Content Doesn't Sound Like Me

The most common reason AI content sounds off is that the brand voice config is too thin. Borker can only sound like you, or like your brand, once you have told it enough about how each of them sounds.

A workspace has two voices, and the channel decides which one wrote the draft you are unhappy with. Check the voice on that channel first: a post that came out corporate when you wanted it personal is often a channel set to Brand.

Diagnose

Go to Settings → Brand & Voice and check:

• About you and About your brand: is there anything in either box beyond your name? These are the only claims Borker may make, so a thin block gives it nothing to say. Check the box for the voice that sounded wrong.
• Standing rules: are there rules that say how that voice actually sounds? Two or three sharp ones beat ten vague ones. Today these are the strongest lever you have on how drafts sound. A rule scoped to one voice never touches the other.
• Suggested rules: is anything waiting for you to confirm? A suggestion steers nothing until it is confirmed.
• Your writing: have you added writing for that voice? Each voice is counted on its own, and from five pieces up its drafts are written with a few of them in view. A brand post with no brand samples uses none rather than borrowing yours.

Add writing for that voice

Paste things that were written unedited, under Your writing. Rough is better than polished marketing copy, for your brand as much as for you, because polish reads as nobody in particular. From five pieces up, Borker drafts with that voice's writing in view; a standing rule is still the sharper tool for anything you can put into one line.

Sharpen your standing rules

Each rule is one line that gets applied to every draft:

Never open with a question.
Always open with a concrete example or a number.
Prefer: ship, iterate. Never use: delve, leverage.

Three to five sharp rules beat ten vague ones. A rule can apply everywhere, to your own posts only, or to your brand posts only.

If Borker flags a phrase that one of your voices genuinely uses, add it to Allowed exceptions under Standing rules instead of rewriting around it, and set it to the voice that uses it: allowing a word for your brand leaves the check running on your own posts.

Refine your topics

Vague topic entries produce vague content. In Settings → Brand & Voice → Topics, replace generic items in the Primary list with specific phrasings.

Instead of "AI productivity tools", try "Hidden time costs of prompt engineering for non-technical founders."

See Topics for more guidance.

Still off?

The most reliable fix: add real writing samples on Brand & Voice. Past posts, newsletter excerpts, anything that captures your tone — Borker uses your own writing as the anchor for how drafts sound.