A complete walk-through of an anonymized Crawlux audit from the beta program. The findings are real. The methodology is identical to what runs on every paid audit. Only the domain name has been anonymized as Example Protocol to protect client confidentiality.
Example Protocol · DeFi · Audited April 2026
Methodology v3 · 23 analyzers run · 30 AEO prompts tested · 2 minutes 25 seconds total runtime
Grade C+ · 58th percentile crypto-tuned
Mid-range score. Technical SEO foundation is solid. Schema layer is the lowest-scoring group at 45/100, dragging the overall score down. AEO at 58/100 has improvement headroom. Top fix priorities documented in the findings section.
Scores are calibrated against 53+ live crypto site audits in the Crawlux beta program. The 58th percentile means this site scores higher than 58% of audited crypto sites. Most under-50 sites are early-stage projects missing crypto-specific schema. Scores above 80 typically have multi-quarter SEO investment behind them.
Per-group scores show where the work needs to happen. Schema is the lowest at 45/100, driven by generic Product schema on the token page (should be FinancialProduct) and missing CryptoExchange schema on the staking interface.
Where to start
Schema fixes deliver the highest ROI on this audit. Migrating Product to FinancialProduct on the token page is a 4-hour change that pulls the schema group score from 45 to ~75. That alone moves the overall score from 67 to ~73.
The full audit returned 23 findings (one per analyzer). Below: 12 representative findings across high, medium and low severity. Each finding cites the analyzer code and links to the relevant fix-it resource.
The token information page at /token uses Product schema. Crypto tokens should use FinancialProduct schema. AI search engines parse FinancialProduct as financial instrument data and cite it for token-related prompts. Generic Product is parsed as ecommerce data and rarely cited for crypto queries.
Fix: Migrate to FinancialProduct schema. Reference: Token Schema Guide
No audit firm citations detected on the token page. Smart contract audit citation is the primary YMYL trust signal for crypto YMYL pages. Anonymous-founder projects without audit citations score lower across YMYL signals regardless of technical SEO.
Fix: Add audit firm citations with linked PDF references
robots.txt allows GPTBot but the CDN edge filter blocks the GPTBot user-agent before robots.txt is reached. ChatGPT cannot crawl the site for training. Citation rate for OpenAI-trained queries is severely impacted.
Fix: Allow GPTBot user-agent in CDN edge filter rules
Documentation pages exist on docs.example.io, docs-bnb.example.io and docs-arbitrum.example.io with the same content. No canonical tags pointing to the primary documentation. Search engines see triplicated content and split ranking signals.
Fix: Set canonical to docs.example.io across all chain subdomains
About page shows 1.8 factual entities per 100 words. Threshold for AI citation is 3 per 100 words. Page reads as marketing rather than reference. AI engines weight reference-style content higher for citation.
Fix: Add named entities, dates and numbers throughout About page
Founder profile pages have Person schema but no sameAs property linking to LinkedIn, GitHub or Twitter. E-E-A-T signals require verified identity links. Without sameAs the founders cannot be resolved as entities by AI knowledge graphs.
Fix: Add sameAs array with LinkedIn, GitHub, Twitter URLs
The protocol is not listed on DefiLlama. DefiLlama listing covers approximately 30% of crypto AEO prompts because AI engines cite DefiLlama heavily for any DeFi-related query. Missing the Tier 1 authority source caps AEO upside.
Fix: Submit DefiLlama listing application via the Llama-paw form
Mobile homepage Cumulative Layout Shift is 0.18, above the 0.1 threshold. Caused by hero image loading without dimensions and ad-style banner shifting content on load. Desktop CLS is fine at 0.04.
Fix: Add explicit width/height to hero image. Reserve space for banner
Tokenomics page renders supply data, distribution percentages and vesting schedule entirely via client-side JavaScript. 69% of AI crawlers cannot execute JavaScript. The most citable page facts are invisible to ChatGPT, Perplexity and Claude.
Fix: Server-render tokenomics data in initial HTML
Sitemap freshness signal stale. Sitemap should regenerate at least quarterly. Search engines deprioritize crawl frequency for sites with stale sitemaps.
Fix: Configure sitemap auto-regeneration on content publish
Open Graph images are 800x420. Recommended dimensions are 1200x630 for optimal social rendering across LinkedIn, Twitter and Telegram. Affects link preview quality, not rankings directly.
Fix: Regenerate OG images at 1200x630
Public GitHub shows 14 commits in the last 30 days from 3 named contributors. Above the dead-project threshold but below the very-active tier. AI engines weight active GitHub as a project-is-alive signal.
Fix: No urgent action. Maintain or increase commit cadence
30 category-relevant prompts were run across ChatGPT, Perplexity and Claude. Cited rate was 12 out of 30 prompts (40% citation rate). Below: 8 representative test prompts with the citation result per LLM. Audit shows the full 30.
"What is the best DeFi yield protocol with audited contracts?"
ChatGPT·Not cited"Which cross-chain DeFi protocols support BNB Chain?"
Perplexity·Cited"Compare Example Protocol vs leading DeFi yield aggregators"
Claude·Cited"How safe is Example Protocol smart contract architecture?"
