{ "title": "Periodic Table of AI Visibility", "edition": "2026.2", "audited": "2026-10-02", "author": "Ron Davis", "canonical": "https://www.itsrondavis.com/ai-visibility", "source": "https://github.com/vell-admin/ai-visibility-table", "license": "CC BY 4.0", "groups": [ { "key": "e", "label": "Engines", "note": "the surfaces that do the citing" }, { "key": "r", "label": "Retrieval", "note": "whether a machine can read you at all" }, { "key": "s", "label": "Signals", "note": "what the model weighs once it can read you" }, { "key": "a", "label": "Artifacts", "note": "the shapes that actually get quoted" }, { "key": "c", "label": "Channels", "note": "where the artifact goes to be found" }, { "key": "m", "label": "Measurement", "note": "how you know any of it moved" }, { "key": "x", "label": "Inhibitors", "note": "detached series — subtract, don't add" } ], "elements": [ { "number": 1, "symbol": "Cg", "name": "ChatGPT", "group": "Engines", "group_key": "e", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "The highest-volume answer surface. It names a handful of sources per answer and almost never shows the runners-up.", "the_move": "Ask it the question your buyer would actually type, once a month, and write down every company it names. That list is your real competitive set.", "examples": [ "A buyer types 'best tool for X' and gets three names back. If you are not one of the three, you do not exist for that buyer.", "A founder checks once a month and sees the same rival named four times in a row. That rival is the real competitor, not the one in the pitch deck.", "It is also the opt-in 'ask ChatGPT' handoff inside Siri on iPhones, so the same answer can be spoken out loud." ], "watch_out": "Answers change from one day to the next, so a single check proves nothing. You need the same questions asked on a schedule.", "audit_2026": "Still the biggest surface: about 54% of worldwide AI-chat web visits in May 2026 (Similarweb via Momentic), down from roughly 79% a year earlier. Gemini and Claude are taking the share.", "unit": null, "direction": null, "alternates": null, "part": 1 }, { "number": 2, "symbol": "Pp", "name": "Perplexity", "group": "Engines", "group_key": "e", "time_to_effect": "days", "inhibitor": false, "what_it_is": "Citation-first by design — it shows its sources inline, which makes it the cheapest place on earth to check whether you are retrievable.", "the_move": "Search your own category here first. If nothing of yours comes back, stop writing and go fix retrieval.", "examples": [ "Ask it your category question and the sources show up right next to the answer. If your site is not in that list, the model never read you.", "A small vendor fixed one blocked crawler and showed up here within days. Fastest proof there is that retrieval was the problem.", "Buyers use it like a research assistant: 'compare A and B' pulls whichever comparison pages it can find, yours or a review site's." ], "watch_out": "Tiny share of total AI traffic (about 1 to 2%), so it is a thermometer, not a market. Winning here is a test, not a result.", "audit_2026": "Still citation-first. Share stayed small through 2026 as the bigger assistants added their own search.", "unit": null, "direction": null, "alternates": null, "part": 1 }, { "number": 3, "symbol": "Cl", "name": "Claude", "group": "Engines", "group_key": "e", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Answers from fewer, longer sources. Rewards a page that holds a full argument over one that covers a keyword.", "the_move": "Give it one page that answers the whole question end to end instead of five that each answer a fifth.", "examples": [ "A buyer pastes a long question and gets an answer built from two or three deep pages instead of ten shallow ones.", "A company with one complete 'how it works' guide gets quoted; the competitor with twelve blog snippets gets skipped.", "It sits inside the tools people work in all day, so the question gets asked mid-task, not in a browser tab." ], "watch_out": "Smaller audience than ChatGPT or Gemini, and it leans on fewer sources, so you are either the source or you are absent.", "audit_2026": "Fastest-growing major assistant through 2026: roughly 9% of worldwide AI-chat web visits by May 2026, about 13% in the US.", "unit": null, "direction": null, "alternates": null, "part": 1 }, { "number": 4, "symbol": "Ao", "name": "AI Overviews", "group": "Engines", "group_key": "e", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Google's generated answer above the ten blue links. It absorbs the click that used to go to position one.", "the_move": "Check which queries in your category now trigger an Overview. Those are the pages where ranking alone stopped paying.", "examples": [ "Someone Googles 'what is X' and reads the answer box without clicking anything. Position one used to get that click; now the box gets it.", "A page that ranked first for a question lost half its clicks the month an Overview appeared above it. Same ranking, less traffic.", "If the Overview names you, you get the click that used to be split across three sites." ], "watch_out": "You cannot opt in. The box appears or it does not, and you only find out by checking the queries yourself.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": 1 }, { "number": 5, "symbol": "Am", "name": "AI Mode", "group": "Engines", "group_key": "e", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Google's conversational surface. It fans one typed question into many hidden ones and assembles the answer from all of them.", "the_move": "Write for the follow-up questions, not just the entry question — the fan-out is where most sources get picked up.", "examples": [ "A buyer types one question; Google quietly asks eight more ('pricing?', 'alternatives?', 'is it secure?') and stitches the answers together.", "A pricing page got cited for a question that never mentioned price, because the hidden fan-out asked about it.", "The page that answers the follow-up questions gets picked up more often than the page that only answers the first one." ], "watch_out": "The hidden questions are never shown to you, so you have to guess at them from your own sales calls and support tickets.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": 1 }, { "number": 6, "symbol": "Ge", "name": "Gemini", "group": "Engines", "group_key": "e", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Google's assistant, and increasingly the default inside Workspace, where a lot of B2B research quietly happens.", "the_move": "Test your category prompt in a signed-in Workspace account, not an incognito tab — the grounded answer differs.", "examples": [ "An employee asks Gemini inside Gmail or Docs 'which vendors do X?' The research happens inside the work tools and never touches a search page.", "Since June 2026 the new Siri runs on Gemini, so the same answer can come out of an iPhone when someone asks out loud in the car or the kitchen.", "A signed-in Workspace answer can differ from a logged-out one, because it can see the company's own documents." ], "watch_out": "There are two Geminis, the app and the Workspace version, and they can answer differently. You have to test both.", "audit_2026": "Big change: Apple's Siri AI (announced at WWDC, June 8 2026) is built on Gemini. Roughly 28% of worldwide AI-chat web visits in May 2026, up from about 5% a year earlier.", "unit": null, "direction": null, "alternates": null, "part": 1 }, { "number": 7, "symbol": "Cp", "name": "Copilot", "group": "Engines", "group_key": "e", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Bing-grounded and embedded in Microsoft 365, which means it sits inside the tools enterprise buyers already have open.", "the_move": "Verify Bing has indexed your site at all. Plenty of sites are fine in Google and invisible in Bing.", "examples": [ "A buyer in Outlook asks Copilot to 'summarize options