Skip to main content
Pharos analyzes your visibility data, prompt responses, content health scores, and competitive gaps to generate prioritized recommendations for improving your AI presence. The Recommendations page has two tabs: Feed, where new recommendations appear, and Tracking, where you manage the ones you are actively working on. Click Generate at any time to refresh your recommendations based on the latest analysis data. Generation typically takes 10–60 seconds while Pharos processes your current visibility scores and gaps.

​
Feed tab

The Feed tab shows all active recommendations waiting for you to review and act on. Each recommendation card includes:
  • Title and description — what to do and why it will improve your AI visibility
  • Impact level — High, Medium, or Low, indicating how much the action is expected to move your visibility score
  • Category — the type of work involved (content, technical, citation, topic coverage, etc.)
  • Topic — the topic cluster this recommendation relates to, if applicable
  • Current performance indicator — for some recommendations, a progress bar showing your current score on the relevant metric
Recommendations are sorted by impact level (High first, then Medium, then Low), with a secondary sort by creation date.

​
Impact summary

At the top of the Feed, a summary bar shows the count of High, Medium, and Low impact recommendations currently in your feed, along with a link to the single highest-priority recommendation.

​
Filtering by category

If your feed contains recommendations across multiple categories, category filter pills appear above the list. Click a category to show only recommendations in that category. Click All to return to the full list.

​
Recommendation types

Topics or questions where competitors are being cited by AI platforms but your brand has no content coverage. Pharos identifies these from the response data it collects and suggests specific content to create or update.
Areas where you have some content but your brand rarely appears in AI responses for relevant queries. These recommendations typically involve improving, expanding, or better structuring existing content.
Pages on your site that are well-positioned to be cited by AI platforms but lack some technical signal (schema markup, extractable summaries, llms.txt coverage) that would help AI systems reference them.
AI-readiness improvements identified from your content health audits: fixing heading hierarchy, adding FAQ schema, allowing AI bots in robots.txt, or creating an llms.txt file.

​
Starting to track a recommendation

When you decide to act on a recommendation, click Track on its card or open it and click Start tracking. This moves the recommendation to the Tracking tab and sets its status to In progress. The Feed tab retains a preview list of untracked recommendations at the bottom of the Tracking tab so you can start tracking them without switching back and forth.

​
Opening the detail sheet

Click any recommendation card to open its full detail sheet. The sheet shows:
  • The full description and reasoning
  • Specific pages related to this recommendation and the issue on each page
  • Step-by-step implementation guidelines
  • AI-generated content (article outlines, code snippets, schema markup, or outreach templates, depending on the recommendation type)
  • Example URLs from competitors or high-performing sites for reference
  • Related prompts that would benefit from this improvement
From the detail sheet you can change the recommendation’s status, start tracking it, or generate a full article from the outline if one is available.

​
Dismissing recommendations

If a recommendation is not relevant to your brand, click Dismiss on the card or in the detail sheet. Dismissed recommendations move out of the feed. To view and restore dismissed recommendations, click the Dismissed filter pill that appears when you have dismissed items. Dismissed recommendations expire after 30 days.

​
Bulk actions

Select multiple recommendations using the checkboxes in the top-left corner of each card, then use the bulk action bar to:
  • Track all — start tracking all selected recommendations at once
  • Dismiss all — dismiss all selected recommendations
Bulk actions apply only to the recommendations visible on the current page. Items hidden by an active category filter are skipped.

​
Tracking tab

The Tracking tab contains all recommendations you have moved into active work. It shows recommendations with a status of In progress or Done.

​
Overall progress

At the top of the tab, a progress bar and stats card show how many tracked recommendations you have completed out of the total you are tracking. The breakdown includes:
  • Completed count
  • In progress count
  • Overall completion percentage

​
Working on a recommendation

Each tracked item shows its current status and the date you started working on it. Use the status controls on the card to advance through the workflow:
1

Start working

Click Start Working to move a recommendation from Pending to In Progress. This records the start date.
2

Mark complete

Click Mark Complete when you have implemented the recommendation. Pharos records the completion date and, if you had a visibility baseline when you started, captures your current visibility score to compute the impact delta.
3

Review impact

After marking complete, the card displays the before and after visibility scores and the delta. If the recommendation linked specific pages, a Re-audit pages action triggers a fresh content health audit on those pages to verify the improvements.

​
Impact measurement

When you start tracking a recommendation, Pharos records your current visibility score as the impact baseline. When you mark the recommendation as done, Pharos fetches your latest visibility score and computes the change. The delta appears on the completed card as a positive or negative percentage point change.
If no visibility data existed when you started tracking, the impact delta cannot be computed. The completed card will indicate this rather than showing a potentially misleading zero delta.

​
Reopening a recommendation

If you started tracking a recommendation but need to revise your implementation, click Reopen on the completed card. This returns the recommendation to Pending and clears the recorded impact delta so a fresh measurement can be captured when you complete it again.
Reopening a completed recommendation clears the recorded impact delta and it cannot be recovered. Confirm the reopen prompt before proceeding.

​
Bulk status updates

Select multiple tracked items and use the bulk action bar to:
  • Mark in progress — set all selected items to In Progress
  • Mark complete — complete all selected items at once
  • Reopen — return all selected items to Pending

​
Pending recommendations preview

If you are on the Tracking tab and have untracked recommendations in the Feed, a Pending section appears below your tracked list, showing up to five untracked recommendations with a Track button on each. Click View all pending recommendations to go back to the Feed.

​
Recommendation settings

Click the settings icon next to the Generate button to open Recommendation Settings. Use these settings to customize how Pharos generates recommendations — for example, specifying focus areas or content types you want to prioritize.