Four parts, one loop.
Voice profile
A generated summary you can inspect against its source writing.
It is written in sentences, not weights
PostLabz reads what you have already published and writes down its observations. You can inspect the resulting tone, audience, pillars, and source counts before generating drafts.
The profile records which supplied website or channel samples informed it. Generated observations should still be reviewed by the account owner.
Short. Fourteen words on average, and you rarely go past twenty five.
With a number or a flat statement of fact. Never a rhetorical question.
The team by role, customers by what they told you. You do not name individuals without asking.
Periods and commas. No em dashes, no exclamation marks, no emoji.
Drafting
Illustrative drafts showing one topic in requested channel formats.
The same Q3 result, written three different ways
The product update is ready. Here is what changed and where to read the release notes.
The product update is ready. Three changes, with the release notes linked below.
A short walkthrough of the product update, based on the supplied release notes.
Approvals
Illustrative queue showing review, edit, reject, and publication checks.
The product update is ready. Here is what changed.
Three things we got wrong shipping the approval flow
How the team runs a Friday retro in 25 minutes
Rejection removes the draft from the active review queue. The current product does not claim to learn a rationale automatically.
Degraded or format-mismatched drafts remain editable and rejectable, but approval is blocked.
Results
Post performance is shown only when real metrics sync from Postiz.
No published metrics yet
After connected Postiz channels publish content and return metrics, results can summarize actual topics, platforms, and hooks. Simulated performance is not presented as customer proof.