# Three scores on every draft, and a gate before anything publishes.

Every blog draft Sphen.ai writes carries a Quality, an SEO and a GEO score. Below the publish floor the draft is marked Not ready to publish yet, and auto-publish waits until it passes.

Canonical: https://sphen.ai/features/blog-quality-score/

## How does Sphen.ai score blog drafts?

Every blog draft gets three scores. Quality checks length, heading structure, title, meta description, FAQ section, structured content and links. SEO measures how well the draft targets its keyword in titles, structure and meta description. GEO measures how citable it is for AI assistants. A draft below the publish floor is marked Not ready to publish yet.

## What do you see in the app?

- **Scores on the draft card**: Each draft in Your article plan carries Quality, SEO and GEO badges, or Scores pending while the worker is still scoring. The Scores panel explains both model scores in one line each.
- **What Quality checks**: Content length, heading structure, title quality, meta description, a FAQ section, structured content such as lists and tables, and links. Length weighs most; the other checks pass or fail. Advisory notes can point at thin evidence, few external links or a generic author without affecting the score.
- **The publish gate**: Publishing is blocked while a draft is below the floor, with the message Not ready to publish yet — improve it before publishing. The AI Blog Planner's Auto-publish mode applies the same floor, so unattended publishing never ships a weak draft.

## The problem

A draft is not an article until someone checks it.

- **Thin posts hurt**: Short, unstructured articles without an answer block or FAQ are rarely cited by assistants and rarely rank.
- **Review time is the bottleneck**: Reading twenty drafts in full to find the three worth publishing is where blog programmes stall.
- **Automation without a gate**: An auto-publisher that ships everything will ship a weak post eventually.

## How it works

1. **Scores land with the draft**: A draft card shows Scores pending until the scores arrive, then Quality, SEO and GEO as numbers.
2. **Read what each one means**: SEO: how well the draft targets its keyword in titles, structure and meta description. GEO: how citable the draft is for AI assistants — direct answers, FAQs and structure.
3. **Improve, then publish**: Below the floor the draft shows Not ready to publish yet. Use Edit with AI or Rewrite with AI, or edit by hand, and publish once it passes.

## You stay in control

- Publishing is always your decision; the scores order the review queue.
- Auto-publish only ships drafts above the floor; everything else waits for you.
- Scores update when you change the draft.

## Questions

### What is the publish floor?

A draft needs a Quality score of at least 85 and a GEO score of at least 70 to publish. Below that it is marked Not ready to publish yet. The floor applies to manual publishing and to Auto-publish alike.

### How do I raise a low score?

Open the draft and use Edit with AI for small changes or Rewrite with AI with an instruction such as add an FAQ section. You can also edit by hand. The scores update with the new content.

### What is the difference between SEO and GEO?

SEO measures how well the draft targets its keyword in titles, structure and meta description. GEO measures how citable the draft is for AI assistants: direct answers, FAQs and clear structure.

### Is the Quality score AI-generated?

No. Quality is computed deterministically from the title, body and meta description, so the same draft always gets the same score. SEO and GEO come from the editorial review step.

### Does scoring use my blog post quota?

No. Scoring is part of every draft. Only generated posts count against the monthly allowance.