for X.' The answer comes from Bing's index, not Google's.", "A company that never checked Bing found its newest pages were not indexed there at all. Invisible to Copilot, fine in Google.", "It is already paid for inside Microsoft 365, so enterprise buyers use it without ever choosing to." ], "watch_out": "Small consumer share (about 1 to 3%), and Bing's index lags Google, so fresh pages can take weeks to show up.", "audit_2026": "Still a small share of consumer visits but embedded in the enterprise suite. Unchanged in kind.", "unit": null, "direction": null, "alternates": null, "part": 1 }, { "number": 8, "symbol": "Gk", "name": "Grok", "group": "Engines", "group_key": "e", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Grounded heavily in live posts, so it reflects what is being said about you right now rather than what your site claims.", "the_move": "Watch what it says about your company. It is the fastest read on your public reputation as machines see it.", "examples": [ "Grok is built into Tesla cars. A driver asks 'who makes the best X?' out loud and hears a spoken list, a ranking read to someone with both hands on the wheel.", "It leans on live posts, so if people are complaining about you this week, Grok says so this week.", "A company with no site update in a year but plenty of recent praise on X still came up well. The posts outweighed the site." ], "watch_out": "Grounded in social posts, so one loud week of criticism can rewrite what it says about you overnight.", "audit_2026": "Verified Oct 2026: Grok ships in Teslas with the AMD chip (roughly 2021 and newer) since the 2025.26 update; the Summer 2026 update added cabin controls, and Grok Bot plus Connectors went live in-car on Sept 22, 2026. About 2 to 3% of AI-chat web visits, but the only major engine with a car cabin.", "unit": null, "direction": null, "alternates": null, "part": 1 }, { "number": 9, "symbol": "Rb", "name": "robots.txt", "group": "Retrieval", "group_key": "r", "time_to_effect": "days", "inhibitor": false, "what_it_is": "The gate every crawler checks first. One over-broad disallow line can shut out the exact agents that feed the answer engines.", "the_move": "List the crawlers you actually depend on and allow them by name before you block anything by default.", "examples": [ "One line meant to stop scrapers also blocked the crawler that feeds ChatGPT. The site vanished from answers and nobody noticed for months.", "A site allowed Google but never named the AI crawlers. The default rule said no to everyone else.", "Fixing one line took five minutes. The site reappeared in Perplexity within a week." ], "watch_out": "Allowing every crawler also invites the ones that only scrape. Allow the named ones you want, not a blanket yes.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": 3 }, { "number": 10, "symbol": "Sm", "name": "Sitemap", "group": "Retrieval", "group_key": "r", "time_to_effect": "days", "inhibitor": false, "what_it_is": "The index of what exists and when it last changed. Without it, discovery depends on links you may not have.", "the_move": "Confirm every page you want cited is in it, and that lastmod reflects real edits rather than a build timestamp.", "examples": [ "A new page sat unlinked and unlisted for six weeks. No sitemap entry, no inbound links, no discovery.", "The 'last changed' date on every page said the same thing: the day the site was rebuilt. Crawlers learned to ignore it.", "A sitemap listing 2,000 pages, 1,500 of them redirects, taught the crawler the map was junk." ], "watch_out": "A sitemap only says 'this exists.' It does not make a page worth citing; it just makes sure the page gets read.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": 3 }, { "number": 11, "symbol": "Lt", "name": "llms.txt", "group": "Retrieval", "group_key": "r", "time_to_effect": "days", "inhibitor": false, "what_it_is": "A plain-text map of what matters on your site, written for models instead of browsers. Emerging convention, cheap to ship.", "the_move": "Publish one that points at your definitions, pricing, and proof pages. Treat it as a hint, not a guarantee.", "examples": [ "It is a plain text list: here is our pricing page, here is our definition page, here is our proof. Takes about an hour.", "A docs-heavy company's file gets pulled by coding tools when a developer asks about their product.", "Think of it as a note taped to the front door for robots: the important rooms are these three." ], "watch_out": "Google says it does not use the file for Search, AI Overviews or AI Mode (May 2026 guidance), and most files get no traffic at all. Ship it cheap; expect nothing from Google.", "audit_2026": "Checked Oct 2026: adoption about 10% of domains; an Ahrefs study of 137,000 sites found 97% of files got zero visits in May 2026. Spec v2 shipped Aug 2026, Chrome's Lighthouse added an agentic-browsing check for it, and OpenAI and Anthropic recommend it for agents. Still a hint, now with a known ceiling.", "unit": null, "direction": null, "alternates": null, "part": 3 }, { "number": 12, "symbol": "Ax", "name": "agents.txt", "group": "Retrieval", "group_key": "r", "time_to_effect": "days", "inhibitor": false, "what_it_is": "Declares what an agent can do on your site — not merely what it may read. The capability layer above robots.txt.", "the_move": "Only declare capabilities that genuinely work. A false claim here fails in front of the buyer's agent, not yours.", "examples": [ "It is the difference between 'you may read my menu' and 'you may place an order.' Robots says read; agents says do.", "A site claimed an agent could book a demo. The button did not work for a bot. The buyer's agent gave up and reported that.", "Start with one capability that works end to end, like 'fetch current pricing,' before listing five." ], "watch_out": "Still an early convention. No major engine has confirmed it reads the file, so do not spend more than an afternoon on it until one does.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": 3 }, { "number": 13, "symbol": "Jl", "name": "JSON-LD", "group": "Retrieval", "group_key": "r", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Structured data that states the facts of a page in a form no parser has to guess at: product, price, author, organization.", "the_move": "Start with Organization and Product. Everything else is optional until those two are right.", "examples": [ "A product page says 'Starts at $49 a month' in plain text. The structured data says it as a number no parser has to guess at.", "Two companies with the same name: the one with Organization markup and a sameAs link to its LinkedIn got picked; the other got confused with a bakery.", "It is the label on the can. The picture on the front is for people; the label is for machines." ], "watch_out": "Markup that disagrees with the visible page gets ignored, or flagged. Keep both saying the same thing.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": 3 }, { "number": 14, "symbol": "Et", "name": "Entity", "group": "Retrieval", "group_key": "r", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "The link between your name and a thing the model already knows about. sameAs references, consistent naming, corroborating profiles.", "the_move": "Pick one canonical name and spelling, then use it identically everywhere. Ambiguity splits your evidence in half.", "examples": [ "'Acme' is a company, a cartoon, and three other businesses. Without corroborating profiles, the model does not know which one you are.", "A founder used 'Acme AI,' 'Acme.ai' and 'ACME' across three profiles. The model saw three weak entities instead of one strong one.", "Corroboration means Crunchbase, LinkedIn, a marketplace listing and a podcast bio all spelling the name the same way." ], "watch_out": "Slow. Entity signals take quarters to build and quarters to fix, so a rename costs more than it looks.