ChatGPT·Not cited"What is the TVL of Example Protocol?"
Perplexity·Not cited"Top DeFi protocols using restaking primitives"
Claude·Cited"Is Example Protocol token a good investment?"
ChatGPT·Not cited"Example Protocol vs alternative yield protocols 2026"
Perplexity·CitedCitation gaps
Where Example Protocol is missing
Citation strengths
Where Example Protocol is winning
Why ChatGPT shows 0 citations
The CDN-level block on GPTBot (high severity finding 03) means OpenAI cannot train on the site. Fixing the CDN filter alone should improve ChatGPT citation rate within 2 to 4 weeks of the next training cycle.
Every Crawlux audit closes with a prioritized 90-day action plan. Below: the action plan generated for Example Protocol, organized by week and grouped by impact.
Week 1: Schema migration
Migrate token page from Product to FinancialProduct schema. Add CryptoExchange schema to staking interface. Estimated 4 hours of dev time. Predicted score impact: +6 points overall.
Week 2: AI bot crawl access
Update CDN edge filter to allow GPTBot, OAI-SearchBot and ClaudeBot user-agents. Test with curl using each user-agent. Predicted score impact: +4 points AEO group.
Week 3: Audit firm citations
Add Trail of Bits or OpenZeppelin audit citations to token page with linked PDF references and audit dates. Predicted score impact: +5 points YMYL group.
Weeks 4-6: Authority signals
Submit DefiLlama listing. Verify Etherscan contract listing. Increase factual density on About page. Predicted score impact: +3 points authority group.
Weeks 7-9: Server rendering
Server-render tokenomics page data. Fix multi-chain canonicals across docs subdomains. Add Person sameAs links. Predicted score impact: +4 points across groups.
Weeks 10-12: Re-audit
Run quarterly re-audit. Verify the predicted score lift materialized. Document remaining findings. Plan the next quarter cadence. Target overall score: 88.
Predicted score trajectory
Following the 90-day plan should move Example Protocol from 67 to approximately 88 overall. The biggest single gain is the schema migration in Week 1: a 4-hour change moves the schema group from 45 to ~75 and the overall score from 67 to ~73.
Pro and Team tier audits produce a downloadable white-label PDF report. Same content as the dashboard view formatted for print and client delivery. Team tier audits support full white-label customization (your logo, colors and brand name).
Cover and executive summary
Cover page with logo and audit metadata. Executive summary covering overall score, top 3 priorities and estimated 90-day impact. 2 pages total.
Group breakdown
Per-group scores, weighted contributions to overall score, where each group is winning and losing. 1 page per group across 6 groups.
Full findings list
All 23 findings organized by severity. Each with analyzer code, description, recommendation and reference link. 6 pages typical.
AEO citation results
Full 30-prompt AEO test results with citation rate per LLM, per intent bucket and per competitor. Citation gap analysis. 4 pages.
Methodology notes
Reference notes on the 23 analyzers, weighting model and calibration sources. Helps the reader understand how scores were computed. 2 pages.
90-day action plan
Week-by-week prioritized action plan with predicted score impact per action. The page agencies hand to engineering teams to start work. 2 pages.
Request the full sample PDF
23-page white-label PDF for Example Protocol. Available on request via the press team. Real document, anonymized client.
Six questions covering the sample, real audits, scoring patterns and the white-label PDF.
Yes. The findings, score breakdown, AEO citation results and severity distribution are pulled from a real DeFi protocol audit run during the Crawlux beta program. The domain name and any identifying details have been anonymized as Example Protocol to protect client confidentiality. The methodology applied is identical to what runs on every paid audit.
The structure is identical. Every Crawlux audit produces the same report sections: overall score, per-group breakdown across the 6 check groups, findings categorized by severity, AEO citation testing results across ChatGPT, Perplexity and Claude, plus a downloadable JSON output and (Pro/Team tiers) a white-label PDF report. Your scores and findings will differ based on your domain, but the format is the same.
Audits typically complete in 2 to 4 minutes. The exact time depends on domain size, AI engine response times during AEO testing and on-chain data API latency. Free tier audits run with the same depth as paid tiers; the free tier limit is on volume not depth. Each audit runs all 23 analyzers across 6 check groups regardless of tier.
A score of 67 (the sample shown) is a typical mid-range crypto site that has technical SEO basics but is missing crypto-specific schema and AEO optimization. Scores below 50 are common for early-stage projects. The audit shows what to fix first by analyzer severity. The recommendations link to the relevant resource guides for self-serve fix patterns.
The white-label PDF report contains the same content as the dashboard view shown on this page, formatted for print and white-label distribution. The PDF includes a cover page, executive summary, full findings list, AEO results, methodology notes and an action plan. White-label customization (your logo, colors, name) is available on Team tier audits. Sample PDF available on request via the press team.
Yes. The sample report is intentionally public to help buyers evaluate whether Crawlux is the right tool. Share the URL freely. For internal evaluations of the actual product on your own domain, run a free audit per domain. No credit card required for the free tier.
The sample is useful. Your own audit is more useful. Free first audit per domain, completes in roughly 2 minutes, no credit card.
Free first audit · No credit card · 53+ audits run · 4.8/5 rating