Skip to main content

Open in AthenaHQ

app.athenahq.ai/prompts?tab=discover

​
Purpose

This page displays the live progress and final results of a specific Prompt Discovery scan. It serves as the primary workspace where marketers can review the AI’s suggested prompts, organize them into logical topics, filter by intent or volume, and select the best candidates to permanently track and monitor. Rather than forcing users to blindly accept all suggestions, this page provides rich context (like search volume, funnel stage, and AI rationale) so customers can make informed decisions about which searches actually matter to their brand.

​
What’s on the page

Page Header Displays a back button to return to the discovery history and the name of the current discovery run. If the user has edit permissions, they can click the pencil icon to rename the run inline. Leg Progress Panel Appears only while the scan is still actively running. It displays a progress bar for each data source being analyzed (e.g., Website analysis, Keyword gaps) along with its status (queued, running, or done) and completion percentage. A note explains that suggestions will appear here as soon as any single source finishes. Discovery Goal Banner Once candidates are available and the run is no longer active, a banner appears summarizing the review state. It will either state how many suggestions are waiting for review or confirm that all suggestions have been reviewed, nudging the user to run discovery again in the future. Table Toolbar Sits just above the candidates table and houses all sorting and filtering controls:
  • Filters dropdown: Contains sub-menus to filter by top volume, source, funnel, intent, or keyword type.
  • Signal filter chips: Quick-toggle buttons for AI signals (like “Competitor Gap” or “Social Signal”). These only appear if the current run actually found prompts carrying these signals.
  • Reset: Clears all active filters.
  • Export: A download icon to export the current view to a CSV file.
  • Search: A text box to search for specific words in the prompt text or topic name.
  • Start tracking: The primary call-to-action that appears when one or more rows are selected. It shows the number of selected prompts.
Discovery Candidates Table spreadsheet-style table where the actual prompt suggestions live. Rows are grouped into collapsible “Topics” (e.g., “Competitors”, “Onboarding”).
  • (Select): A checkbox column to select individual prompts or entire topic groups for tracking. The header checkbox selects all currently filtered suggestions.
  • Prompt: The suggested search query. For topic headers, this shows the topic name (editable via a pencil icon), candidate count, and aggregate metrics. For candidate rows, it displays the prompt text, an optional colored signal badge, and an info icon ([i]) you can hover over to see exactly why the AI suggested it.
  • Source: Which discovery engine found this prompt. Tooltip: “How this prompt was discovered / Logical: market analysis from your website, segments and products / Brand: AI brand research with custom inputs / GSC: Google Search Console keyword clustering / Social: Reddit threads, YouTube comments and videos / Keyword sources: brand and competitor keywords, Reddit, YouTube”
  • Funnel: The buyer journey stage. Tooltip: “Where the prompt sits in the buyer journey / TOFU: top of funnel, awareness and discovery / MOFU: middle of funnel, evaluation and consideration / BOFU: bottom of funnel, decision and purchase”
  • Intent: The user’s underlying goal. Tooltip: “The searcher’s goal behind the prompt / Informational: looking to learn or research / Commercial: comparing products or services / Transactional: ready to buy or take action / Navigational: looking for a specific brand or page”
  • Type: The keyword category. Tooltip: “The keyword category for this prompt / Head: broad, high-volume keywords / Mid tail: moderate specificity and volume / Long tail: specific, lower-volume queries”
  • Volume: Tooltip: “Estimated monthly search volume”
  • Value: Tooltip: “Estimated keyword value (CPC)”
Already Tracked Section A collapsible section pinned to the very bottom of the table. It acts as a graveyard for suggestions the AI found but realized you are already monitoring.

