Understanding your scores

SEOBetter shows two numbers for every article: the GEO Score and the SEOBetter Score. Both run locally on your site and both are free on every plan.

GEO Score

GEO stands for Generative Engine Optimization: how likely AI answer engines (ChatGPT, Perplexity, Google AI Overviews, Gemini, Claude) are to cite your article.

The GEO Score is a weighted average of 14 to 15 individual checks: readability, citations, statistics, expert quotes, tables, lists, section openings, keyword density, freshness, and more. A score of 65 to 80 on a first generation is normal. 80+ is strong.

SEOBetter Score

Same checks, different grouping. The GEO Score tells you how good the article is overall, but not where the missing points went. The SEOBetter Score re-groups the checks into layers so you can see which area is weak:

SEO foundation: readability, keyword density, freshness, a direct answer up top
AI citation quality: citations, quotes, factual density (weighted highest, because research shows statistics, quotes, and citations are what AI engines actually cite)
Extractability: tables, lists, section openings that AI engines can lift cleanly
Schema coverage: how complete your structured data is

The two numbers can differ slightly. That's expected; they weight the same checks differently.

The suggestions list

Under the scores, SEOBetter lists specific suggestions, with the high-priority ones first and the rest collapsed under "additional suggestions". Each one maps to a failing or weak check, so working through the high-priority items is the fastest way to raise the score.

The Re-analyze loop

Scores don't update by themselves while you type. The workflow is:

Open the article in the WordPress editor. The SEOBetter panel shows the current scores and suggestions.
Fix something from the suggestions list (add a statistic, break up a wall of text, add a table).
Click Re-analyze. The scores recompute on the spot.
Repeat until you're happy.

Scores also recompute automatically every time you save the post.

Don't chase 100

The score is a proxy, not the goal. An article at 85 with accurate, useful content beats a 95 stuffed with checkbox filler. Use the suggestions to fix real weaknesses, then publish.