AI Visibility
Track how AI search engines mention, cite, and rank a brand across thousands of buyer-question prompts — with a single 0–100 score, drill-down into every answer, and a 4-week trend line.
What you can do
AI Visibility runs the same buyer-question prompts customers ask AI tools (ChatGPT, Claude, Gemini, Perplexity, Grok, Microsoft Copilot, Google AI Overview) and captures every brand mention, every cited source URL, and the sentiment in which the brand was discussed. The result is a single 0–100 visibility score per brand, recomputed nightly over a 4-week rolling window, with full drill-down into every scan answer.
Setting up a brand
- Add the brand. Pick a primary domain owned by the brand, optional associated domains (subdomains, regional variants), an industry category, an optional writing voice, an optional brand logo, plus the brand's default market and default answer language. When the primary domain has a country-coded TLD (like .co.il, .de, or .co.uk), the market auto-suggests from it; for generic TLDs (.com, .org) the default starts at the United States and English (United States).
- Configure prompts. Browse a curated catalog of 480 buyer-question prompts across 20 industries (in 8 languages), or write custom prompts, or generate prompts on demand. The AI generator reads an auto-extracted profile of what the brand actually does (built from the brand's primary domain) so the generated prompts target the brand's real product space, not just its name. A short manual description in the brand Settings tab overrides the auto-extracted profile when filled in.
- Add competitors. Auto-discovered on the first scan from SERP results plus an LLM fallback — review and confirm/dismiss the suggestions, or add competitors manually.
- Run a scan. Pick which prompts and engines to include, confirm or adjust the market and answer language (pre-filled from the brand's defaults), choose sample count, and submit. Scans take 1.5–3 minutes for the API engines plus ≈36 seconds per prompt for Microsoft Copilot's async scrape.
The visibility score
A single 0–100 number summarizing how visible a brand is across AI search engines for the configured prompts, geo, and language, over a 4-week rolling window. The score is shown as a circular gauge with a letter grade:
| Score | Grade |
|---|---|
| 80–100 | A |
| 60–79 | B |
| 40–59 | C |
| 20–39 | D |
| 0–19 | F |
The score is a weighted sum of 5 normalized metrics:
- Mention share (30%) — fraction of scan answers that mention the brand at least once.
- Citation count (30%) — citations to the brand's owned domains across all scan answers, normalized against the per-engine median.
- Sentiment (15%) — average sentiment of brand mentions, mapped from negative/neutral/positive to 0–1.
- Top position (15%) — how early in the AI's answer the brand is mentioned (top of answer scores higher than buried at the bottom).
- Engine coverage (10%) — fraction of configured engines that produced at least one brand mention or citation.
Top recommendations
The Overview tab opens with a Top Recommendations strip — prioritized actions the system surfaces based on the brand's current scan data. Each card carries a destination action button that jumps to the relevant tab pre-scoped to the right context (e.g., a specific source, prompt, or page). The same recommendations engine also feeds the Technical Health tab, scoped to technical-category findings.
Citations, mentions, and history
Beyond the score, every scan run has full drill-down:
- Citations — every URL the AI engines cited, with the engine, position rank, and the prompt the citation came from.
- Mentions — every brand mention extracted from raw answer text (target brand, competitors, and unrelated brands), with sentiment, position, and the AI's reasoning for that sentiment.
- History — one card per scan run with status mix (succeeded / pending / failed counts), total cost, prompts count, and engines used. Re-open any past run to inspect its citations + mentions.
Sources tab
The Sources tab aggregates every third-party domain cited in scan answers about a brand into a single ranked table. Each row carries an automatic classification (publisher, forum, marketplace, review, social, video, encyclopedia, academic, Reddit UGC, owned, competitor, other), an influence score that weighs citation count, engine diversity, and competitor overlap, plus filters by classification + four sort axes.
Clicking a row opens the drill-down modal: per-engine progress bars showing which AI engines cited the source, a competitor-overlap list (which competitors got cited from the same source), a Reddit-UGC subreddit breakdown when applicable, and the top prompts that triggered citations. Each top-prompt row exposes "Improve via Article" and "Improve via FAQ" CTAs that pre-fill the relevant generation flow.
Paid plans see a "Sources to influence" gap-opportunity panel above the main table — domains where competitors are cited but the brand is barely cited or absent (citation_count ≤ 2 with non-empty competitor overlap). Free-tier customers see a placeholder card describing the gap-analysis feature with a link to upgrade.
Save to Trusted Sources. Each source-domain card and the drill-down modal expose a Save action. Saved domains land on a per-brand shortlist (capped at 100, viewable and removable from the Brand Configure modal's Trusted Sources sub-tab). Saving a domain serves two purposes:
- Track partnership-priority domains — the saved list lets a customer keep a list of authority domains worth prioritizing for outreach, content placement, or partnership work.
- Influence AI article generation — when the brand owner writes a new article via the editor or AI Bots cron with the brand's domain selected as the publishing target, the saved Trusted Sources are sent to the AI as a soft preferred-citation hint. The AI decides whether to apply the hint based on topical relevance — a successful injection surfaces a confirmation toast at the end of the stream. This is a soft preference, not a forced citation.
