Example report
What a Repliq audit looks like, applied to a real site.
Repliq audited figma.com as a demonstration, using the same method and the same deliverable format as for a white-label client. The full document (summary, score breakdown, AI crawler access, key findings and action plan) is available as a PDF.
Live search testing confirms Figma is largely absent from its own “alternatives” and comparison narrative, a space currently owned by a direct competitor, Penpot. Meanwhile, a three-page robots.txt allow-list cuts ChatGPT and Perplexity off from its richest content.
AI Citability Analysis Report
Search and Generative Engine Optimization Audit for Figma
Overall GEO Score
Good
Preview of page 1 of 12. The full report covers the six-dimension score breakdown, AI platform readiness, AI crawler access status, key findings and a prioritized action plan.
A second example is available: the Novotel audit (in French), or download its PDF directly.
Independent audit conducted by Repliq for demonstration purposes, based on public data collected on July 14, 2026. Repliq is not affiliated with Figma, Inc. Brands mentioned belong to their respective owners.
The full data behind this audit
Everything the PDF contains, in plain HTML: score, platforms, crawlers, findings and action plan.
Figma scores 73/100 on GEO, a rating driven by a striking split: Brand Authority is near-maximal at 94/100, anchored by Wikipedia, LinkedIn, YouTube, G2 and Capterra, while AI Citability and Platform Optimization lag at 58/100. The gap is structural rather than content-driven: ChatGPT's and Perplexity's live-citation bots are confined to a robots.txt allow-list covering essentially only the homepage, pricing and release notes, cutting them off from Figma's richest content. Live search testing also confirms Figma is largely absent from its own “alternatives” narrative, a space competitor Penpot currently owns. No critical issues were found; the prioritized quick wins could raise the score to approximately 85/100 within 30 days.
GEO score breakdown
| Component | Score | Weight |
|---|---|---|
| AI citability and visibility | 58/100 | 25% |
| Brand authority | 94/100 | 20% |
| Content quality and E-E-A-T | 73/100 | 20% |
| Technical foundations | 80/100 | 15% |
| Structured data | 68/100 | 10% |
| Platform-by-platform optimization | 58/100 | 10% |
AI platform readiness
| Platform | Score | Status |
|---|---|---|
| Google AI Overviews | 72/100 | Good |
| ChatGPT | 42/100 | Below average |
| Perplexity | 40/100 | Below average |
| Gemini | 68/100 | Moderate |
| Bing Copilot | 68/100 | Moderate |
AI crawler access
| Crawler | Platform | Status | Recommendation |
|---|---|---|---|
| GPTBot | ChatGPT (training) | Blocked | Deliberate IP-protection stance; revisit periodically as a business decision. |
| ChatGPT-User / OAI-SearchBot | ChatGPT (live search) | Restricted | Expand to pattern-based Allow rules covering /blog/, /customers/, /solutions/. |
| PerplexityBot | Perplexity | Restricted | Same allow-list expansion as the ChatGPT bots. |
| Claude-SearchBot | Claude (search) | Restricted | Same allow-list expansion as the ChatGPT bots. |
| anthropic-ai | Claude (training) | Blocked | Deliberate IP-protection stance; revisit periodically. |
| Google-Extended | Gemini (training) | Blocked | Doesn't affect current indexing; revisit for future model freshness. |
| Googlebot | Google Search / AI Overviews | Allowed | Keep allowed. |
| Bingbot | Bing Copilot | Allowed | Keep allowed. |
| CCBot | Common Crawl (training) | Blocked | Deliberate stance; affects models trained on Common Crawl. |
| ClaudeBot | Claude (live browsing) | Allowed | Falls under the permissive wildcard rule; keep allowed. |
Key findings
High
Live AI search bots confined to a 3-page allow-list
ChatGPT-User, OAI-SearchBot, PerplexityBot and Claude-SearchBot are restricted to essentially the homepage, /pricing/ and /release-notes/, so ChatGPT and Perplexity cannot retrieve Figma's deep feature pages or blog content when answering live queries.
High
No llms.txt file
figma.com/llms.txt returns a 404. Publishing one would provide a lower-maintenance, authoritative index of key content for AI systems.
High
Five major AI training crawlers fully blocked
GPTBot, anthropic-ai, Google-Extended, CCBot and cohere-ai all carry a blanket Disallow: /, excluding Figma's content from the training corpora of OpenAI, Anthropic, Google and Common-Crawl-derived models.
High
Organization schema missing Wikipedia, Wikidata and Crunchbase links
The homepage Organization JSON-LD links six social platforms via sameAs but omits the three highest-value entity-disambiguation signals for a well-known public company.
High
Blog author schema lacks credentials
Person objects in the BlogPosting schema carry name only (no sameAs, jobTitle or worksFor), even though the credentials are stated in the visible page text.
High
Product schema missing offers, blocking rich results
The /design/ and /figjam/ schema blocks lack the required offers property and use Product instead of SoftwareApplication, the type competitors like Canva and Miro use.
High
No first-party comparison content
Figma has no owned “Figma vs. [competitor]” or “Figma alternatives” content; Penpot currently owns and ranks the comparison narrative that AI systems cite instead.
Medium
Highly uneven content citability
The flagship blog research report scores 80 to 88/100 for AI citability, but the commercially important marketing pages average under 30. The pages most visitors land on are the least quotable.
Medium
Sitemap freshness signal is broken
Every sampled sitemap URL carries the same lastmod timestamp: the sitemap's build time rather than true per-page modification dates.
Medium
Fragmented company-history narrative
No single About/Company page exists; /about/ duplicates the homepage, and institutional facts are scattered across /newsroom/.
Medium
Flagship statistic is stale
The widely-used “95% of Fortune 500 uses Figma” claim is self-dated March 2025, roughly 16 months old as of this audit.
Medium
Zero structured data on blog listing, careers and legal pages
These page categories ship no JSON-LD at all, despite containing clearly structurable facts such as the company's legal address and open job postings.
Prioritized action plan
Quick wins (this week)
- Add Wikipedia, Wikidata and Crunchbase to the Organization schema's sameAsEffort : Low · Impact : High
- Publish an llms.txt curating the highest-citability contentEffort : Low · Impact : High
- Convert the AI-bot robots.txt allow-list to pattern-based rulesEffort : Low · Impact : High
- Enrich blog author Person schema with jobTitle, worksFor and sameAsEffort : Low · Impact : Medium
- Fix sitemap lastmod values to true per-page modification datesEffort : Low · Impact : Medium
Medium term (this month)
- Restructure the AI-search-bot allow-list to cover blog, customers and solutionsEffort : Medium · Impact : High
- Add hreflang annotations linking English pages to the 14 locale variantsEffort : Medium · Impact : Medium
- Publish 2–3 first-party “Figma vs. [competitor]” comparison pagesEffort : Medium · Impact : High
- Consolidate a single About/Company page with founding story and milestonesEffort : Medium · Impact : Medium
- Add SoftwareApplication schema (with offers) to product pagesEffort : Medium · Impact : High
Strategic initiatives (this quarter)
- Run a recurring first-party comparison and “alternatives” content programEffort : High · Impact : High
- Adopt a formal AI-crawler access policy separating training from retrievalEffort : Medium · Impact : High
- Publish editorial standards and methodology for the original researchEffort : Medium · Impact : Medium
- Build a dedicated customer case-study hub to strengthen Experience signalsEffort : High · Impact : High
- Periodically revisit the blanket blocking of AI training crawlersEffort : Low · Impact : Medium
Your clients have the same gaps, without knowing it.
Repliq delivers this level of detail for every audit, under your brand.