You explain a complex idea on a sales call and the room goes quiet for a beat. Exec presence. Then you draft the follow-up Slack message and spend thirty minutes making it smaller, safer, less you. This gap isn't a language problem.
The gap between how you sound and how you write
You're fluent in spoken English. Your colleagues say so. Your performance reviews say so. On a Zoom call with a US executive, you're sharp, present, quick with the details. Your accent is there, but your intelligence is too. People follow your reasoning.
Then you switch to Slack.
The message sits in the compose field. You read it once. The tone feels wrong. Too formal. You read it again. The sentence is too long. You read it a third time and rewrite the opening. Now it sounds stiff. You delete that and try again. Three or four iterations later, you send something that accomplishes the task but doesn't sound like you. It sounds like the careful version of yourself.
This pattern isn't new to you. It happens on almost every async message: follow-ups after calls, updates to your team, deal notes, proposals. The pace of writing forces a kind of self-editing that speech never requires.
Why the existing tools make this worse
Grammarly catches a misplaced comma. Google Translate catches vocabulary. But neither gives you back the thirty minutes. Neither preserves the speed at which you actually think.
What they do is formalize the process of writing, layer on another step. Grammarly suggests a rewording that's technically more polished and emotionally further from what you meant. Google Translate smooths out your accent, which sounds good in theory until you realize it also smooths out the person inside it.
The fundamental assumption behind most writing tools is that English speakers are trying to write better. But you're not trying to write better. You already know how to write correctly. You're trying to write faster, and you're trying to preserve the voice that survives when you speak.
The real bottleneck
The issue isn't grammar or vocabulary. It's translation friction and the async medium's pace.
On a call, you have real-time feedback. If someone misunderstands you, they interrupt. You clarify. The conversation moves. Your brain operates at call speed, which is the speed you're comfortable with. You don't have time to second-guess yourself, so you don't.
In Slack, there's no time pressure except the one you create. You have the luxury of revision, which becomes a trap. Because now you can hear every possible tone and worry about each one. Is this professional enough? Is this too casual? Does this sound native? The rewriting loop starts. Three iterations. Four. You're still not sure.
The gap isn't between your abilities. It's between the speed at which you think and the speed at which writing lets you operate.
What changes when you draft at speaking pace
Maria leads a PM team at a Berlin software company where English is the working language but the culture is async-first and heavily document-driven. In calls, she's sharp. On the written side, where everything from Slack to specs becomes the record that her Swedish and US counterparts reference, she'd spend 45 minutes on a 300-word follow-up email: 5 minutes to draft, then four separate rewrites before sending.
When she could draft at the speed she speaks, the first version already sounded like her. Already had the pace and confidence of the call. The refinement still happened, but the foundation was right. She kept her voice and lost the forty minutes per message.
Recitey runs locally on your device using Whisper's open-source model, processes speech in real time, and outputs clean text without metering, without caps, without assuming you need help with English. You speak. You get back what you recognize.
The identity question
Writing is where you show up async. It's where you document your thinking, build credibility with stakeholders you don't see every day, leave a record of your judgment. If the medium forces you to sound like the careful version of yourself, then your async presence, your documented self, isn't authentic.
The gap isn't your English; it's the medium itself, which gives you unlimited time to second-guess and revise yourself smaller. You get back the half hour, and more importantly, you get back to sounding like yourself.