continuity. Above five, you hit diminishing returns unless you have a large engaged following already. Five posts per week is the ceiling for most solopreneurs because quality drops after that without a team.
LinkedIn does not detect or penalize AI-written content directly. What it penalizes is generic, low-engagement content, which is what most AI-written posts are without a voice system. Posts written with a trained voice prompt perform identically to manually written posts in reach and engagement metrics.
Tuesday through Thursday, 7:30 to 8:30am in your audience's timezone, consistently outperforms other slots. Monday posts compete with weekend catch-up. Friday posts get buried by end-of-week drop-off. But consistency matters more than timing. Posting at 9am every day beats posting at the "perfect" time sporadically.
Yes. Batching is how you maintain consistency without burning out. Write all five posts on Monday, schedule them across the week. The only exception: if a major industry event happens mid-week, swap one scheduled post for a timely reaction post. Keep one slot flexible for this.
The quiz tells you which AI marketing system fits your business right now. Takes 2 minutes.
Take the Free Quiz