Adding new entries is a paid-plan feature; viewing and removing existing entries works on every plan so a customer who downgrades can still clean up their list.
Content tab
The Content tab runs content audits against pages on the brand's own domain — surfacing AI-readability issues that hurt how AI engines understand and cite the page. Each audit produces actionable findings; AI-generated fixes can be reviewed inline and pushed back to WordPress with one click (for WordPress-connected sites).
For pages outside WordPress, the fix text is shown for manual application.
Technical Health tab
The Technical Health tab answers a different question from the strategic tabs: can AI engines actually reach the brand's pages? Strategic tabs measure how AI engines talk about the brand; Technical Health measures the plumbing that lets them get there in the first place.
Four KPI tiles summarize the state: Discoverability % (fraction of audited pages passing the AI-readability bar), AI crawler events (hits from AI bots like GPTBot, ClaudeBot, PerplexityBot on connected CDNs), Blocked requests (AI bots refused by robots.txt or firewall), and Open issues (audit findings awaiting fix).
A Domain Health card per audited domain surfaces robots.txt, sitemap, and canonical hygiene plus an AI-bot access matrix — which of the major AI crawlers (GPTBot, ClaudeBot, PerplexityBot, Google-Extended, Bingbot, and others) are allowed or blocked. The Audited Pages table below lists every crawled URL with its audit findings; click into a row for the per-page drill-down (shared with the Content tab).
AI crawler events are populated by CDN integrations — connect Cloudflare, Akamai, or Fastly via the CDN integrations page to fill the events and blocked columns. Pages can be (re-)audited from the Audit pages button on this tab; the audit shares the monthly Content audits cap with the Content tab so technical and content work draw from the same budget.
This tab works in "All brands" aggregate mode as well as brand-scoped, so a multi-brand operator can review domain hygiene across every account brand in one view.
Commerce tab
The Commerce tab is a product-catalog visibility tracker measuring where AI shopping engines surface the brand's products across retailer pages. Structured shopping signals (product cards, retailer carousels) come from Microsoft Copilot and Google AI Overviews — the engines that return shopping-formatted answers. The other AI chat engines (ChatGPT, Claude, Perplexity, Grok) don't return structured commerce data; their brand-mention signal flows through regular visibility scans and contributes to the per-product Visibility sub-tab. Four KPI tiles — Commerce Presence %, Average Share-of-Shelf %, Retailers Covered, Products Tracked — sit alongside a per-retailer breakdown and a per-product drill-down (Overview / Visibility / Edit sub-tabs).
The brand's product catalog is built from one of six feed sources, all configurable from the Product Feeds card on this tab:
- Paste URLs — manual list of product URLs
- Sitemap URL — regex-pattern product detection from a sitemap.xml
- WordPress auto — extends to WooCommerce products via the existing site assets cache
- Shopify /products.json — paginated import (capped at 4 pages × 250 items)
- Google Merchant XML feed — standard Google Shopping feed format
- CSV upload — documented 11-column template
Customer-edited fields on a product (title, keywords, image, retailer links) preserve across feed re-syncs via a 24-hour grace window. Commerce scans count against a separate plan-tier cap (independent of the visibility-scan cap) so commerce work doesn't compete with regular visibility budgets — the Run Scan modal carries a scan-kind toggle (Visibility / Commerce / Both) so you choose what to spend on.
Attribution tab
The Attribution tab is a roadmap placeholder for measuring downstream traffic and conversions that AI engines drive to the brand's sites. Planned signals include proxy-correlation with GA4, AI-referrer header tracking, and UTM correlation. The tab is visible on the dashboard but the metrics it surfaces are work-in-progress; full functionality lands in a future release.
Filtering across tabs
A global filter bar above the tab strip scopes every tab to a consistent slice of data — date range, AI engines, AI bots, and brand domain. Changing a filter on one tab updates the data on every other tab without a reload, so a question like "What's our visibility on ChatGPT in the last 7 days for our .co.il domain?" stays answered as you move between Overview, Citations, Mentions, and Sources.
Scheduling automatic scans
One brand maps to one schedule. Open the brand's Configure modal → Schedule tab and pick a cadence (weekly or monthly) plus the day-of-week and hour-of-day to fire. The scan runs the brand's persistent prompts, competitors, default market, and default answer language — matching what a manual "Run scan" produces — with results visible on the dashboard within minutes of completion. Set the brand's defaults in the Settings tab so every scheduled scan runs in the right market and language without per-run intervention.
Plan caps
Brand count, monthly AI Visibility spend, and daily scan count are all plan-tier capped. Junior Starter starts at 1 brand, $50/month spend cap, and 1 scan/day; Independent tiers scale from 10 brands / $250 spend cap / 15 scans/day (Starter) up to unlimited brands, scans, and spend (Advanced). Content audits use a rolling 30-day monthly cap (500–unlimited on Independent tiers). See Plans & usage limits.