What Counts As A Citation, According To Bradley Benner
Most people use citation and mention interchangeably. Bradley Benner draws a line, and it changes what you should be counting.
Most people use citation and mention interchangeably. Bradley Benner draws a line, and it changes what you should be counting.
Every model can write your blog post already. It cannot have been in the room. Here is why we run a podcast instead of a content calendar.
Gisele Navarro of HouseFresh measured her own visitors by source instead of trusting vendor claims: ChatGPT, Claude and Gemini referrals stay about two minutes, Reddit and YouTube stay six to eight across multiple pages. Here is the eight-step audit and the page rebuild it drove.
HouseFresh measured it. AI referrals leave in two minutes. Reddit and YouTube referrals stay six to eight, across multiple pages. Gisele Navarro rebuilt the site around that gap.
I pulled 46 complete search results and scored 303 ranking pages against their own backlink anchors. Exact-phrase anchors showed nothing. Anchors matching the query’s tokens, word order ignored, showed a real effect. Here is the method, the numbers and the 12.87 USD it cost.
Patrick Stox, Ben Wills and Ben Senescu split AI visibility into two surfaces on two clocks – the retrieved pages you can work on this week, and the training data you cannot. Here is the practitioner version.
Search Console under-reported clicks on all four sites I measured, by 2% to 16%. Not dark clicks – a second gap on top of them. Here is the method and seven prompts to test your own site.
Most blog programs carry topics nobody can defend against the product catalog, protected by an email click. Laura Hoot of GPO gives the eight-step audit: name the destination page and CTA for every topic, kill or re-aim the ones that cannot, and re-seed the calendar from the catalog outward.
Google’s AI Max now picks which page on your own site a paid click lands on. Lisa Raehsler of Big Click Co lays out the eight-step audit and exclusion pass to run before the February rollout makes it mandatory.
The median Search Console property attaches a query to just 34.9% of its clicks. I measured the dark-click gap across 35 properties and confirmed the one export that does not lose them.
Published CTR curves describe somebody else’s site. Three copy-paste prompts that pull your real click-through rate out of Search Console, plus the one thing you should never ask any tool to build.
Alex Radetic of Nuvo Agency still does keyword research by opening the search and looking at it. His SERP-reading method, why Search Console impressions are inflated by AI Overviews, and the sixty-seven ChatGPT leads that reframed his AEO pitch.
AJ Kohn checked how many pages in the “Crawled – currently not indexed” bucket were actually indexed. Across 24 clients it ran from 99.8% down to 0.5%. Seven steps to score your own bucket before you spend a penny fixing it.
Erika Braeger of Ten Speed builds Claude skills that absorb the hour of data-gathering before an analysis starts, and stop there. Eight steps to scope, build, QA and measure one for your own team.
ChatGPT invents URLs on your domain and sends real people to them. Your 404 report is the list. The step-by-step for turning invented pages into real ones, from Stephanie Long’s Unscripted SEO interview.
Fiverr and Upwork’s “vetted” section says vetted by AI in the fine print. Stephanie Long of Junction Jack on how to actually hire SEO help, plus the verification habit that catches what the tools invent.
A viral thread says prompt ChatGPT ten times, screenshot the sources, and go get profiles wherever they repeat. I ran it across 13 runs with SERP controls. No source appeared in a majority of runs, 56 percent of citations pointed at the firms own websites, and on long-tail queries ChatGPT matched Google top 4 exactly.
Before you commit six months of SEO to a keyword set, buy three weeks of ads on it. The three-keyword validation test Gert Mellak runs when a client’s keyword choice is wrong.
Gert Mellak has already pulled impressions out of his client reports, because bots create them and Google does not filter them. Eight practitioner changes from the episode: kill the impressions column, stop pricing off search volume, settle keyword arguments with three weeks of Google Ads, and track high intent instead of traffic.
Paste a regex and find out whether Search Console, GA4 and Microsoft Clarity will actually run it. Catches lookaheads, backreferences and the other PCRE constructs Google’s RE2 engine silently refuses.
Done-For-You
Podcast Production Service FlagshipWe host, book guests & hand you transcript + clipsPodcast-Powered Link BuildingShow-notes backlinks money can’t buyPodcast SEO & AI Visibility GEOGet cited by ChatGPT & AI OverviewsLink Building
Link Building ServicesTwo ways that actually workCommunity-Cleanup EventsWhite-label local links that do goodWhite-Label ReportsClient-ready reportingListen
The Unscripted NetworkOur full SEO podcastWatch on YouTubeFull video interviewsListen on CastosEvery episodeExplore
Podcasting Resource GuideThe complete guidePodcast SEO StrategyLinks, entities & AI visibilityOur 8-Step ProcessConversation to conversionFree Libraries
SEO Tactic Library 60 FreeExecutable plays by difficultySEO Strategy LibraryThe frameworks behind tacticsSEObitsBite-size SEO power-upsRead
SEO Knowledge HubEverything in one placeBlogDeep dives & teardownsSEO SOPsStep-by-step operating proceduresThe Tool
Keyword Forecasting ToolFind keywords worth ranking forGet Your ForecastCustom keyword researchRequest a SampleSee a demo forecastGet Started
Create a Free Account FreeNo card requiredLog InExisting accountsHow It Works2-minute walkthrough