Kristian paid for Wispr Flow. Fourteen dollars a month. The free tier convinced him voice drafting actually worked, three cold emails sounded like him instead of corporate templates. So he paid. Problem: Wispr doesn't ship on Windows.
Here's what happens next if you're on Windows.
You find the web version. Bookmark it. Tell yourself you'll use that for now. Three days later the workflow friction kills it. It's another tab, another login, and by the time you've switched to the web app and back to Linear, you've lost momentum. You end up typing.
The Mac-first assumption
Every premium voice-writing tool that's launched in the past three years shipped Mac first. Wispr. Otter. Superwhisper. The logic makes sense from a business angle. Creators skew Mac. You optimize for velocity and ship where the money is. Fair enough.
But if you're a founder on Windows, you notice it everywhere. Cursor went native on Windows from day one. Notion runs the same on both. The moment someone charges $14 a month though, the Windows version becomes "coming later" or doesn't exist. You learn to stop hoping.
Kristian's Thursday mornings are blocked off. One hour. He calls it his "distribution batch." Cold outreach DMs to design partners. Investor updates. Support replies to early customers. Notion posts. That one hour sets whether the week happens or stalls. A tool that adds friction costs him 15 minutes of productive time, which costs him the week's momentum, which costs him revenue. At pre-PMF every minute compounds.
Why voice wins for first drafts
He used to type the cold emails. They came out stiff. Formal. Like he was afraid of the recipient. Voice changed that. He speaks them out loud and something shifts. The draft sounds natural. Conversational. Actually him. He sent three. Got two replies. Text hadn't done that.
But here's the trap: voice-first means living in a web app, and a web app is just another place your attention has to go.
You can't speak into Slack while drafting a team reply without context-switching to another tab. You can't document a bug in GitHub at 11pm without jumping to a browser and back. You definitely can't narrate what you just built in Cursor without leaving the code.
Wispr Flow is genuinely good at what it does. The voice quality is clean. The editing is intuitive. But "good" and "usable at the moment you need it" aren't the same thing.
Every context switch costs attention. The research calls it "attention residue." You switch to the web app, come back, and part of your brain is still there. The friction isn't obvious, it's not slow. It's just enough. Just enough to break momentum. Just enough to make you type instead of speak.
The Windows-native move
Recitey was built around this exact friction.
It runs natively on Windows. Not a web app. Not another login. You speak in Slack, in Cursor, in your email, in Linear. The first draft comes out rough. You polish it in whatever app you're already using. No context switch. No overhead.
Kristian tested it one night at 11pm in Cursor. He was documenting a bug. Spoke it out loud. Listened to the cleanup. Sent it. No tabs. No attention residue. Just speak and ship.
The time-value math clicks instantly. Thirty minutes a day saved on communication alone pays for itself. At any salary. At pre-revenue that's not overhead. That's runway.
It's not about Wispr being bad
Wispr is genuinely good. Kristian still uses it for some things. Longer writing that needs structure. Things that benefit from the web experience. The problem isn't Wispr.
The problem is the assumption baked into every $14-a-month voice tool: that the person paying is on a Mac. And the quiet resentment that builds when you realize you're using the second-class version.
Windows has been 77 percent of the market for years. But look at the premium tools. The creator tools. The founder tools. You'd think Windows was a rounding error.
The unspoken math
There's something else here that doesn't make it into the marketing.
He used to draft that cold email four times. Four versions. Each one trying a different angle to sound less corporate, less afraid. Then he'd send the safest one. Or not send it at all.
That's the real cost of friction: not that it slows you down, but that it kills your confidence. Every context switch, every extra step, every login becomes a moment to second-guess. Is it worth the effort? Maybe I'll do this later.
Now he speaks it once. Listens. Polishes in place. Ships.
That's not faster typing. That's a different kind of confidence. The kind that comes from not pretending the draft is more polished than it is. The kind that comes from the tool getting out of the way, so the speed of your thinking matches the speed of your typing.
It's also the kind of confidence that compounds. One cold email a day. Five a week. Some turn into partnerships. Some don't. But the difference between "I'll draft this later" and "I'll speak this now" is the difference between shipping and deferring. And at pre-revenue, that's everything.