A content quality score is Autogram's AI rating of a draft before it publishes, assessing how well the post meets your standards. It acts as a quality gate so weak drafts can be rewritten or held back rather than going out. Together with a publish threshold, the score lets you automate publishing while keeping a consistent bar for what reaches your channel.
A score is only as meaningful as the criteria behind it. A rubric typically asks whether the draft covers its topic, holds the intended tone, stays within a sensible length, and avoids the vague filler that generated text drifts toward. Because these are judgements rather than measurements, two drafts a point apart are effectively equal; the number is useful for sorting, not for ranking.
The threshold is the minimum score a draft must reach before it publishes without a human looking at it; anything below is held for revision instead. Raising it means fewer posts go out unattended and more wait in your queue; lowering it means the channel keeps its cadence but weaker drafts slip through. Start strict, watch what gets held, and loosen only once the holds stop being justified.
Scores are not stable across changes you make elsewhere. Rewrite the instructions, switch the model, or shift the topic and the same threshold suddenly passes or blocks a different share of drafts. Treating a threshold as set-and-forget is the common mistake; it needs rechecking whenever the generation side changes, otherwise a gate calibrated months ago is quietly letting everything through.
Automate your Telegram publishing with AI content, quality scoring, and smart scheduling.