Topic
Analytics
29 articles on Analytics from the Rankry team.
The AI Visibility Tool Shake-Up of Fall 2026: Funding, Features and Platform Changes
From September to early November 2026, AI visibility tools raised big rounds, shipped similar features and had to adjust to platform changes such as Copilot access. Here is what changed, what it does to your numbers, and seven questions for any vendor.
Rankry Team · November 5, 2026
Poisoned SERPs: Why Google Feeds Rank Trackers Swapped Results and How to Check
Since September 2026, Google has answered some rank-tracker queries with normal-looking pages of swapped results: 8% of pages in Nozzle's data. Here is what is known, how it touches AI Overview tracking, and a five-check triage.
Rankry Team · October 28, 2026
Is GEO a Distraction? What the October 2026 Traffic Data Actually Says
Not entirely, but on traffic the skeptics have the stronger numbers: in one 145-million-session analysis, ChatGPT sent 0.62% of what Google organic sent. Here is what the October 2026 data says, and how to size a GEO budget.
Rankry Team · October 24, 2026
AI Overviews on 88% of Commercial Searches: What It Means for "Best" and "Review" Pages
In a Semrush panel of 400 US commercial keywords, AI Overviews appeared on 88% in August 2026, up from 45.75% a year earlier. Here is what the number means, and how to keep "best" and "review" content earning clicks and citations.
Rankry Team · October 22, 2026
Bing Topics Portfolio: How to Turn Free AI Search Topic Data Into a Content Plan
Bing's new Topics Portfolio groups the AI grounding queries where your site appears into topics, with sizes and opportunity queries where others get cited and you don't. Here is how to turn it into a content plan.
Rankry Team · October 13, 2026
Google AI Overviews Cut Organic Clicks by About 40%: What You Lose and What You Gain
The best evidence so far says AI Overviews cut organic clicks by about 40% on informational searches where they appear. Here is what you lose, what you gain when you're the cited source, and how to measure both.
Rankry Team · October 11, 2026
GPT-6 in ChatGPT: What Changed for Brand Answers and How to Re-Baseline
GPT-6 has been rolling out in ChatGPT since October 7. No study yet compares brands per answer before and after the switch, so here is what is known, what isn't, and how to re-baseline your tracking.
Rankry Team · October 9, 2026
Do AI Engines Prefer Local Websites? What the Country-by-Country Data Shows
Often yes, but the trigger is the language of the question more than the country of the user. Here is the country-by-country citation data, how much it varies by engine, and what it means if you sell in several markets.
Rankry Team · October 6, 2026
Branded Prompts Inflate Your AI Visibility Score (91.6% vs 19.6%)
AI engines mention a brand almost every time a prompt names it, and far less often when buyers ask about the category. If your tracked set mixes the two, your score is inflated. Here is how to split it.
Rankry Team · October 1, 2026
What AI Visibility Tracking Really Costs per Prompt: How to Compare Tools Fairly
A tool's monthly price says little until you divide it by prompts, engines and runs. Normalized per prompt-engine-run, plans that look far apart land close together. Here is the formula and the six questions to ask any vendor.
Rankry Team · September 25, 2026
September 2026 AI Citations: LinkedIn +39%, YouTube -42%, Reddit Cut in ChatGPT
September 2026 data shows LinkedIn up 39% and YouTube down 42% in AI citations. Reddit wasn't halved across engines: its collapse happened inside ChatGPT, starting August 8. Here is what moved, where, and what to do.
Rankry Team · September 22, 2026
How to Read Your GSC Data to Find AI Search Opportunities
Impressions up with clicks down is usually AI answering on the page and citing you, not broken SEO. Here are the four patterns worth hunting in your own Search Console data, each pointing at one specific action, and the blind spot no Google report will ever cover.
Rankry Team · September 16, 2026
Do AI Answers Change by City? Geo-Variation in Brand Recommendations
This is not the near-me problem: the same prompt, asked from different cities, returns a differently ranked list even for a business with no physical location. Money and market-specific categories swing hardest. Here is how to measure it per market and close the gaps where you sell.
Rankry Team · September 15, 2026
What 500 Runs of One ChatGPT Question Look Like (and How to Measure It Yourself)
Run one buyer prompt hundreds of times and "who ranks #1" stops being a fact: in the modeled distribution used here, twelve brands took the top spot at least once and the leader held it in about a fifth of runs. Here is the shape of that variance, and how to measure it on your own brand.
Rankry Team · September 7, 2026
The Prompt Volume Problem: How to Know Which Buyer Questions to Even Track
Every prompt you add has to be run many times to be trustworthy, which makes a sprawling list worse than a focused one. Here is how to pick the 30 to 50 questions that matter, rank them by how close they sit to a purchase, and when to revise the set rather than quietly breaking your own trend.