​
What you can do here

  • Rename the run: Hover over the run’s name in the top page header and click the pencil icon to edit it. Press Enter or click away to save.
  • Rename a topic: Hover over any dark gray Topic row in the table and click the pencil icon. If you type the exact name of an existing topic, the system will prompt a confirmation dialog: “Merge into ‘<name>’?” to combine the two groups.
  • Expand or collapse topics: Click the arrow chevron next to a topic name to hide or show its prompts. You can also click the chevron in the top “Prompt” column header to expand or collapse all topics at once.
  • Select prompts for tracking: Check the boxes next to individual prompts, check a topic’s box to select all prompts inside it, or check the master box in the table header to select everything currently visible. (Tip: Hold Shift while clicking checkboxes to select a large range at once).
  • Start tracking: Once prompts are selected, click the blue “Start tracking (N)” button in the toolbar. This opens a confirmation dialog noting how many selections are new vs. duplicates. Confirming will add them to your permanent monitoring list.
  • Filter and search: Use the “Filters” dropdown, the signal chips, or the search bar to narrow down the list. Click “Reset” to clear them.
  • Sort the table: Click any column header (like Volume or Funnel) to sort the topic groups and the rows within them. Clicking repeatedly cycles between Ascending, Descending, and no sort.
  • Resize columns: Drag the invisible dividers between column headers to make columns wider or narrower. Your preferences will save to your browser.
  • Export as CSV: Click the download icon in the toolbar. This exports exactly what is currently filtered on your screen.
  • View Already Tracked: Click the “Already tracked” row at the bottom of the table to expand it. Hovering over the info icon here explains: “These prompts are already covered by your current monitoring. They matched existing prompts with high similarity (85%+), meaning the same search intent is already being captured.”

​
Data shown

The candidates are generated by the AI discovery scan against your website’s market context, competitor gaps, and optionally connected Search Console data. The table overlays estimated search volume and CPC data alongside the AI’s internal categorizations.

​
Common workflows

Review and track newly discovered prompts
  1. Open a discovery run from the Discover history list.
  2. If the run is still active, wait for sources to populate the table.
  3. Browse the candidate suggestions, grouped by topic.
  4. Use the signal chips or volume filters to focus on high-priority searches.
  5. Select the checkboxes for the candidates you want to monitor.
  6. Click Start tracking (N) in the toolbar.
  7. Review the counts in the confirmation dialog and confirm.
  8. Once successful, click View in Manage on the success dialog to see your newly tracked prompts.
Rename or merge a topic
  1. Scroll to the topic group row you want to change and hover over it.
  2. Click the pencil icon next to the topic name.
  3. Type the new name and press Enter.
  4. If the new name matches an existing topic elsewhere in the table, a dialog will ask you to confirm merging the two groups into one.
Export filtered suggestions to CSV
  1. Apply any desired filters (e.g., Intent: Commercial, Funnel: BOFU).
  2. Click the download icon in the top right of the toolbar.
  3. A CSV file will download containing only the rows and columns currently visible in your filtered view.

​
Empty, loading, and error states

  • Loading: A centered animated spinner displays while the run’s details fetch from the server.
  • Empty: If the run finished but found zero suggestions, a file icon appears with the text “No prompts discovered”. If the run actually failed or timed out, it adds “This run could not finish. Run discovery again later.” If you apply filters that hide every row, the table simply displays “No prompts found”.
  • Error: If the specific run ID doesn’t exist or failed to load entirely, a centered “Run not found” message appears.
  • Linked from: The main Discover history list (/prompts?tab=discover) when clicking on a specific past run.
  • Links to: The Prompts Manage view (/prompts) via the success dialog after tracking new prompts, and the Discover history list via the top back button.

​
Common support questions

Why can’t I click “Start tracking” or rename topics? You likely have “Viewer” or “Read-only” access to Prompts. You need Write access to permanently track new queries or edit topic names. Hovering over the disabled button will display a tooltip explaining the missing permission. Why did my CSV export prompt me to upgrade my plan? Exporting data from the app is a premium feature. If your organization is on a free plan, clicking the export icon will open a dialog outlining our paid subscription options instead of downloading the file. Why are some suggestions locked in the “Already tracked” section? The AI surfaces everything it finds relevant, but does a final check against the prompts you are already monitoring. If a newly discovered prompt shares 85%+ similarity with something you already track, it is safely moved to the “Already tracked” section so you don’t accidentally waste credits monitoring duplicates.