Behind the Build: How I Ship 5 LinkedIn Posts a Week in 15 Minutes

This is the founder's own workflow, not a customer case study. I'm sharing the exact system I use to run LinkedSignal's content so you can steal the playbook.

When I started building LinkedSignal, I had two problems. First, I needed to be visible on LinkedIn — that's where my customers are. Second, I had roughly zero extra hours per week to spend writing posts. Every hour I spent drafting content was an hour not shipping product.

So I built a workflow that gets 5 posts out the door every week in about 15 minutes of focused work. Here it is, start to finish.

The 15-Minute Weekly Sprint

Every Sunday evening I block 15 minutes. That's it. No "content day," no batching sessions, no Notion doc filled with half-written drafts. The entire weekly content operation runs in a single sitting.

The sprint has four steps:

1. Harvest (3 minutes). I open my week's notes, Slack, and customer conversations. I grab 5 raw ideas — things I learned, customers I talked to, mistakes I made, product decisions I shipped.
2. Generate (5 minutes). I paste each idea into LinkedSignal, pick the post type (story, hot take, how-to, listicle), and let the AI draft a full post. Five posts, five generations, about a minute each.
3. Edit (5 minutes). I read each draft and tighten the opening line, kill filler, and make sure it sounds like me. This is the only step that requires real attention.
4. Schedule (2 minutes). I drop all 5 into the content calendar — one per weekday morning, staggered 7:30–8:30 AM. Done.

Total: 15 minutes. Five posts out the door. No staring at a blank page.

Why This Works (The Leverage Points)

Raw material is the bottleneck, not writing. The hardest part of LinkedIn content is not the writing — it's the thinking. What do I have to say this week? What did I learn? Once I have 5 genuine observations, the drafting is mechanical. I use the week itself as my content feedstock: customer calls, product decisions, bugs I hit, conversations with other founders. If I'm working, I already have content.

AI compresses the drafting step, not the thinking. I don't ask the AI to come up with ideas — ideas are mine. I ask it to turn my half-sentence idea into a structured post with a hook, body, and CTA. That's what generative AI is actually good at: structure and first drafts. I keep the human layer on what matters (voice, truth, opinion) and delegate the scaffolding.

Batching is overrated. Frequency isn't. Most content advice tells you to batch-write 20 posts in one sitting. I tried that. It doesn't work for me because by post 8 my ideas feel stale and recycled. A 15-minute weekly sprint keeps posts fresh and tied to whatever is actually happening in the business that week.

Editing is where voice lives. Every AI draft I ship has been edited. I cut roughly 20–30% of the generated text, rewrite the opening line, and add one specific detail that only I would have (a real number, a customer name, a dumb mistake I made). That specific detail is what makes the post sound like a human wrote it.

The Post Types That Earn Their Keep

Across the last ~3 months of running this workflow, four post formats have consistently pulled weight for me:

• Story posts. A specific thing that happened, what I learned, what changed. Highest comment rate of any format I ship.
• Hot takes. A strongly-held, mildly contrarian opinion about LinkedIn, founders, or B2B SaaS. Generate the most replies — and occasionally the most unfollows, which is fine.
• Listicles. "7 things I got wrong about X." Saves drive reach even when engagement looks flat in the first hour.
• How-tos. A concrete system or workflow (exactly like this post). These do the best long-tail — they keep getting impressions weeks after posting.

Polls, milestones, and generic "inspiration" posts underperform for me consistently. Your mileage will vary — track your own top formats instead of trusting mine.

The Prompts I Use Inside LinkedSignal

For each idea, I fill out three fields: topic, post type, and tone. Here's a concrete example from last week:

Topic: A customer asked if we use GPT-4 and I declined to answer.
Post type: Hot take
Tone: Witty, direct

That's the entire input. LinkedSignal generates a full post with a scroll-stopping hook, 3-paragraph body, and a question at the end. I edit for 60 seconds and ship it.

The mental model: treat the AI like a skilled ghostwriter who needs a 10-second brief. If you can't brief a ghostwriter, you can't brief the AI. The quality of the output is proportional to the quality of your raw idea — not the prompt.

What I Track (And What I Ignore)

I look at three numbers at the end of each month:

1. Comments per post (median). If this is trending up, my hooks and questions are working.
2. Profile views (weekly). The leading indicator for inbound. Posting frequency is the primary driver here.
3. Top post by saves. Saves mean someone will reference this later — that's the strongest long-tail signal.

I ignore impressions, likes/reactions, and follower count. They fluctuate too much week-to-week to inform any real decisions, and optimizing for them creates noise.

Steal the System

If you want to copy this workflow exactly:

1. Block 15 minutes on your calendar every Sunday evening. Make it recurring.
2. Keep a scratch doc (Notes, Apple Notes, Obsidian — doesn't matter) where you capture one-line ideas as they happen during the week.
3. Use an AI post generator that lets you pick post type and tone. I built LinkedSignal for exactly this, but any tool with structured templates will work.
4. Edit every draft. Non-negotiable. AI writes scaffolding, you provide the voice.
5. Schedule all 5 at once. Don't hand-post during the week — the friction will kill consistency.

The point of all of this isn't to "crush LinkedIn" or become a creator. The point is to be consistently visible in your professional space without content eating your calendar. 15 minutes a week is a price I'm willing to pay. I hope this helps you find a price you're willing to pay too.

If you want the exact tool I use: LinkedSignal is free for 5 posts a month, which is exactly the weekly sprint I described. You can try the full workflow without paying.