Primary Industry Datasets
Original Empirical Research
We crawl and analyze large production cohorts to publish original findings that AI search competitors cannot fabricate. Verifiable methodology, open datasets, and actionable engineering takeaways.
Original Data ReportSample: 100 live production domains (91 valid)
100 Websites Audited: What Real-World Data Reveals About Speed, Schema, and AI Bots
An empirical forensic benchmark of 100 live websites measuring server latency, mobile LCP bottlenecks, Knowledge Graph adoption, and llms.txt AI discovery readiness.
76.9%Mobile LCP Failure Rate70 of 91 valid domains exceeded Google 2.5s mobile LCP benchmark
1,034msAverage Server TTFBAverage Time to First Byte recorded across all live production domains
28.6%Missing Schema.org Rate26 of 91 sites lacked any JSON-LD structured data graph
96% vs 27%AI Readiness DisparitySaaS / DevTools vs Agencies with active llms.txt endpoints
Key Empirical Findings:
- •76.9% of audited production websites failed Google mobile Largest Contentful Paint target (LCP <= 2.5s), with an overall cross-sector average of 4.58 seconds.
- •E-Commerce brands exhibited the heaviest payloads (849.5KB average HTML) and slowest load times (6.51s average LCP), driven by uncompressed assets and marketing trackers.
- •Indie Hackers delivered the leanest HTML payloads (141.9KB average) and fastest load times (2.86s average LCP), but suffered the lowest Schema.org adoption at just 59.1%.
- •A striking AI discovery divide emerged: 96.0% of SaaS and Developer Tools have published llms.txt files, compared to only 27.3% of Digital Agencies and 36.4% of Indie Hackers.
- •3.3% of audited production domains still serve uncompressed raw HTML over the wire, completely missing Gzip and Brotli compression.
Original Data ReportSample: 500 active production domains
The State of WordPress Performance 2026: 500-Site Empirical Benchmark
An empirical teardown of mobile Core Web Vitals, page weight, DOM depth, and cache hit ratios across production websites.
32%Mobile LCP Pass Rate160 of 500 sites met the 2.5s LCP threshold at the 75th percentile
3.4MBMedian Mobile Page WeightDriven primarily by uncompressed media and third-party marketing pixels
380msAverage Server TTFBMeasured across 8 geographic testing regions
1,420Average DOM Node CountHigher node counts correlated with complex visual page builders
Key Empirical Findings:
- •68% of audited WordPress sites in our sample did not meet Google-s mobile Largest Contentful Paint target (LCP <= 2.5s).
- •The median WordPress site in our cohort shipped 1.8MB of JavaScript, with 62% marked as unexecuted during initial viewport render.
- •Sites built on lightweight native block themes (GeneratePress, Kadence) averaged significantly lower mobile LCP times than those on legacy drag-and-drop builders.
- •Sites configured with server-level edge HTML caching demonstrated an average TTFB of 145ms versus 620ms on standard shared hosting environments.
Original Data ReportSample: 100 digital agency homepages
100 Agency Websites: What Their Homepages Reveal About Speed and SEO
We audited 100 digital agency homepages to document real-world performance bottlenecks and structured data adoption.
74%Autoplay Video RateObserved on 74 of 100 analyzed homepages
48%Valid Schema Entities48 sites contained validated Organization JSON-LD markup
4.2MBAverage Hero WeightInitial visual assets were frequently uncompressed
240msAverage Long Task DurationMain-thread execution measured during interaction emulation
Key Empirical Findings:
- •74% of agency homepages in our cohort loaded background video, adding an average of 4.2MB to initial mobile payloads.
- •52% of agency sites lacked complete Schema.org Organization structured data linking their services and entity identifiers.
- •In our sample, agencies with sub-1.5s mobile load times also showed a 40% higher average domain rating, correlating with more active technical maintenance.
- •Heavy animation scripts (GSAP, Locomotive Scroll, Three.js) were loaded globally on 44% of homepages, correlating with elevated main-thread long task durations on simulated mobile profiles.
Original Data ReportSample: 200 technical publishing URLs
AI Search Readiness Study: How AI Crawlers Parse Technical Content
Observational research examining GPTBot, PerplexityBot, and ClaudeBot crawling behavior across 200 technical documents.
3.2xObserved Citation FrequencyObserved across structured claim-and-evidence passages in synthetic test queries
3-5 DaysAverage Recrawl WindowObserved recrawl cycle for actively updated technical domains
98%Static HTML Read RateStatic crawlable HTML achieved the highest parsing reliability
Key Empirical Findings:
- •Domains with explicit llms.txt endpoints were crawled by verified AI crawler user-agents within our monitoring window, though citation inclusion varied significantly by query context.
- •Passages structured with concise factual claims followed by specific data points were cited more frequently in our synthetic answer retrieval tests.
- •Client-side JavaScript rendering without server-rendered HTML resulted in incomplete text extraction for AI crawlers that do not execute full browser scripts.
Original Data ReportSample: 250 B2B SaaS marketing homepages
How Much JavaScript Do SaaS Websites Actually Ship?
Benchmarking 250 B2B SaaS marketing homepages for script execution time, tracking pixel bloat, and hydration overhead.
2.4MBMedian Script PayloadJavaScript code transferred over wire
64%Third-Party Tag ShareShare of main thread execution consumed by tracking tags
820msAverage Blocking TimeTotal blocking time recorded on simulated mobile hardware
Key Empirical Findings:
- •The median B2B SaaS homepage in our sample executed 2.4MB of JavaScript before user interaction.
- •Third-party marketing tags (chat widgets, tag managers, analytics) accounted for an average of 64% of total main-thread blocking time.
- •Static-generated marketing sites (Next.js SSG, Astro) loaded significantly faster than client-rendered SPAs on mobile network profiles.