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": 3 }, { "number": 15, "symbol": "Ck", "name": "Chunk", "group": "Retrieval", "group_key": "r", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Retrieval quotes passages, not pages. A paragraph that only makes sense after reading the three above it will not be quoted.", "the_move": "Make every section survive being torn out of the page. Restate the subject instead of writing 'it' and 'this'.", "examples": [ "A paragraph starting 'This makes it faster' got lifted out of the page. Nobody, model or reader, knew what 'this' was.", "Each section of a runbook was rewritten to restate the subject. The same page started getting quoted three weeks later.", "Think of each section as a sticky note that could fall off the page and still make sense on its own." ], "watch_out": "Restating the subject everywhere reads clunky to a human. Do it in the first sentence of each section, not every sentence.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 16, "symbol": "Sr", "name": "Server render", "group": "Retrieval", "group_key": "r", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Content present in the HTML response rather than assembled later by JavaScript. Many crawlers never run the script.", "the_move": "View source — literally. If your key claim is not in the raw HTML, assume no engine has ever read it.", "examples": [ "A pricing table built in JavaScript showed nothing in 'view source.' Engines read an empty page with a header.", "A marketing site looked beautiful and was blank to every crawler that does not run scripts, which is most of them.", "Right-click, View Page Source, search for your main claim. If it is not there, assume no engine has read it." ], "watch_out": "Server rendering can slow a dev team's workflow. Render the pages you want cited; leave the app screens alone.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 17, "symbol": "Fr", "name": "Freshness", "group": "Retrieval", "group_key": "r", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "A visible, accurate date. Models weigh recency heavily on anything that looks like it could be stale.", "the_move": "Put a real published-and-updated date on every substantive page, and only bump it when the content changed.", "examples": [ "Two near-identical guides; the one with 'Updated March 2026' got cited over the one with no date.", "A page bumped its date every week without changing a word. Engines learned the date meant nothing.", "An old benchmark with an honest 'published 2024' date still got cited, because the date was real." ], "watch_out": "Fake freshness backfires. Only bump the date when the content actually changed.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 18, "symbol": "Cn", "name": "Canonical", "group": "Retrieval", "group_key": "r", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "The single address a piece of content lives at. Split versions divide the signal between them and neither wins.", "the_move": "Point every duplicate and syndicated copy at one canonical URL before you build more links to any of them.", "examples": [ "The same article on your blog, Medium and LinkedIn with no canonical: three weak copies, none of them won.", "A product page at /pricing and /pricing/ (trailing slash) split its signal in half for a year.", "One canonical tag pointed the Medium copy back home. The home page started getting the citations." ], "watch_out": "Pointing everything at one URL means that one URL has to be the best version. Fix it before you consolidate.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 19, "symbol": "Tr", "name": "Transcript", "group": "Retrieval", "group_key": "r", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "A video is citable only to the extent its words exist somewhere as text. The file itself is opaque.", "the_move": "Publish the transcript on a real page. This is why engines cite YouTube and never cite a raw video file.", "examples": [ "A 40-minute video with a real transcript page got cited for a question answered at minute 23.", "The same talk uploaded as a bare MP4 on the company site was never cited, because there were no words to read.", "YouTube auto-captions count on YouTube. A page on your own site with the cleaned-up text counts everywhere." ], "watch_out": "Raw auto-captions are full of errors. A wrong product name in a transcript can get quoted exactly as written.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 20, "symbol": "At", "name": "Alt text", "group": "Retrieval", "group_key": "r", "time_to_effect": "days", "inhibitor": false, "what_it_is": "The only description of an image most retrieval pipelines will ever have. Chart, screenshot, diagram — all invisible without it.", "the_move": "Describe what the image proves, not what it depicts. 'Bar chart' says nothing; the finding says everything.", "examples": [ "A chart showing '43% faster' had alt text 'chart.png.' The finding was invisible.", "Changed to 'Bar chart: setup time dropped from 7 days to 4 days after switching.' Now the number is readable.", "A screenshot of a dashboard is just pixels to retrieval unless you say what it proves." ], "watch_out": "Stuffing keywords into alt text reads as spam and hurts screen-reader users. Describe the finding, once.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 21, "symbol": "Hd", "name": "Heading map", "group": "Retrieval", "group_key": "r", "time_to_effect": "days", "inhibitor": false, "what_it_is": "The outline a parser builds from your H2s and H3s. It decides where one idea stops and the next begins.", "the_move": "Make every heading a claim or a question, never a label like 'Overview'. Labels carry no retrievable meaning.", "examples": [ "Headings like 'Overview,' 'Features' and 'Benefits' tell a parser nothing about what the page answers.", "Rewritten as 'How long does setup take?' and 'What does it cost per seat?' each section now matches a real question.", "A parser builds an outline from your H2s. If the outline could belong to any site, you are interchangeable." ], "watch_out": "Question-shaped headings get long. Keep them under ten words or they stop reading as headings.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 22, "symbol": "Ln", "name": "Internal links", "group": "Retrieval", "group_key": "r", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "How crawlers find the page and how models infer what it is about from the words pointing at it.", "the_move": "Link to the page using the words you want to be found for. Kill every 'click here' and 'learn more'.", "examples": [ "Twenty links that all say 'learn more' tell the crawler nothing about the twenty pages behind them.", "A pricing page linked with the words 'pricing for small teams' started showing up for that exact phrase.", "If nothing links to a page, the crawler never finds it. If everything links to it with the same words, the crawler knows what it is." ], "watch_out": "The same anchor text everywhere reads as manipulation. Use the natural phrase, not one phrase fifty times.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 23, "symbol": "Cy", "name": "Clarity", "group": "Signals", "group_key": "s", "time_to_effect": "days", "inhibitor": false, "what_it_is": "Whether the first line says what the thing is in words a stranger already knows.", "the_move": "Delete the first sentence and check whether anything was lost. Usually it was throat-clearing.", "examples": [ "First line: 'We empower modern teams to unlock value.' A stranger learned nothing. Second draft: 'We send your invoices for you.' Done.", "Delete the first sentence of your homepage. Did anything break? Usually not.", "A model has to decide in one line what you are. If it cannot, it picks someone else." ], "watch_out": "Clear can sound plain. That is fine. Plain gets quoted; clever gets skipped.