AI Search Performance Report (Public)

​

Purpose

This page is the public, read-only AI Search Performance Report for a single company pitch. It bundles every metric Athena tracks for that company. Share of Voice, brand mention rate, model-by-model performance, citation sources, competitor breakdown, brand-trait comparisons, and the underlying individual AI responses: into one shareable URL that anyone with the link can open. The report exists so a customer (or someone the customer is pitching, like a prospective client or stakeholder) can see how a brand performs across AI search platforms. ChatGPT, Claude, Perplexity, Gemini, Google AI Overview, Copilot, AI Mode, Grok, and DeepSeek: without needing an Athena login. It sits at the end of the Pitch Workspace flow: customers configure a pitch internally, run it, and then share this report URL externally. For pitches that came in through the public landing-page flow, the page also doubles as a sales surface: it includes a CTA banner inviting the viewer to see pricing or book a demo. For pitches created inside an authenticated Pitch Workspace, the same data is visible but the CTA banner is hidden. Because the URL is fully public, the page is intentionally simple: no filters beyond a response search box, no editing, no destructive actions: just metrics, charts, and a way to read each individual model response in detail.

​

What’s on the page

​

Page header

A sticky header showing the company logo (with a fallback initial if the logo isn’t available) and the company name. On the right are the Copy Link and Share buttons. On small screens those two collapse into a single ⋯ (More) menu.

​

Report header

The title “AI Search Preview Report,” a “Powered by AthenaHQ” link to athenahq.ai, and an “Audited [date]” line showing when the pitch was run.

​

Metric cards row

Two small summary cards sit side-by-side at the top:

• Share of Voice: the percentage of all brand+competitor mentions that belong to the target brand. The info tooltip reads: “The percentage of mentions your brand receives compared to all brands.” Clicking Learn more opens a modal with the plain-English definition, the formula (rendered in math notation), and bullets explaining why SoV matters.
• Brand Mentions: the percentage of all AI responses (across all models and prompts) that mention the brand at all. The info tooltip reads: “Percentage of AI responses that mention your brand.” Its Learn more modal explains Brand Mention Rate with the same formula + insights pattern.

​

AI Search Share of Voice (SOV) chart

A donut chart with a numbered legend ranking the brand and each competitor by their share of mentions. Hovering a slice shows the share-of-voice percentage and the absolute mention count. The brand’s slice is always indigo; competitors rotate through other colors. If there’s no SoV data yet, a blurred placeholder version of the chart is shown with a “No share of voice data available” message.

​

AI Model Performance card

Displays the overall mention rate (one big number: % of all responses that mention the brand) and a horizontal segmented bar showing mention rate by AI model. Below the bar is a legend with each model’s name and its individual mention-rate percentage in parentheses. The first four models are color-coded indigo / pink / purple / emerald; any beyond that are gray.

​

Positive Brand Traits / Negative Brand Traits cards

Two radar (spider) charts sitting side-by-side. Each axis is a single attribute name (e.g., “Reliable,” “Easy to use”), and each chart compares the target brand against “All Competitors” pooled together. Positive traits use indigo + gray; negative traits use orange + gray. Up to 8 attributes are shown per chart, sorted by total mentions. If there are no positive (or negative) traits at all, that card is hidden, not just empty, so the page may show one, both, or neither of these cards.

​

Top Citation Sources card

A horizontal progress-bar list of the top 10 domains cited as sources in AI responses, ranked by mention count. Each row shows the domain name, the mention count, and a bar whose width is proportional to the top source’s count. Google Places sources are labeled google.com (Google Places) to distinguish them from regular google.com citations.

​

Competitor Analysis card

A simple table listing every tracked competitor with their share of total mentions. Columns:

• Company: competitor name with logo. Long names are truncated; hover shows the full name.
• Mention %: percentage of all combined brand+competitor mentions attributed to this competitor.

​

Responses Table card

The core table at the bottom of the page: one row per AI response. Above it sit small icons showing which AI models contributed responses (active models are full-color, models that produced nothing are dimmed). On the right is a Search responses input. Columns:

• Model: colored badge showing which AI model produced the response (ChatGPT, Claude, Perplexity, Gemini, AI Overview, Copilot, AI Mode, Grok, DeepSeek).
• Mentioned. Yes / No badge indicating whether the target brand was mentioned in the response.
• Competitors Mentioned: stacked logo icons of competitors detected in the response. Hover shows competitor names; if more than 4, a +N badge expands to a tooltip listing the rest.
• Prompt: the specific prompt variation that was sent. Search-query matches are highlighted in yellow.
• Response: truncated preview of the raw response text with search-query matches highlighted. Click the row to open the full response in a drawer.

