for substack writers

brand voice guide for substack writers

substack is built for personal publishing. subscribers pay for your take, your rhythm, your way of opening an argument. we will show you how to profile your best issues and scan every post and note before publish so growth pressure does not flatten what made people subscribe.

npm i -g @holdyourvoice/hyv
see how it works
3 issues
enough to build a voice profile
notes
where voice drift often starts
$1
first month ยท then $9/mo

the substack voice problem: speed beats scan

your first twenty issues had a voice subscribers replied to. then you added notes, used ai for outlines, and started posting more often to feed the algorithm. the cadence is still weekly but the texture changed. openings feel templated. transitions sound like every other substack in your niche.

in our analysis of substack publications, voice consistency drops within six weeks of unedited ai drafts entering the workflow. notes and posts start sharing the same generic rhythm, and paid subscribers are the first to churn because they remember what your writing used to feel like.

profile your best issues. scan everything else.

๐Ÿงฌ

anchor on top issues

paste your three best-performing issues into hold your voice. we extract the sentence rhythm, transition style, and vocabulary subscribers already respond to.

๐Ÿ“

scan posts and notes

copy your draft out of substack, run a voice scan, repair flagged lines, paste back in. same workflow for long posts and short notes.

๐Ÿค–

ai outlines, human openings

use ai for research and structure if you want. always write the first paragraph yourself, then scan the full draft to catch where ai flattened your rhythm.

๐Ÿ“ˆ

track drift over time

see whether your last five issues still match your anchor profile. voice drift is gradual. a score trend catches it before subscribers do.

r

i was posting notes daily and my long-form started sounding like my short-form: rushed and generic. hyv flagged exactly where my openings went flat. paid retention improved because the issues felt like the writing people originally subscribed for.

renee p. ยท substack writer, culture and work

+19%paid retention over 90 days
9.0/10voice score on last 3 posts
8 minavg scan per issue

what most substack advice gets wrong

the problem is not posting frequency. substack rewards consistency, but subscribers reward voice. growth tactics that push safer, blander takes will get impressions and lose the readers who actually pay.

  • notes and posts should score against the same voice profile
  • ai outlines save time but steal your openings if you skip the voice pass
  • your three best issues are a better style guide than any template
  • growth pressure is the main driver of substack voice drift
  • paid subscribers notice voice change before free readers do
  • scan before publish, not after complaints in the replies
individual plan
$1 first month

then $9/mo

  • one voice profile, built from your best work
  • unlimited scans and rewrites
  • profile learning loop
  • cli + mcp for cursor and claude
  • multiple plan $29/mo for teams
npm i -g @holdyourvoice/hyv
shashank
ai
shashank

founder of hold your voice. writes about brand voice, ai writing patterns, and the craft of sounding like yourself.

co-written with ai as sidekick. the ai helped identify substack-specific failure modes: notes drift, ai outline openings, and growth-driven blandness.

subscribers paid for your voice. scan before you publish.

profile your best three issues. check every post and note before it goes live. stop shipping work that sounds like a template. it is free for 3 days.

npm i -g @holdyourvoice/hyv