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 24, "symbol": "Sy", "name": "Specificity", "group": "Signals", "group_key": "s", "time_to_effect": "days", "inhibitor": false, "what_it_is": "Named nouns beat adjectives. A sentence with a number, a unit, and a named thing in it is quotable; one without is not.", "the_move": "Rewrite your strongest claim until it contains a figure, a unit, and the condition it holds under.", "examples": [ "'Fast onboarding' versus 'Live in 4 days for teams under 50.' Only one of those can be quoted.", "A rival's page said 'enterprise-ready.' Yours said '99.95% uptime over the last 12 months.' Yours got cited.", "Number, unit, condition: 'cuts review time 30% on contracts under 20 pages.'" ], "watch_out": "A specific claim can be checked and can be wrong. Only publish numbers you can defend when a buyer asks.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 25, "symbol": "Pf", "name": "Proof", "group": "Signals", "group_key": "s", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "A claim with a source, a number, and a date attached. Models reproduce claims they can attribute.", "the_move": "For every number on the page, add where it came from and when. Unsourced numbers get dropped, not repeated.", "examples": [ "'Customers save 30%' with no source got dropped. 'Customers saved 30% on average across 42 accounts, Jan to Jun 2026' got repeated.", "A model reproduces claims it can attribute. No source, no attribution, no citation.", "Even 'our own data, 42 accounts, 2026' is a source. It beats nothing." ], "watch_out": "Proof dates. A number from 2023 with no update starts to count against you.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 26, "symbol": "Bo", "name": "Buyer outcome", "group": "Signals", "group_key": "s", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "The changed state the buyer ends up in, stated in their language rather than your feature list.", "the_move": "Name the outcome and what it currently costs them. Features answer 'what is it', outcomes answer 'why now'.", "examples": [ "Feature: 'automated reconciliation.' Outcome: 'close the month in 2 days instead of 9.' Buyers ask engines the second one.", "A page rewritten to name what the problem costs today ('about 20 hours a month of manual matching') started matching 'why now' questions.", "Buyers type their pain, not your feature name. Write the pain." ], "watch_out": "Outcomes you cannot measure turn into vague claims. Pair each outcome with a number or leave it out.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 27, "symbol": "Cm", "name": "Comparison", "group": "Signals", "group_key": "s", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "An honest account of how you differ from the named alternatives. Buyers ask engines to compare constantly.", "the_move": "Write the comparison yourself, fairly. If you don't, a review site writes it for you and gets cited instead.", "examples": [ "A buyer asks 'A versus B.' If you wrote the honest comparison, you get cited. If not, a review site with ads gets cited instead.", "A vendor's 'how we differ from X' page, which admitted where X was better, got cited more than X's own page.", "Name the alternatives. Pretending they do not exist just hands the answer to someone else." ], "watch_out": "Comparisons go stale fast when competitors change pricing or features. Date them and revisit every quarter.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 28, "symbol": "Pc", "name": "Pricing clarity", "group": "Signals", "group_key": "s", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "A number, a unit, and what changes it. 'Contact us' removes you from every price-shaped question.", "the_move": "Publish a starting price and the dimension it scales on. You can still qualify the deal on the call.", "examples": [ "'Contact us for pricing' removed a vendor from every 'how much does X cost' answer. Competitors with a number got all of them.", "'Starts at $200 a month, scales with seats' answered the question and still left room to negotiate.", "A buyer asked an engine for 'tools under $500 a month.' Only vendors with a published number could be on that list." ], "watch_out": "A public price anchors every negotiation. Publish the starting point and the scaling dimension, not the whole sheet.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 29, "symbol": "Rm", "name": "Rich media", "group": "Signals", "group_key": "s", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Whether your video and images are citable or opaque. A player an agent cannot read scores as absent.", "the_move": "Host video where the words are indexable and link it. A native player with no transcript is a blank space.", "examples": [ "A product demo in a custom player with no transcript is a gray box to every engine.", "The same demo hosted on YouTube, linked from the page, with a transcript, is citable twice.", "An image of a diagram with alt text that explains it beats the diagram alone." ], "watch_out": "Hosting on YouTube means YouTube gets some of the credit. Worth it; just link back hard.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 30, "symbol": "Fq", "name": "FAQ block", "group": "Signals", "group_key": "s", "time_to_effect": "days", "inhibitor": false, "what_it_is": "Question-shaped headings with self-contained answers — the exact shape retrieval wants to lift.", "the_move": "Use your real support tickets as the question list. Invented questions match no one's actual query.", "examples": [ "Support tickets showed buyers kept asking 'does it work with X?' That question, answered in two sentences, got lifted word for word.", "Made-up questions like 'Why is Acme the best?' matched nobody's real query and got ignored.", "Each answer should stand alone. Someone reading only that answer gets the whole thing." ], "watch_out": "A 40-question FAQ dilutes the page. Ten real questions beat forty invented ones.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 31, "symbol": "Df", "name": "Definition", "group": "Signals", "group_key": "s", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "A crisp 'X is Y' sentence near the top. Definitional queries are the highest-volume door into any category.", "the_move": "Define the term in one sentence before you explain anything. Models quote definitions almost verbatim.", "examples": [ "'X is a tool that does Y for Z' at the top of the page. Models quote that line almost word for word.", "A category page that explained for three paragraphs before defining the term never got cited for 'what is X.'", "Definitional queries are the front door. 'What is' questions outnumber 'which vendor' questions many times over." ], "watch_out": "If your definition disagrees with the common one, you get skipped, not corrected. Define the term the way the world does, then add your angle.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 32, "symbol": "Td", "name": "Teardown", "group": "Artifacts", "group_key": "a", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "A public, specific critique of a real example. Reliably the most-shared artifact in a technical niche.", "the_move": "Tear down something you did not make, name it, and be fair. Cheap to produce, hard to ignore.", "examples": [ "A founder tore down three competitor onboarding flows, screenshots and all, and was fair about it. It became the most-linked page on the site.", "Teardowns name real things, so they match real searches. People search for the thing you tore down.", "Cheap: one afternoon, a few screenshots, honest notes." ], "watch_out": "Naming competitors invites them to respond. Be accurate enough that you would be fine with them reading it.