A footer line shows “Showing X of Y responses” plus a Load More button.

​

Landing Page CTA banner (conditional)

Only appears when the pitch was created from Athena’s public landing-page flow. Headline: “Track this brand across every AI surface,” subtext mentions continuous monitoring across all 9 supported models. Two buttons: See pricing (links to athenahq.ai/pricing) and Book a demo (opens the Athena Calendly). This banner is hidden for pitches created inside the Pitch Workspace.

​

Response Detail drawer

Opens when a row in the Responses table is clicked. Slides in from the right and shows:

• Top metadata grid (3×2):
  • Date: date the response was collected.
  • Model: colored badge of the AI model.
  • Prompt Type. “Non-Branded” (Discovery) or “Branded” (High-Intent).
  • Mentioned. Yes / No / N/A badge for whether the brand was mentioned.
  • Market Position: the brand’s rank in the response (1, 2, 3…) or None / N/A.
  • Competitors Mentioned: pill list of competitor logos+names mentioned in the response.
• Base Prompt: the original prompt that was sent.
• Variation: the prompt variation, shown as a chat-style bubble on the right.
• Response body: full markdown-rendered response. Brand mentions are highlighted indigo; competitor mentions are highlighted orange. For Google AI Overview responses where no overview was shown, displays “No AI Overview was shown for this query.”
• Sources: list of cited source pills, each with citation number [n] (or [-] if not directly referenced in the text), favicon, and domain. Sources from the brand’s own root domain are colored indigo; competitor sources are orange; everything else is gray. Clicking a source opens it in a new tab.
• AI Overview only. Organic Results (collapsible): list of organic results returned alongside the AI Overview, with position, title, snippet, and displayed link. Results from the brand’s domain are highlighted indigo; competitor-domain results are orange.
• AI Overview only. People Also Ask (collapsible): the question, snippet, and source link for each PAA item.
• AI Overview only. Related Searches (collapsible): pill list of related search queries, each linking to the search.

​

What you can do here

• Copy Link (top-right): copies /report/[pitchId] (the full report URL) to your clipboard. The button briefly animates to a checkmark.
• Share (top-right): on devices that support native sharing (mostly mobile), opens the OS share sheet pre-filled with the report URL and a subject line. On desktop, falls back to copying the URL.
• More menu (⋯, mobile only): on small screens the Copy Link and Share buttons collapse into this dropdown.
  • Copy Link (inside ⋯): same as the desktop button.
  • Share (inside ⋯): same as the desktop button.
• Learn more (Share of Voice card): opens a modal explaining the SoV metric.
• Learn more (Brand Mentions card): opens a modal explaining the Brand Mention Rate metric.
• Search responses: filters the Responses table client-side by prompt or response text. While a search is active, infinite scroll is disabled: search only matches against responses already loaded.
• Click a Responses table row: opens the Response Detail drawer.
• ↑ / ↓ arrows in drawer header: navigate to the previous / next response (also works with keyboard arrow keys while the drawer is open).
• ✕ in drawer header: closes the drawer.
• Load More (Responses footer) or scrolling near the bottom of the table: fetches the next 50 responses.
• Organic Results / People Also Ask / Related Searches (drawer, AI Overview only): click each section header to expand or collapse.
• See pricing (landing-page CTA only): opens athenahq.ai/pricing in a new tab.
• Book a demo (landing-page CTA only): opens the Athena Calendly booking page in a new tab.
• Powered by AthenaHQ (report header): opens athenahq.ai in a new tab.
• Source pills, organic result links, PAA links, related search links (all in drawer): open the corresponding URL in a new tab.

There are no destructive actions, no exports, no kebab/overflow row actions, no bulk-edit toolbars, and no card-reordering on this public page. (Inside the authenticated Pitch Workspace the same report supports drag-and-drop card reordering, but on the public /report/[pitchId] URL it is disabled.)

​

Data shown

Everything on the page is derived from the data Athena collected for this specific pitch:

• The list of prompts and prompt variations configured for the pitch.
• AI responses generated by running those prompts against ChatGPT, Claude, Perplexity, Gemini, Google AI Overview, Copilot, AI Mode, Grok, and DeepSeek.
• Mention detection results: whether the brand was mentioned in each response, what rank, and which competitors appeared.
• Sentiment classifications (positive / neutral / negative) attached to each response.
• Citation sources extracted from each response.
• Brand attributes (positive and negative traits) tagged to the pitch and the responses.
• Competitor records associated with the pitch.
• Logos for the brand and competitors, fetched from Athena’s logo storage.

If the pitch is still being processed when the report is opened, the page auto-refetches every 30 seconds so visitors see new responses appear live until the run finishes.

