Josh Blyskal

Josh Blyskal leads AI strategy at Profound. Working from more than 250 million AI responses and about 3 billion citations, he has published findings that many teams now treat as baseline context.

Two results travel especially widely. Conventional SEO metrics explained only 4 to 7 percent of AI citations in the analyses he has presented. Separately, Reddit's share of ChatGPT citations grew more than eightfold in 2025 in the patterns he reported.

Scale is the point. Plenty of agencies can run a few thousand prompts. Blyskal's datasets are large enough that a finding can survive the usual "maybe your sample is weird" objection.

Profound's commercial position means Blyskal's talks also function as category education for buyers evaluating AI visibility platforms. Even readers who never buy the software still use the citation stats as orientation for what normal looks like at scale.

If classic rank and authority metrics explain only a thin slice of citations, then GEO programs that only remix SEO dashboards are under-instrumented. Teams need AI-specific observation: which domains get reused, in which answer types, and how that mix shifts over months. That measurement gap is why studies like Rand Fishkin's inconsistency work still matter beside Blyskal's scale.

The Reddit finding pushed a parallel conversation about community and UGC sources in generative answers. Whether a brand should participate on Reddit is a separate strategy question. Blyskal's contribution was showing the citation share moving in public data so the debate could start from evidence.

His architect role on this list comes from measurement that changed what peers believe is knowable about AI citations.

The weak correlation with classic SEO metrics is the finding managers still underestimate. It implies a separate instrumentation project, not a renamed SEO report.

Reference-point stats only matter if peers keep citing them. Blyskal's findings cleared that bar quickly, which is why they appear in so many 2026 planning decks.

Start with his Tech SEO Connect presentation analyzing 250 million AI search responses. It is the clearest single artifact for how he walks audiences through the scale of the dataset and the headline citation findings.

Conference audiences often remember the Reddit multiple and the 4 to 7 percent SEO-metric explanation as a pair. Together they say AI citation behavior is both shifting toward community sources and poorly predicted by yesterday's SEO dashboards.