Recommendations
The recommendations panel is the heart of every analysis. It translates your scores into specific, ordered actions — telling you exactly what to change, why it matters, and how difficult it is to implement.
Priority levels
Fundamental issues that significantly block AI engines from extracting or citing your content. Fix these first — they represent the largest score gains. Examples: no answer in the first 200 words, missing H1, page title doesn't match target query.
Important improvements with substantial impact. Should be addressed after Critical items in the same sprint. Examples: missing FAQ schema, no author byline, no summary section at top of article.
Meaningful optimizations that compound over time. Good to include in your next content update. Examples: improve heading hierarchy, add cause-effect language, add comparison table.
Fine-tuning and polish. Implement these when you revisit the page for other reasons. Examples: add figcaption to images, improve CTA copy, minor semantic HTML improvements.
Reading a recommendation
Each recommendation card contains:
Filtering and sorting
Use the filter controls above the recommendations list to:
- Filter by priority level (Critical only, High+, etc.)
- Filter by dimension (e.g., show only E-E-A-T recommendations)
- Sort by estimated score impact (highest first)
- Sort by effort (Quick wins first — great for fast improvements)
Implementation workflow
Here's the workflow we recommend:
- Run analysis and export the recommendations list as a CSV (use the Export button).
- Add recommendations as tasks in your project management tool with the priority and effort fields.
- Implement Critical items in your next content update — these are typically simple structural changes.
- Schedule High items for your next planned content revision.
- After implementation, re-run the analysis to verify your score improved.
- Use Projects (Agency plan) to track score history over time.