> ## Documentation Index > Fetch the complete documentation index at: https://docs.athenahq.ai/llms.txt > Use this file to discover all available pages before exploring further. # Responses > Browse, search, and filter every individual AI-model response collected for the website, with a drawer to inspect and edit each response's details. `app.athenahq.ai/responses` ## Purpose The Responses page allows you to browse, search, and filter every individual AI-model response collected for your tracked website. It provides a granular look at exactly what AI models are saying in response to your prompts. This page is essential for diagnosing why certain metrics might be shifting. You can dive deep into the raw text of responses, inspect which competitors and sources are being cited, and verify sentiment and market position. The page also features a slide-out drawer where you can inspect, edit, or flag specific response details. ## What's on the page ### Page header Shows the "Responses" title. If you navigated here from a specific prompt or drilldown, a Back button will appear so you can easily return. The far right of the header hosts action buttons like Columns and Export on desktop, or inside a mobile menu. ### Filter row A sticky bar containing your active filter chips (like Date Range, Models, or Sentiment) alongside an "Add filter" button to attach more criteria. This row also houses the Saved Views dropdown to manage your filter presets, and a "Clear All" button if any non-default filters are applied. ### Responses table A spreadsheet-style table displaying every response row. You can resize, reorder, and sort columns. ### Footer bar Located at the bottom of the table, showing your current viewing progress (e.g., "Showing 40 of 1,234 responses") alongside a "Load More" button to fetch the next page of data. ### Response detail drawer A side drawer that opens when you click a row in the table (or navigate using arrow keys). It displays the full response metadata, competitors, sources, attributes, personas, locations, and the raw text. It also contains controls to edit or flag the response. ### Export dialog A modal that appears when you click the Export button, allowing you to choose how many responses to export to a CSV file. It includes custom amount inputs, quick-select presets, and an "Export all" option, along with an estimated time to completion. ## What you can do here * **Back:** Click the back arrow in the page header to navigate back to your safe return location (e.g., returning to a specific prompt's drawer). * **Add filter:** Click "Add filter" in the filter row to open a menu and select a new filter criteria. * **Clear All:** Click "Clear All" in the filter row to reset all non-default filters back to their original state. * **Saved Views dropdown:** Click the "Views" dropdown to save your current filter configuration as a named view (with an optional icon), load a previously saved view, or update/delete an existing view. * **Columns:** Click the "Columns" button in the top right to open a popover/drawer that lets you toggle column visibility and drag to reorder them. * **Export (download icon button):** Click the download icon in the top right to open the Export Responses dialog to configure and download a CSV of your filtered responses. *Note: Data export requires a paid subscription plan.* * **Load More:** Click "Load More" at the bottom of the table to fetch and append the next page of responses. * **Click a table row:** Click any response row to open the Response Details drawer. This will also update the URL with a `?responseId` parameter so you can share the link, and it prefetches adjacent rows for faster navigation. * **Arrow Up / Arrow Down (while drawer open):** Use your keyboard's up and down arrow keys to quickly navigate between response rows without closing the drawer. * **Drawer: Edit response (pencil icon):** Inside the drawer, click the pencil icon to open the Edit Response dialog. Here you can manually adjust whether the brand was mentioned, rank, sentiment, and competitor mentions. *Disabled for read-only users.* * **Drawer: Flag response (flag icon):** Inside the drawer, click the flag icon to open the Flag Response dialog to report an issue with the response content. *Disabled for read-only users.