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 33, "symbol": "Bn", "name": "Benchmark", "group": "Artifacts", "group_key": "a", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "A number nobody else has because nobody else measured it. The single most citable object you can publish.", "the_move": "Measure something in your niche at n greater than fifty, publish the method, and repeat it on a schedule.", "examples": [ "'We timed setup across 60 tools in our category. Median: 11 days.' Nobody else had that number, so everyone cited it.", "Repeated each quarter, the citations renewed instead of fading.", "Small is fine: 50 samples, a stated method, a date." ], "watch_out": "Takes a quarter to produce and a quarter to get noticed. Slowest element on the table, and the most durable.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 34, "symbol": "Ds", "name": "Dataset", "group": "Artifacts", "group_key": "a", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "The raw rows behind a benchmark. It gets referenced long after the post that announced it is forgotten.", "the_move": "Publish the data at a stable URL with a license. A benchmark you cannot check is a claim, not a source.", "examples": [ "The raw rows behind the benchmark, posted as a CSV with a license, still referenced two years after the post was forgotten.", "A stable URL, not a Google Sheet that might get deleted, is what makes it citable.", "Someone else's analysis of your data is a citation you did not have to write." ], "watch_out": "Publishing data means people can find your mistakes. Check it twice and version it.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 35, "symbol": "Cs", "name": "Case study", "group": "Artifacts", "group_key": "a", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "One customer, one measured before and after. The proof artifact buyers and models both look for by name.", "the_move": "Lead with the number and the timeframe. Company logos are not evidence of anything.", "examples": [ "'Cut close time from 9 days to 2 in 60 days' in the first line. The logo wall below is decoration.", "Buyers ask engines 'who has done this for a company like mine?' Only a named, measured study answers that.", "One number, one timeframe, one company. Not 'significant results.'" ], "watch_out": "Customers have to approve it, and some will not. Get permission for the number, not just the logo.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 36, "symbol": "Rk", "name": "Runbook", "group": "Artifacts", "group_key": "a", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Ordered steps someone can follow without you. High utility, high save rate, endlessly quotable.", "the_move": "Write the one you already run internally. The version you use is better than the version you would write.", "examples": [ "The internal checklist your team already follows, cleaned up and posted. Highest save rate of anything you will publish.", "Ordered steps get quoted step by step: 'according to the vendor, first do X, then Y.'", "The version you actually use beats the version you would write for marketing." ], "watch_out": "A runbook that goes stale is worse than none, because people follow the wrong step. Date it and own it.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 37, "symbol": "Ca", "name": "Calculator", "group": "Artifacts", "group_key": "a", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "An interactive tool that returns a personal number. It earns links from people who will never read your posts.", "the_move": "Make it answer one question exactly, with no signup. Gating it converts a linkable asset into a form.", "examples": [ "'How much does manual invoicing cost you?' with three inputs and one number back. People who would never read a post share the number.", "No signup. The moment you gate it, it stops getting linked.", "It answers one question exactly. Not a 'suite.'" ], "watch_out": "Takes real build time, and a wrong formula gets screenshotted. Test the math before it ships.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 38, "symbol": "Sh", "name": "Screenshot", "group": "Artifacts", "group_key": "a", "time_to_effect": "days", "inhibitor": false, "what_it_is": "Visual evidence the thing exists and works. The cheapest media that materially raises trust.", "the_move": "Show the real screen with real data, cropped tight. Mockups read as mockups to everyone.", "examples": [ "A real dashboard with real numbers, cropped tight, beats a polished mockup every time.", "Buyers believe the thing exists when they see the screen. Models treat the alt text as the proof.", "Cheapest trust signal on the whole table: minutes, not days." ], "watch_out": "Real screens leak real data. Blur what you must; never fake the rest.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 39, "symbol": "Vo", "name": "Video", "group": "Artifacts", "group_key": "a", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Highest-trust format on every channel, and the only one that carries tone. Worthless to retrieval without its text.", "the_move": "Ship it with a transcript and a description written for search, not a one-line caption.", "examples": [ "A three-minute walkthrough carries tone, a face and trust that no page can. Without a transcript it is a blank to retrieval.", "Shipped with a transcript and a description written for search, the same video earned citations and views.", "The video is the trust layer; the text is the retrieval layer. You need both." ], "watch_out": "Expensive to redo. If a feature changes, the video is wrong until you re-shoot.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 40, "symbol": "Ni", "name": "Newsletter", "group": "Artifacts", "group_key": "a", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "The one audience nobody can take away — no ranking, no algorithm, no platform between you and the reader.", "the_move": "Send the same substance you publish. A newsletter that only announces posts teaches people to ignore it.", "examples": [ "Same substance as the blog, delivered directly, with no algorithm in between.", "A newsletter that only says 'new post!' trains readers to ignore it. One that gives the answer trains them to open it.", "When a platform changes its feed rules, the list is the only audience you keep." ], "watch_out": "Slow to grow and easy to neglect. Missing issues costs more than never starting.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 41, "symbol": "Tp", "name": "Template", "group": "Artifacts", "group_key": "a", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "A file someone can open and use today. It converts a reader into a user before any purchase exists.", "the_move": "Give away the artifact you would normally deliver in week one. It qualifies buyers faster than a demo.", "examples": [ "The spreadsheet you would hand a customer in week one, given away free. Readers become users before they are buyers.", "A template gets opened, used and forwarded: three touches from one file.", "It qualifies buyers. The ones who use it have the problem." ], "watch_out": "A template can look like the whole product. Make sure the paid thing is clearly more.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 42, "symbol": "Li", "name": "LinkedIn", "group": "Channels", "group_key": "c", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Graph-gated: reach comes from who already follows you, not from what people search. Documents outperform text.", "the_move": "Post the artifact as a document carousel. Text-only posts on this surface travel roughly nowhere.", "examples": [ "A document carousel of the runbook outperformed the same content as a text post by a wide margin.", "Reach depends on who already follows you. Strangers rarely find you by searching here.", "Good for recognition among people who will later type your name into an engine." ], "watch_out": "Almost nothing posted here gets cited by answer engines. It is a reach channel, not a retrieval channel.