it's thursday morning. you have an hour for distribution batch. by the time your first email doesn't sound like marketing copy, half of it is gone. you try to sound like yourself, but you sound like everything you've read. so you delete and rewrite.
the thursday morning tax
Kristian (oslo-based founder, $8k MRR) built in public by shipping cold emails, social posts, investor updates. every thursday morning he blocks the hour. what should take 40 minutes takes 60 because of the rewrites.
the first draft always sounds stiff. too formal. the tone's off. so he rewrites. same email, different phrasing. reads it again. still not quite right. rewrites a third time. now it sounds like him, but it took 15 minutes for a single email. he has four more to send.
the math is brutal. four emails. three rewrites per email. that's 45 minutes of typing and rewriting before he even gets to send. he loses the social posts. the investor update waits until next week.
it's not laziness. it's not perfectionism. it's the friction of turning thought into text fast enough that the thought stays intact.
why speech is different
typing forces you to edit in real time. you think "i should reach out about this partnership" and your fingers spell out "dear sir or madam" because you're half-asleep and half-trying to sound professional. then you delete. then you retype. then you delete again.
speaking skips that. you say the email out loud. it comes out closer to how you'd say it on a call. closer to how you actually think. you still edit, but you're editing from "that sounded right" not "that sounded wrong and i have to guess what right sounds like."
wispr flow proved this works. $14 a month, thousands of founders swear by it. except it only exists on mac. if you're kristian (windows machine, vs code, cursor), wispr is the demo you watch on product hunt and then close the tab.
the wispr problem for windows
here's what happened: every premium voice tool launched on mac first. wispr. macwhisper. others. the reasoning is obvious (mac users write, mac users buy software). the outcome is: windows founders get treated like the problem is theirs.
the customer call says it all. "i tried wispr's web demo and it was great, but i'm on windows so it doesn't exist for me."
that's not hyperbole. wispr has no windows app. the web version is a toy. so windows founder shops for alternatives. finds otter.ai (too expensive, too many features, too much ui). finds speechify (not built for writing, built for listening). finds nothing that's actually good for the specific job (draft cold emails in 40 minutes on thursday morning).
so kristian types. still rewrites four times. still loses thursday batch time. still resents that the tool world decided he doesn't count as a real market.
what changes with local, windows-native voice
recitey was built for this specific gap. runs whisper on your device (no word limit, no metering, no cap on free tier). works in slack, email, github issues, your text editor, everywhere you type. windows-native, no browser, no login tax.
kristian records his first email. 90 seconds to speak what would take 5 minutes to type. the draft is messier than a typed draft (more pauses, some ums), but it's also more honest. sounds like him talking, not him performing.
he reads it. edits once (removes a false start). reads again. hits send. total time: 7 minutes instead of 15.
now he has three more emails left in the hour. he actually sends the social posts. the investor update goes out.
it's not magic. he didn't become a faster writer. but the friction dropped enough that he didn't lose the batch to perfectionism.
the thing about first drafts
once you start drafting by voice, something shifts. you stop trying to sound polished. first drafts are supposed to be rough. that's the feature.
kristian's emails sound like kristian now, not like a template. they get responses. they don't all get responses, but the ones that do tend to be actual conversations, not the "thanks but no thanks" email that comes after a perfect cold outreach.
the trade-off: you lose the ability to hide. typed prose lets you copy tone from other emails, from hn comments, from anywhere. voice makes you write in your own voice or not at all. for a founder, that's usually a win, even though it feels risky at first.
it takes about 30 seconds of using a voice tool to realize this is real, and another week to stop feeling self-conscious about talking to your computer.