I recently stumbled across a workflow that completely changed how I think about writing speed. You know that feeling when you’re staring at a blank doc, fingers hovering over the keyboard, and the words just won’t come? Turns out, the problem might not be writer’s block. It might be the keyboard itself.
This LinkedIn creator shared a process that’s genuinely impressive: they write a newsletter for nearly 400,000 weekly subscribers, and they don’t type a single word of the first draft. Instead, they talk. The tool that makes it happen is called Wispr Flow, a voice-to-text AI that captures your spoken words and turns them into clean, written text.
The reasoning is simple and backed by real numbers. Most people type at around 40 words per minute. But when you talk? You hit roughly 143 words per minute. That’s more than 3x faster. And here’s what makes it interesting: your spoken voice carries a natural rhythm, personality, and tone that typed words often flatten out. Speaking your ideas first means your writing sounds more like you and less like everyone else.
🎙️ The Full Voice-to-Newsletter Workflow
The original poster laid out a clear, repeatable process. Here’s how it works, step by step:
- Download Wispr Flow from wispr.ai. It’s free to start with, and it runs quietly in the background on your machine.
- Set your activation key. The author uses the Shift key. This means you hold Shift whenever you want Wispr to start listening. No clicking through menus, no opening a separate app.
- Open your notes app (any text editor works) and hold your activation key down.
- Talk for about 60 seconds about your topic. Don’t filter yourself. Don’t try to sound polished. Just speak naturally about what you want to cover, as if you’re explaining it to a friend.
- Let Wispr transcribe everything. If it misspells a name or technical term, correct it once. The tool remembers your corrections permanently, so it won’t make the same mistake twice.
- Take your spoken draft to Claude AI. Copy and paste this prompt:
“I want to write a newsletter on [TOPIC]. I want to [GOAL] + [CONTEXT]. Make sure we build the outline first. Use ‘AskUserQuestion’ to ask me questions.”
- Edit hard. The AI gives you roughly 70% of the way there. Your taste, your judgment, and your editorial eye handle the remaining 30%. Cut any sentence that sounds robotic or generic.
Total time for a solid first draft? Under 10 minutes.
🧠 Why This Process Actually Works
There are a few things happening here that make this more than just a productivity hack.
- Speed removes friction. When you can dump your thoughts at 143 words per minute instead of 40, you spend less time in the painful “getting started” phase and more time in the creative zone.
- Voice captures personality. Typed text tends to default to safe, generic phrasing. When you speak, you use your natural cadence, your favorite expressions, your real voice. That’s what makes writing feel authentic.
- AI handles structure, you handle soul. The prompt shared by the expert is smart because it tells Claude to build an outline first and ask clarifying questions. That means the AI isn’t just guessing what you want. It’s collaborating with you.
- The correction memory is underrated. Wispr remembering your fixes means it gets better over time. Industry-specific jargon, names of people you mention often, brand terms: correct them once and move on.
✍️ Tips to Get the Most Out of This Workflow
If you want to try this yourself, here are a few practical pointers that’ll help you get better results from day one:
- Don’t script your speech. The whole point is to capture raw, unfiltered thinking. If you rehearse what you’re going to say, you lose the natural voice advantage.
- Use the prompt template as a starting point, not a rigid formula. Swap in your specific topic, goal, and context. The more detail you give Claude upfront, the less editing you’ll do later.
- Be ruthless in step 7. The 70/30 split the author mentions is key. AI-generated text often has a telltale smoothness that reads as flat. Your edits are what turn a draft into something worth reading.
- Batch your voice dumps. If you have multiple newsletter ideas, record them back to back in separate notes. Then process them through Claude one at a time. You can capture a week’s worth of content ideas in under five minutes of talking.
🔑 The Bigger Takeaway
Typing makes you sound like everyone else. Your voice doesn’t.
That line from the original post really stuck with me. We spend so much time optimizing our writing tools, but the most powerful differentiator might just be speaking your thoughts out loud before you ever touch a keyboard. The combination of voice capture plus AI structuring plus human editing creates a workflow that’s fast, personal, and surprisingly effective.
Whether you’re writing newsletters, blog posts, or even long LinkedIn updates, this voice-first approach is worth experimenting with. The best writing sounds like a real person talking. So why not start by actually talking?
Check out the full LinkedIn post for more details from the person who shared this workflow.