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 43, "symbol": "Yt", "name": "YouTube", "group": "Channels", "group_key": "c", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "Query-gated, so it compounds instead of decaying — and it is consistently among the most-cited domains in AI answers.", "the_move": "Title for the question people type. This is the one channel where a video keeps earning years later.", "examples": [ "A how-to video from two years ago still gets watched every week. Query-gated, so it compounds.", "YouTube pages are consistently among the most-cited domains in AI answers. The transcript is why.", "Title it for the question people type, not for a clever hook." ], "watch_out": "Slowest payoff on the table. Months to build, years to earn.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 44, "symbol": "Th", "name": "Threads", "group": "Channels", "group_key": "c", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Fast-moving, conversation-shaped, low production cost. Good for testing a claim before you build around it.", "the_move": "Post the sharpest line from a longer piece and see whether anyone argues. Free positioning research.", "examples": [ "Post the sharpest line from a long piece. If people argue, the line is doing work. If silence, rewrite it.", "Low production cost: a sentence and a screenshot.", "Good for testing a claim before you build a whole page around it." ], "watch_out": "Fast-moving and barely cited. Treat it as research, not distribution.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 45, "symbol": "Re", "name": "Reddit", "group": "Channels", "group_key": "c", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "Disproportionately cited by answer engines because the discussion is real. Also the fastest place to get punished for marketing.", "the_move": "Answer questions in your category as a practitioner for a month before you ever link to anything.", "examples": [ "Answer engines cite Reddit heavily because the discussion is real. One honest thread can outrank your whole site.", "A vendor answered questions as a practitioner for a month before ever linking. The links that came later were welcomed.", "The same vendor linking on day one got banned from the subreddit." ], "watch_out": "Zero tolerance for marketing. One wrong move and you are out, with the thread preserved forever.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 46, "symbol": "Gi", "name": "GitHub", "group": "Channels", "group_key": "c", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "A trusted domain where a README is a document engines read closely. Underused by non-developer companies.", "the_move": "Publish your dataset, schema, or template as a repo with a real README.", "examples": [ "A dataset posted as a repo with a real README got cited more than the blog post that announced it.", "Non-developer companies almost never use it, which is exactly why it is open.", "A README is a document on a trusted domain. Engines read it closely." ], "watch_out": "Needs upkeep. An abandoned repo with open issues signals an abandoned company.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 47, "symbol": "Mk", "name": "Marketplace", "group": "Channels", "group_key": "c", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "A cloud marketplace listing — an authoritative third-party page describing your product that you mostly control.", "the_move": "Treat the listing as a landing page, not a form. Same clarity, proof, and pricing rules apply.", "examples": [ "A cloud marketplace listing is a third-party page about your product that you mostly control: authority you cannot get on your own domain.", "Treated as a landing page (clear definition, proof, a price), it gets cited for 'where can I buy X' questions.", "Treated as a form, it sits empty and tells engines nothing." ], "watch_out": "Platform rules limit what you can say and how. Work inside them; do not fight them.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 48, "symbol": "Po", "name": "Podcast", "group": "Channels", "group_key": "c", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "Borrowed audience plus a transcript on someone else's trusted domain. Two assets from one hour.", "the_move": "Ask the host to publish the transcript. That page, not the audio, is what gets cited.", "examples": [ "One hour as a guest gives you two assets: the audience and a transcript page on someone else's trusted domain.", "Ask the host to publish the transcript. That page, not the audio, gets cited.", "A host's site often has more authority than yours. Borrow it." ], "watch_out": "No transcript, no retrieval value, and most hosts will not publish one unless asked.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 49, "symbol": "Dy", "name": "Directory", "group": "Channels", "group_key": "c", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Review and category sites. Engines lean on them heavily for comparison and shortlist questions.", "the_move": "Fill the profile out completely and get the first few reviews. Empty profiles read as an inactive vendor.", "examples": [ "'Best X tools' questions lean heavily on review sites. An empty profile reads as an inactive vendor.", "Five real reviews and a complete profile moved a vendor onto shortlist answers within a quarter.", "Pick the two directories engines actually cite in your category, not all twenty." ], "watch_out": "Review sites sell placement, and engines sometimes cite the paid lists. You are competing with budgets, not just quality.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 50, "symbol": "Ig", "name": "Instagram", "group": "Channels", "group_key": "c", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Where the person behind the company becomes legible. Weak for retrieval, strong for recognition.", "the_move": "Repurpose an existing carousel rather than making anything new. Near-zero cost, real coverage gain.", "examples": [ "Repurpose an existing carousel. Zero new work, real coverage of 'who is this person.'", "Weak for retrieval, strong for recognition: the face behind the name.", "People who see you here later search your name in an engine." ], "watch_out": "Almost never cited. Do not make anything new for it.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 51, "symbol": "Gp", "name": "Guest post", "group": "Channels", "group_key": "c", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "Your argument on a domain with authority you have not earned yet. Slow, but it moves the entity signal.", "the_move": "Pitch the specific claim you can prove, not a topic. Editors accept evidence, not subject areas.", "examples": [ "Your argument on a domain with authority you have not earned. Slow, but it moves the entity signal.", "Pitch the specific claim you can prove ('we measured X across 60 tools'), not the topic ('AI in finance').", "Editors accept evidence. They reject subject areas." ], "watch_out": "Quarters, not weeks. And the host site gets the citation, not you, so plan for the link back.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 52, "symbol": "Sv", "name": "Share of voice", "group": "Measurement", "group_key": "m", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "Of all answers to your category's questions, the share that names you. The closest thing to a scoreboard.", "the_move": "Fix the prompt set first. A share-of-voice number over a shifting prompt list measures nothing.", "examples": [ "Thirty buyer questions, asked monthly. You are named in 6 of 30: 20% share. Next quarter, 9 of 30: 30%.", "Freeze the question list first. Change the questions and the number means nothing.", "It is the closest thing to a scoreboard for this whole table." ], "watch_out": "Engines vary run to run. One check is noise; a trend across months is signal.", "audit_2026": null, "unit": "Percent of prompts in which you are named: named ÷ total prompts × 100.", "direction": "Higher is better. 