You sound sharp on the call. Five minutes later, you're rewriting the same thought for Slack. Three times. Four times. The words you said effortlessly become cautious, stilted, smaller on the screen.
The hidden cost of async writing in a second language
Maria is a senior account executive at a European B2B SaaS company. On video calls with US enterprise clients, she's sharp. She asks the right questions. She anticipates objections. She closes deals. Her manager notes her exec presence. She sounds like the person the client wants to work with.
But when the call ends, she moves to Slack, writing up the next steps, summarizing outcomes for her team in Boston, crafting a follow-up email to the procurement lead. Something shifts. The version of her that shows up on screen isn't the version they heard on the call. It's more careful. More hesitant. Smaller.
It's not a vocabulary problem. It's not that she doesn't know English. She's had eight years of enterprise sales conversations in English. The problem is the medium. Speaking forces you to think. Writing in a second language asynchronously, where every word gets parsed before you hit send, forces you to over-think.
Maria isn't alone. Swedish PMs, German senior engineers, Spanish-speaking account managers across Europe; competent, confident in meetings, hesitant on the page. They read their Slack drafts three, four times before sending, checking for something that was never wrong. The cost isn't vocabulary. It's time. Knowledge workers spend roughly 23% of their day editing and polishing prose. For someone working in a second language, that number climbs higher. Not because they're incompetent. Because the medium strips away the confidence that speaking preserves.
The rewrite cycle nobody talks about
Here's what the 30-minute Slack message actually looks like:
Maria types: "I think we should push the demo to next week so we can integrate their feedback on pricing."
She reads it back. It sounds stiff. Smaller than how she said it on the call. Too verbose. She rewrites:
"Given their feedback on pricing, I'd suggest we push the demo to next week to integrate their input."
Better, but now it sounds formal in a way that doesn't match her usual Slack tone. Impersonal. She tries again:
"I'd recommend pushing the demo to next week. That gives us space to integrate their pricing feedback."
Still something off. Is "I'd recommend" too much? Does it sound presumptuous? She rewrites:
"Based on the pricing feedback, pushing the demo to next week would give us time to integrate their input before we move forward."
She reads it. Passive. Careful. Not her. Not how she described it on the call two minutes ago. One more time:
"Let's push the demo to next week; that way we can fold in their pricing feedback before we present again."
Fourteen minutes from thought to send. For a 15-word Slack message.
This happens for the other draft too. A follow-up to the procurement contact:
She speaks it naturally: "Thanks for the call. I wanted to circle back on the pricing tiers you asked about."
She writes: "Thank you for taking the time to discuss. I'm following up on the pricing tier structure you raised."
Rewrites: "I wanted to follow up regarding the pricing tiers we discussed."
Rewrites: "Following our conversation, I'd like to circle back on the pricing tier question."
Settles on: "Quick follow-up on the pricing tiers from today. I have some thoughts."
Eight minutes.
That's the hidden tax. Not vocabulary. Not grammar. Time. And the distance between how you sound when you're thinking in real-time and how you sound when you edit.
Why Grammarly doesn't solve this
Grammarly catches typos. It suggests more concise phrasing. But it doesn't solve the real problem, which isn't grammar. It's friction.
When you write in a second language, you're doing double work. You're thinking the thought, then translating it, then reading it back to check if the translation sounds right, then rewriting because it doesn't sound like you, then rewriting again because "concise" and "authenticity" seem to pull in opposite directions.
That cycle, the one a native speaker does once or twice, you do it four or five times before you hit send.
Grammarly assumes you're a beginner. You're not. You're someone who is sharp on calls and is fighting the medium, not the language. Hemingway Editor offers similar value: shorter sentences, clearer structure. But neither tool addresses why you're rewriting in the first place.
The problem isn't that your writing is bad. It's that you don't trust it until you've read it five times, which is a confidence problem, not a language problem. And no grammar checker fixes confidence. It just adds another layer of editing.
What actually changes when you speak first
Speaking first, writing second changes the math.
When Maria uses voice writing, she doesn't draft. She speaks the thought exactly as she would say it on a call: "Let's push the demo to next week so we can integrate their pricing feedback before we present again."
The system runs Whisper locally on your device, so there's no variable cost, no word cap, no metering. Your voice comes through; the actual cadence and phrasing you use in meetings, without the translation step. The rough draft is ready to send in under two seconds. It isn't perfect. It's good enough. It sounds like you.
She hits send. No rewrite cycle. No fourteen-minute tax. The version of Maria on Slack sounds like the version her client heard on the call.
She saves roughly seven minutes per message on average. Over a week of enterprise sales communication, that's roughly forty minutes. Over a month, that's two and a half hours she gets back.
More importantly, she gets to keep her voice.
The performance review gap that isn't real
Your last review probably included: "excellent exec presence on calls" and, tucked into the writing section, "could be more concise in writing."
You know that second feedback isn't actually true. It's just the medium. Speaking forces clarity because you don't have time to second-guess yourself. Writing in a second language, in Slack, where every message is permanent before it's sent, forces caution.
The gap between those two pieces of feedback isn't a skill gap. It's a confidence gap. And a time gap. And a medium that doesn't preserve the version of you that shows up when you're thinking in real-time.
That's not a writing problem. That's a friction problem.
The trade-offs worth considering
Voice-first writing isn't frictionless. There's a learning curve to speaking your drafts. You might feel self-conscious saying your Slack message out loud, especially in an open office or on a video call.
Some moments will still be keyboard moments: late-night work, sensitive conversations, moments where you need to think through text. And that's fine. No tool is universal.
But for the bulk of what you write, follow-ups after calls, summaries for your team, status updates, deal notes, internal recaps, the ones that currently take 30 minutes and come out sounding smaller than intended? Those become a clear thought into your microphone. Sent. Done.
You also don't need to maintain a separate account or tool. Voice writing in Recitey works in Slack, email, browsers, your terminal, your notes app, anywhere you type on Windows. No context switching. No special interface. Just speak, and what you meant lands on the page.
The version of you that belongs everywhere
On a call, you're sharp. Your client hears you. They trust you because of how you think in real-time, not because of how carefully you've edited.
Async communication deserves the same version of you. Not the rewritten, over-thought, filtered-through-self-doubt version. The version that closed the deal. The version that's actually you.