* * **Drawer: Previous/Next response (chevron buttons):** Inside the drawer, click the up or down chevron buttons in the top right to move to the adjacent row. * **Drawer: Close (X):** Click the X or click outside the drawer to close it and clear the `?responseId` parameter from the URL. * **Export dialog: quantity +/- buttons and quick-select (1,000 / 5,000 / 10,000):** Adjust how many responses will be included in your CSV export. * **Export dialog: Reset to 1,000 / Export all (N):** Instantly set the export quantity to the default 1,000 or to the full available result count. * **Export dialog: Export:** Click to trigger the CSV export. A progress toast will appear while it compiles, followed by a success or error message. * **Export dialog: Cancel:** Closes the dialog without starting the export. * **Mobile ⋯ menu: Columns (submenu):** On mobile devices, tap the three-dot ⋯ menu and select "Columns" to toggle column visibility from a checklist. * **Mobile ⋯ menu: Download:** On mobile devices, tap the three-dot ⋯ menu and select "Download" to open the Export Responses dialog. ## Data shown This page displays your tracked AI-model responses pulled directly from our analytics systems. It also utilizes your organization's configured prompt tags, personas, locations, and brand/competitor logos to enrich the response data. ## Common workflows ### Investigate a prompt's AI responses 1. Arrive from a specific prompt's analytics drawer where you clicked to view responses. 2. The table loads automatically scoped to that single prompt. 3. Click any row to open the drawer and read the full response text and metadata. 4. Use the Back button in the page header to return to the originating prompt drawer when you are finished. ### Find responses missing brand mentions but mentioning competitors 1. Click "Add filter" and set the **Brand Mentioned** filter to "No". 2. Add the **Competitors** filter and select your relevant competitors. 3. Scan the table for rows with the "No" Mentioned badge highlighted in orange alongside a populated Competitors Mentioned column. 4. Click a row to open the drawer and inspect the full text to understand why the AI favored competitors. ### Export a filtered set of responses to CSV 1. Apply your desired filters (e.g., date range, specific models, sentiment). 2. Click the Export (download icon) button in the top right. 3. Adjust the export quantity in the dialog (or choose "Export all" if you want the entire dataset). 4. Click "Export" and wait for the progress notification to complete your CSV download. ### Edit or flag a response 1. Click a response row in the table to open the details drawer. 2. Click the pencil icon to open the Edit Response dialog to adjust mention status, rank, sentiment, or competitor mentions, then save your changes. 3. Alternatively, click the flag icon to open the Flag Response dialog and submit a reason if the response data appears broken or inaccurate. ### Save a customized filter view 1. Configure several filters (like specific models, a 30-day date range, and negative sentiment). 2. Open the "Views" dropdown in the filter row and choose to save the current view, assigning it a name and optional icon. 3. In the future, open the same dropdown to load that view and instantly restore all your configured filters. ## Empty, loading, and error states * **Empty:** If no responses match your current filters, a "No responses found." message is displayed in a single centered row in the table. * **Loading:** While responses or dynamic filters are loading, an animated spinner appears in place of the table rows. The table headers remain visible. * **Error:** Fetch failures, such as an export failing, surface via toast notifications in the corner of the screen. There is no dedicated inline error state on the table itself. ## Linked from / links to * **Linked from:** * The Prompts page's per-prompt drawer (via "View responses" drilldown links). * The Sankey, Olympus, or Content dashboards via drilldown links. * Guided product tours (specifically the "prompt-detail" tour sequence). * **Links to:** * A Prompt's analytics drawer (via the return/back button when arriving from the Prompts page). * CSV file downloads (client-triggered, not a separate page). ## Common support questions **Why is the "Mentioned" badge orange?** The "Mentioned" badge turns orange with a "No" when your brand was omitted from the response, but one or more of your tracked competitors *were* mentioned. It highlights missed opportunities where AI recommended alternatives instead of you. **I adjusted my columns. Will they stay that way?** Yes! Your column visibility, order, and customized widths are saved to your browser automatically. They will persist the next time you visit the Responses page on that same device and browser. **Why does my CSV export have a maximum limit?** Exports are capped by default to ensure fast downloads and prevent browser timeouts on massive datasets. You can adjust the limit inside the Export dialog, or select "Export all" if your dataset isn't overwhelmingly large. **Why can't I edit or flag a response?** If the pencil and flag icons in the drawer are disabled, your account likely has read-only/viewer permissions for this website. Only users with admin or editor permissions can alter response data. **If I double-click a column edge, what happens?** Double-clicking the resize handle between two column headers will automatically snap that column's width to fit its longest visible content.