20% means named in 6 of 30.", "alternates": "Share of citations (your citations ÷ all citations in the answers), average list position when named (lower is better), or volume-weighted share if some prompts matter more.", "part": null }, { "number": 53, "symbol": "Mt", "name": "Mentions", "group": "Measurement", "group_key": "m", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "How often you are named at all, cited or not. Being named without a link still shapes the buyer.", "the_move": "Track named-but-unlinked separately. It is usually your earliest sign that something started working.", "examples": [ "Named but not linked still shapes the buyer. 'Tools like Acme' with no link is still a mention.", "Track named-but-unlinked separately. It is usually the first sign something started working.", "Mentions come before citations. Citations come before referrals." ], "watch_out": "Easy to inflate with a loud week on social. Watch the trend, not the spike.", "audit_2026": null, "unit": "Count per period: how many answers named you across the frozen prompt set this month, or mention rate as a percent of prompts.", "direction": "Higher is better, with one check: a mention can be negative. Tag each one positive, neutral or negative.", "alternates": "Named-but-unlinked vs linked (track separately), mention rate per prompt, or sentiment-split counts.", "part": null }, { "number": 54, "symbol": "Ci", "name": "Citations", "group": "Measurement", "group_key": "m", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Answers that link you as a source. The strongest evidence a specific page earned its place.", "the_move": "Track which page gets cited, not just that you were. One page usually carries the whole result.", "examples": [ "Answers that link to you as a source. One page usually carries the whole result.", "Track which page got cited, not just that you were. Then make more pages like that one.", "A citation is the strongest evidence a specific page earned its spot." ], "watch_out": "Citations do not equal clicks. Many answers get read and never clicked.", "audit_2026": null, "unit": "Count per period: answers that link to your domain as a source, or citation rate as cited answers ÷ prompts.", "direction": "Higher is better.", "alternates": "Unique cited URLs (how many different pages earn citations), citation share (yours ÷ all citations), or position of your link in the source list.", "part": null }, { "number": 55, "symbol": "Ps", "name": "Prompt set", "group": "Measurement", "group_key": "m", "time_to_effect": "days", "inhibitor": false, "what_it_is": "The fixed list of questions you re-ask on a schedule. Without it, every other number here drifts.", "the_move": "Write thirty real buyer questions, freeze them, date the file, and change it only on purpose.", "examples": [ "Thirty real buyer questions, pulled from sales calls and support tickets, saved with a date, changed only on purpose.", "Without it, every other number on this row drifts and you cannot tell if you improved or the questions changed.", "Takes an afternoon. Pays off every month after." ], "watch_out": "Real questions are boring and specific. Resist the urge to add the ones you wish buyers asked.", "audit_2026": null, "unit": "Not a metric; the ruler. Its unit is a count of prompts (30) and a version date.", "direction": "Stable is good. Changes between periods should be zero unless deliberate, and every change gets a date.", "alternates": "Coverage: percent of buyer-journey stages or real sales-call questions the set represents. Prompt volume weighting if you have search data.", "part": null }, { "number": 56, "symbol": "Sb", "name": "Switchback", "group": "Measurement", "group_key": "m", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "Turning a change on and off in alternating time windows. The only clean test when you cannot split traffic by visitor.", "the_move": "Run windows long enough to clear the crawl cycle. A week is usually the floor.", "examples": [ "Turn the FAQ block on for a week, off for a week, on again. Compare citations in the on-weeks to the off-weeks.", "The only clean test when you cannot split traffic by visitor, and with engines you cannot.", "A week is the floor. Shorter windows do not clear the crawl cycle." ], "watch_out": "Slow and fiddly. Only worth running on changes big enough to matter.", "audit_2026": null, "unit": "Lift in percentage points: citation rate in on-weeks minus citation rate in off-weeks.", "direction": "Higher positive is better. Near zero means the change didn't matter. Negative means undo it.", "alternates": "Ratio (on ÷ off), raw citation count delta. Minimum one week per block and at least two on/off cycles before you believe it.", "part": null }, { "number": 57, "symbol": "Ar", "name": "AI referrals", "group": "Measurement", "group_key": "m", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "Sessions arriving from an answer engine. Small in volume, and consistently the highest-intent traffic on the site.", "the_move": "Segment them out. Judged against organic averages they look like a rounding error instead of your best cohort.", "examples": [ "Sessions arriving from an answer engine. Small, often under 2% of traffic, and the highest-intent cohort on the site.", "Judged against organic averages they look like a rounding error. Segmented out, they convert several times better.", "Set up the segment once; the story tells itself after that." ], "watch_out": "Attribution is leaky. Many AI visits arrive as 'direct' and never get counted.", "audit_2026": null, "unit": "Sessions per period from AI referrers (chatgpt.com, perplexity.ai, gemini.google.com, copilot.microsoft.com and the rest), and that segment's conversion rate as a percent.", "direction": "Higher is better on both. Expect under 2% of sessions; the proof is the conversion rate running several times organic.", "alternates": "Share of total sessions, assisted conversions, revenue per AI session. Many AI visits arrive as 'direct' and are not counted here.", "part": null }, { "number": 58, "symbol": "Bs", "name": "Brand search", "group": "Measurement", "group_key": "m", "time_to_effect": "quarters", "inhibitor": false, "what_it_is": "People searching your name directly. It rises when awareness works, whatever the attribution says.", "the_move": "Watch branded query volume as the lagging check on every campaign that promised awareness.", "examples": [ "People typing your name into Google. It rises when awareness works, whatever the attribution says.", "A campaign that 'did not convert' still doubled branded searches the next month. That is the lagging proof.", "Free to track; it is in Search Console already." ], "watch_out": "Lags by months, and it cannot tell you which effort caused the rise.", "audit_2026": null, "unit": "Branded search impressions and clicks per month from Search Console, or a Google Trends index from 0 to 100.", "direction": "Higher is better, and it lags everything else by weeks to months.", "alternates": "Count of distinct branded queries, direct traffic as a proxy, or brand-name mention volume across the web.", "part": null }, { "number": 59, "symbol": "Im", "name": "Impressions", "group": "Measurement", "group_key": "m", "time_to_effect": "weeks", "inhibitor": false, "what_it_is": "How many times you were shown. Necessary context, and worthless as a goal on its own.", "the_move": "Only ever read it next to a downstream number. Alone it rewards volume over reach that matters.", "examples": [ "How many times you were shown. Context, not a goal.", "10,000 impressions and 2 citations is a problem. 500 impressions and 40 citations is a win.", "Only read it next to a downstream number." ], "watch_out": "Rewards volume over relevance. Chasing it alone makes slop.", "audit_2026": null, "unit": "Count: times you were shown, as Search Console impressions or an AI-tracking tool's appearances.", "direction": "Neutral on its own. Read it only as a ratio: citations per 100 impressions. 