Blog

• Why AI Does Not Quote Your Best Articles: A Five-Blocker Structure Audit — Score a page against five structure blockers: late answers, unnamed claims, label headings, facts that are not text and nothing original. Then fix the worst first.
• B2B Content Marketing AI Statistics: What the 2026 Research Actually Shows — 95% of B2B organisations use AI and 39% say it improved content performance. Here are the figures with their sample sizes and field dates, plus why one widely quoted growth gap does not mean what it is used to mean.
• Google's AI Generated Content Policy: Four Rules From the Documentation — Google does not prohibit AI-written content. It prohibits volume with nothing added, and since August 2026 it also names manipulating AI responses as spam. Here are the four rules, quoted from the source.
• Share of Model: The Formula, the Vanity Charge, and the One Fix — Share of model counts how often assistants name your brand against rivals. The formula is settled enough to use; the single blended figure most teams report is the part that deserves the vanity charge.
• AI Visibility Tracker or a Spreadsheet? Seven Questions and the Real Prices — Whether to pay for an AI visibility tracker comes down to prompt allowances and run counts, not features. Here is the side-by-side, the current entry prices, and the seven questions that settle a trial.
• AI Visibility Metrics: Why the Number Moves, and How to Report It Honestly — AI answers vary between identical runs, which breaks the way most dashboards report visibility. Here is the mechanism, the formulas behind a tracker's score, and the sample size at which a rate becomes a finding.
• ChatGPT Intelligent UI: Four Claims, Traced to Their Sources — Four claims about ChatGPT's Intelligent UI, checked against what the launch coverage and the Help Center actually state. Two hold with caveats; two have no evidence behind them yet.
• AEO vs SEO vs GEO: A Claim Check on What the Vendors Say — Three vendor guides define the same acronyms three different ways, then repeat the same five statistics. We traced each claim to its primary source and found one that holds with a date attached, one that collapses, and one that nobody cites correctly.
• AI Search Visibility: A Seven-Blocker Audit You Can Score in an Hour — Score seven reasons assistants answer your question without naming you, from edge-level bot policies and facts trapped behind tabs to a brand nobody else writes about. Each blocker has a five-minute test and one decision.
• Zero Search Volume Keywords: When the Zero Hides a Buyer — A reported zero describes a tool's reporting floor, not the absence of demand. Here is where the zero comes from, a five-question filter for deciding which ones earn a page, and why your own Search Console hides nearly half the picture too.
• Track Brand Citations in AI: A Four-Step Spreadsheet Playbook — A free four-step method for tracking whether assistants name and cite your brand, with the run count set by two 2026 studies rather than by convention. Copy-paste field layouts for every step.
• What ChatGPT Reads on Your Page, and the Bot That Decides If You Appear — Three OpenAI user agents can reach your site and only one governs whether you show up in ChatGPT answers. A January 2026 sentinel test and logged research sessions show what each assistant can and cannot see in your markup.
• Generative Engine Optimization: Six Popular Statistics, Traced to Source — Most generative engine optimization guides recycle the same four figures without a sample size. We traced six of them to their original pages and marked which hold, which do not, and which nobody has documented.
• WebMCP Explained: How Your Site Could Tell AI Agents What It Can Do — WebMCP is a draft browser API that lets a page declare named actions to AI agents instead of making them guess from the layout. Where the spec stands in October 2026, what Chrome has shipped, and who should act now.
• Your PR Machine Has Five Blockers: Score Yours in 45 Minutes — Five things stop a B2B brand earning third-party coverage, and all five are testable in under an hour. Score each from 0 to 2, then fix the two that scored highest before you spend a cent on an agency.
• If Your Site Is Not in English, ChatGPT May Research Your Topic Without You — One vendor study says much of ChatGPT's background research for non-English prompts runs on the English web. A larger 2026 dataset says the citations it ends up showing are mostly local. Both can be true, and the gap between them is where the work is.
• Most Marketing Insights Are Just Summaries: How to Write One That Is Not — A summary arranges what happened. An insight names a cause, commits to an action and says how confident you are. Here is the test, and the reporting habit that produces them.
• LinkedIn Newsletter for B2B Teams: Launch It, Then Keep It Alive — Four steps to a LinkedIn newsletter: confirm the option exists, write one promise, build issue one on a single claim, and start an email list the platform cannot reclaim.
• Growth Plays: A Plain Definition and Five a Small Content Team Can Run — A growth play is one signal, one rule, one action and one number. Here is the definition, the four parts, and five plays a team of three can start this quarter.
• Find Decaying Content With Free Search Console Data in One Hour — A four-step playbook: line up the same quarter a year apart, flag the real losers, diagnose each with impressions and click-through rate, then repair by section.