​

Common workflows

View the full report

1. Open the /report/[pitchId] URL.
2. Wait for the loading animation (the indigo Athena temple logo) to finish.
3. Read the metric cards, then scroll through SoV, model performance, brand traits, citation sources, and competitor analysis.
4. Scroll to the Responses table to see the underlying data.

Read individual AI responses

1. Scroll to the Responses table.
2. Optionally type into Search responses to filter.
3. Click any row to open the Response Detail drawer.
4. Use ↑ / ↓ in the drawer header (or arrow keys on the keyboard) to step through responses.
5. Click ✕ to close.

Share the report with a stakeholder

1. Click Copy Link in the page header.
2. Paste the URL into email, Slack, etc.
3. On mobile, tap ⋯ → Share to use native sharing instead.

Load all responses for review

1. Scroll to the bottom of the Responses table.
2. The next 50 responses load automatically as you reach the bottom.
3. Or click Load More in the table footer.
4. Repeat until “Showing X of X” matches and the Load More button is gone.

Understand a metric

1. Click Learn more on the Share of Voice or Brand Mentions card.
2. Read the definition and formula in the modal.
3. Click outside or press Escape to close.

​

Empty, loading, and error states

• Loading (full page): a centered, animated AthenaHQ temple logo (indigo, with a gentle pulse) is shown while the initial data loads.
• Loading (per card): each card shows a skeleton shimmer in place of its content until that section’s data resolves.
• Empty Responses table: “No responses found.”: typically only seen if a search query has filtered everything out.
• Empty SoV chart: a blurred placeholder version of the donut is shown with the message “No share of voice data available.”
• Empty Brand Traits: the Positive and/or Negative Brand Traits cards are hidden entirely if no attributes have been recorded: they don’t render an empty state.
• Pitch not found: the page returns a standard 404 not-found page.
• AI Overview without overview: the drawer shows “No AI Overview was shown for this query” instead of the response body.
• Failed to load more: a toast appears reading “Failed to load more responses. No more attempts will be made.” and infinite scroll stops.
• Failed to copy link: a toast appears reading “Failed to copy link.”

​

Linked from / links to

Linked from

• Anyone who has the /report/[pitchId] URL: sent via Copy Link, Share, email, Slack, embedded in a deck, etc.
• The authenticated Pitch Workspace, where customers generate and share this URL.

Links to

• athenahq.ai (Powered by link in the report header).
• athenahq.ai/pricing (See pricing CTA: landing-page pitches only).
• cal.com/team/athenahq/athena-intro (Book a demo CTA: landing-page pitches only).
• Individual cited source URLs (from the drawer Sources section).
• Organic result URLs, People Also Ask source URLs, and Related Search URLs (drawer, AI Overview responses only).

​

Common support questions

“Do I need to log in to view this report?” No. Reports at /report/[pitchId] are fully public: anyone with the URL can open them. That’s by design so customers can share reports with stakeholders who don’t have Athena accounts. “Can I rearrange the cards on the report I shared?” Not on the public report URL. Card reordering is only available inside the authenticated Pitch Workspace; the public page intentionally locks the layout. “Why don’t I see Positive (or Negative) Brand Traits?” Those cards are hidden when no attribute data exists for the pitch. If your pitch hasn’t generated any positive or negative trait mentions, that card simply won’t render. “What does the [-] in front of a source mean?” It means that source is one of the citations Athena collected for the response, but the response text didn’t actually reference it with a numbered citation marker like [1] or [2]. Sources with [1], [2], etc. are explicitly cited inline in the response. “My search isn’t finding a response I know exists. Why?” The search box only filters responses that have already been loaded into the table. If you’ve only loaded the first 50 of 200 responses, search won’t see the remaining 150. Click Load More (or scroll to the bottom) until everything is loaded, then search. “What’s the difference between Share of Voice and Brand Mentions?” Brand Mentions is the percentage of all responses that mention your brand at all (out of every prompt × model combination). Share of Voice is the percentage of all brand+competitor mentions that belong to your brand specifically: it ignores responses where nobody was mentioned and tells you how dominant your brand is relative to competitors. Both metrics have a Learn more modal with the formula. “The data on the page is changing while I look at it. Is something wrong?” No, that means the pitch is still processing. While a pitch is running, the page auto-refreshes every 30 seconds so you see new responses appear in real time. Once the pitch is fully complete it stops auto-refreshing. “Why is there a ‘See pricing’ / ‘Book a demo’ banner on this report but not another one?” The CTA banner only appears on reports that originated from Athena’s public landing-page flow. Reports created inside a paid Pitch Workspace don’t show it.