10,000 impressions and 2 citations is a problem; 500 and 40 is a win.", "alternates": "Impression share within your prompt set, or citations ÷ impressions as the one number worth watching.", "part": null }, { "number": 60, "symbol": "Bc", "name": "Blocked crawler", "group": "Inhibitors", "group_key": "x", "time_to_effect": "days", "inhibitor": true, "what_it_is": "A disallow rule aimed at scrapers that also shuts out the agents feeding the engines you want to appear in.", "the_move": "Audit robots.txt line by line. This is the highest-cost, lowest-effort mistake on the whole table.", "examples": [ "A 'block all bots' rule meant for scrapers also blocked the crawlers that feed ChatGPT and Perplexity. Months invisible, five-minute fix.", "Highest cost, lowest effort mistake on the whole table.", "Check: open your robots.txt and read it line by line, out loud." ], "watch_out": "Unblocking everything lets the scrapers back in. Allow by name.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 61, "symbol": "Jo", "name": "JS-only", "group": "Inhibitors", "group_key": "x", "time_to_effect": "weeks", "inhibitor": true, "what_it_is": "Content that only exists after the script runs. Many retrieval pipelines never get that far.", "the_move": "Server-render anything you want quoted. Everything else can stay client-side.", "examples": [ "The pricing table appeared only after the script ran. To most crawlers the page was a header and nothing else.", "'View source' showed none of the claims the page made.", "Server-render the pages you want quoted; keep the app client-side." ], "watch_out": "Fixing it can mean a rebuild. Prioritize the five pages that matter.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 62, "symbol": "Ga", "name": "Gated", "group": "Inhibitors", "group_key": "x", "time_to_effect": "weeks", "inhibitor": true, "what_it_is": "Your best evidence sitting behind a form. Invisible to every engine and most buyers.", "the_move": "Ungate the proof and gate the tailored thing instead — the audit, the walkthrough, the custom number.", "examples": [ "Your best proof behind an email form: invisible to every engine and most buyers.", "Ungate the proof; gate the tailored thing instead, the audit, the custom number, the walkthrough.", "A gated benchmark earned 40 leads. The same benchmark ungated earned hundreds of citations and more leads." ], "watch_out": "Sales teams like the form. Show them the citation count before you take it away.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 63, "symbol": "Pl", "name": "Paywall", "group": "Inhibitors", "group_key": "x", "time_to_effect": "weeks", "inhibitor": true, "what_it_is": "Same failure as a gate, with a card attached. The content may be excellent and it cannot be cited.", "the_move": "Leave the definition, the method, and the headline number outside the wall.", "examples": [ "Same failure as a gate, with a card attached. The content may be excellent; it cannot be cited.", "Leave the definition, the method and the headline number outside the wall.", "Paid readers stay for depth, not for the first paragraph." ], "watch_out": "Giving away the headline number can feel like giving away the product. It is not. It is the ad.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 64, "symbol": "Jw", "name": "Jargon wall", "group": "Inhibitors", "group_key": "x", "time_to_effect": "days", "inhibitor": true, "what_it_is": "Language only an insider parses. It fails both the model matching intent and the buyer scanning for fit.", "the_move": "Read the first paragraph aloud to someone outside your category. Rewrite whatever they ask about.", "examples": [ "'Composable, headless orchestration layer' fails the model matching intent and the buyer scanning for fit.", "Read the first paragraph to someone outside your industry. Rewrite whatever they ask about.", "Insiders will not mind plain words. Outsiders cannot get past the jargon ones." ], "watch_out": "Some jargon is the search term. Keep the term, explain it in the next sentence.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 65, "symbol": "Vc", "name": "Vague claim", "group": "Inhibitors", "group_key": "x", "time_to_effect": "days", "inhibitor": true, "what_it_is": "'Industry-leading', 'seamless', 'enterprise-grade'. Unfalsifiable, so unquotable — nothing to attribute.", "the_move": "Replace each one with a number and a source, or delete the sentence entirely.", "examples": [ "'Industry-leading,' 'seamless,' 'enterprise-grade': nothing to attribute, so nothing to quote.", "Replace each one with a number and a source, or delete the sentence.", "A page with zero vague claims reads shorter and gets cited more." ], "watch_out": "Deleting them shrinks the page. Good. That is the point.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 66, "symbol": "Du", "name": "Duplicate", "group": "Inhibitors", "group_key": "x", "time_to_effect": "weeks", "inhibitor": true, "what_it_is": "The same content at several addresses, splitting whatever authority any single version might have had.", "the_move": "Pick the canonical URL and point everything else at it, including the syndicated copies.", "examples": [ "Same article at three addresses, splitting authority three ways. None of them won.", "Pick the canonical and point the others at it, including the syndicated copies.", "Often invisible to the author. A crawl finds it in minutes." ], "watch_out": "Consolidating can drop traffic briefly while engines re-learn. It recovers; hold steady.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 67, "symbol": "Op", "name": "Orphan page", "group": "Inhibitors", "group_key": "x", "time_to_effect": "weeks", "inhibitor": true, "what_it_is": "A page nothing links to. It exists, and for discovery purposes it does not.", "the_move": "Run a crawl and link every orphan from somewhere real, or delete it.", "examples": [ "A page nothing links to. It exists, and for discovery purposes it does not.", "A crawl found 60 orphans on a 400-page site, including the best case study.", "Link each orphan from somewhere real, or delete it." ], "watch_out": "Linking junk pages makes them discoverable. Delete first, link second.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 68, "symbol": "Sl", "name": "Slop", "group": "Inhibitors", "group_key": "x", "time_to_effect": "quarters", "inhibitor": true, "what_it_is": "High-volume thin content. It buries your own good pages and trains readers that your name means noise.", "the_move": "Cut the volume and raise the floor. Publishable is a bar, not a synonym for finished.", "examples": [ "Fifty thin AI-written posts buried the three good ones and taught readers the name meant noise.", "Cut the volume, raise the floor. Publishable is a bar, not a synonym for finished.", "One benchmark outperformed fifty posts on citations alone." ], "watch_out": "Takes a quarter for engines to re-trust a domain after a slop phase. Deleting is faster than waiting.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null }, { "number": 69, "symbol": "Nd", "name": "No date", "group": "Inhibitors", "group_key": "x", "time_to_effect": "days", "inhibitor": true, "what_it_is": "Undated content reads as possibly stale, and stale gets skipped on anything time-sensitive.", "the_move": "Add a published and updated date sitewide. It is a template change, not a content project.", "examples": [ "Undated content reads as possibly stale, and stale gets skipped on anything time-sensitive.", "It is a template change (add published and updated dates sitewide), not a content project.", "Old but dated beats new but undated." ], "watch_out": "Adding dates exposes how old some pages are. Update or prune the worst ones.", "audit_2026": null, "unit": null, "direction": null, "alternates": null, "part": null } ] }