Fractl Research & Insights
Explore Fractl’s original research on how people discover brands, what earns media coverage, and why content gets shared. Find data and insights on AI search, digital PR, content marketing, and consumer behavior to inform your reporting and marketing strategy.
Explore original research columns by Fractl’s founders and leadership team. Discover insights on AI visibility, SEO, content marketing, and digital PR to inform your next marketing decision.
We measured 11,573 AI answers across 15 industries — what AI recommends, which sites it cites, and why getting featured on those sites is the new way to win AI recommendations.
We spent two weeks trying to beat AI detectors — built the detector ourselves, ran a census of 326 models, and trained our own rewriters. Detection is solved; the ‘humanizer’ promise runs into a real wall: you can evade, or keep the facts, not both.
We read a real AI model’s training data and ran 6,000+ controlled retrieval experiments across nine models. Most of what’s sold as ‘AI visibility’ measures neither engine that actually produces AI answers.
Get more useful AI campaign ideas with a clearer brief, different generation methods, and human review. See what Fractl’s tests found.
Get past repetitive AI ideas with outside perspectives and better follow-up questions. Learn what Fractl’s research suggests for your next brainstorm.
We tracked about 10,000 AI answers across 16 industries to see which brands and which review sites AI recommends.
Creator deals alone barely beat ordinary TikTok posts. See what 79 million posts show about boosts, and how to judge a deal.
AI blog ideas often match existing articles. See what Fractl found and how to check your shortlist before commissioning another familiar post.
Search isn’t just changing—it’s being rewritten. In just 18 months, AI went from novelty to default, and marketers are scrambling to redefine visibility.
Consumer trust in AI search dropped from 82% to 54% in a year. Fractl surveyed 1,008 consumers and 150 marketers on brand visibility for 2026.
The publisher is the prompt: Explore how AI publisher partnerships structure latent space and influence what assistants say about your industry.
If your agency pitch leads with GEO or AEO, you may be losing the room. Here is what marketing decision-makers actually want to hear.
A new volatility view tracks how SEO thought leaders are framing AI search and how often they’re changing their stance.
A data-driven look at niche media sites, podcasts, and communities that quietly influence decision-makers and how to use them for earned media.
We tracked how over 8,500 brands show up in ChatGPT, Gemini, and Claude. Five percent of leaders are vanishing from AI. Here’s who, and why it’s happening.
Fractl tested the biggest prediction in search marketing with real keyword data. The results will change how you think about SEO in the age of AI.
Marketers are redefining AI search through the language they use. Discover which terms are shaping hiring, strategy, and engagement.
As AI chatbots become companions, not just tools, discover how this shift impacts relationships, emotional well-being, and societal norms.
Most people read the AI answer, then click anyway. Fractl surveyed 1,007 Americans to find out what still earns the click in 2026.
Fractl surveyed 1,003 AI users across four generations. Baby boomers trust AI answers most, Gen Z writes the longest prompts, and 78% pay nothing.
Organic search leads at just 21 of 100 brand sites in 2026. New data on which channels actually drive traffic, and where AI answers fit.
We modeled the paid equivalent of organic traffic across 96 brands and surveyed 228 respondents about SEO cuts, CAC, and replacement cost awareness.
We compared refreshed and unrefreshed pages on 152 domains. Refreshed pages earned twice the median AI citations; traffic gains were selective.
New data on which publishers AI answers cite most, how the source mix differs by platform, and how publisher compensation is changing.
We compared cited sources across Google’s top 10, AI Overviews, AI Mode, ChatGPT, and Claude for 478 queries. Only AI Overviews were more concentrated.