Rankry Team · September 6, 2026
The Hidden Cost of Being AI's Favorite: When Citations Don't Convert
The visits AI sends are excellent. The trap is the page so completely restated inside the answer that no visit is needed: you get the credit and give away the value. Here is which content gets consumed, which forces the click, and how to earn citations that still leave a reason to visit.
Rankry Team · September 4, 2026
AI-Referred Visitors Convert Better: The Business Case for LLM Visibility
AI referrals are a small share of traffic and convert several times better than typical organic, because those visitors read a summary that vetted you before they clicked. Here is how to argue for the budget on value per visitor rather than volume, and which slide actually lands.
Rankry Team · August 17, 2026
Why Your Direct Traffic Is Exploding (And What It's Really Telling You)
A link copy-pasted out of an AI answer arrives with no referrer, so analytics files your highest-intent visitors under Direct, next to people typing your address. Four tells separate real direct from AI in disguise, and rising direct is usually proof you are being cited.
Rankry Team · August 16, 2026
Bots Now Outnumber Humans on Your Site: How to Read Your Real Traffic
Automated requests crossed half of all web traffic in 2026, so a traffic graph can double while revenue stays flat. Your logs hold three populations, humans, crawlers, and agents, and a single sessions number treats a buyer's agent and a training crawler as the same event. Here is how to separate them.
Rankry Team · August 14, 2026
Agentic Search vs Zero-Click: The Two Different Ways You're Losing Traffic in 2026
Zero-click means a human read your answer and never clicked. Agentic means no human searched at all, because an AI agent researched and decided for them. The two look identical in analytics, they are not the same problem, and the playbook that fixes one does nothing for the other.
Rankry Team · August 13, 2026
How Many Times Should You Ask ChatGPT Before You Trust the Answer?
A single AI visibility check is a coin toss, not a measurement. At 20 runs a 50% carries roughly a 22-point margin, and precision grows only with the square root of the run count, so halving the error means quadrupling the runs. Here is the arithmetic, and the rule it gives you for reading a weekly move.
Rankry Team · August 12, 2026
Mention Tracker vs Rank Tracker vs Visibility Tracker vs Citation Tracker: Four Names, Four Different Metrics
Four words people treat as one product measure four different things: whether you were named, where you landed, what share of runs include you, and whether the AI linked your domain. They move independently. Here is how to tell them apart, and the one question that exposes what a tool actually counts.
Rankry Team · August 11, 2026
Introducing AI Traffic: See the Visitors AI Sends to Your Site
Rankry now tracks the real visitors AI engines send you, not just what those engines say. One script, ten AI sources, geography, language, device, engagement, landing pages, and the query behind the visit, next to your visibility data.
Rankry Team · August 6, 2026
Which AI Sends the Most Traffic? ChatGPT vs Perplexity vs Gemini
ChatGPT sends 87.4% of AI referral traffic in Conductor's 2026 benchmark, but volume is not value: Gemini referrals grew 388% year over year and Perplexity skews to research intent. How to find your own mix.
Rankry Team · August 5, 2026
How to Track AI Referral Traffic in GA4 (2026 Guide)
The full GA4 setup for AI referral traffic: custom channel group, the regex, Explore for history, and GA4's native AI Assistant channel. Plus the two blind spots no configuration can close, and when a dedicated script is the better tool.
Rankry Team · August 4, 2026
Why Is My Direct Traffic Increasing? (It's Probably AI)
Direct traffic climbing with no campaign behind it is usually AI referrals GA4 has misfiled. Here is what sits inside your Direct bucket, four checks that tell AI from real direct, and the honest limit on what any tool can recover.
Rankry Team · August 4, 2026
Do AI Answers Actually Drive Traffic and Revenue? The Honest Data
AI referral traffic is about 1% of visits but converts 4-5x better than organic search. The honest data on whether AI answers drive revenue, and how to measure it on your own site.
Rankry Team · July 22, 2026
What Is AI Share of Voice and How to Measure It
AI Share of Voice is your slice of all brand appearances in AI answers. How to calculate it with a worked example, the three kinds that matter, what counts as good, and how to improve it.
Rankry Team · June 30, 2026
AI Referral Traffic: Does ChatGPT & Perplexity Send Visitors, and How to Track It
ChatGPT, Perplexity, Gemini, Claude, Grok, Copilot, and Google AI Overviews send real, high-intent visitors. Here is how to find them in GA4, set up tracking that survives the gaps, and why visibility and referral traffic are two different numbers.
Rankry Team · June 13, 2026