Rank tracking told you where you stood. This tells you if you exist.
AI assistants answer a growing share of the queries you used to win in blue links. The discipline that replaces rank tracking starts at your own server: who reads you, what they read, and whether you show up in the answers.
From positions to presence.
Rankings measured positions
For twenty years the deliverable was a position on a results page, and the click curve did the rest. Every SEO tool is built around that unit of measurement.
AI answers skip the page
When an assistant answers directly, there is no position - there is inclusion or absence. You are either part of the answer, cited under it, or invisible. None of that shows up in a rank tracker.
Crawls are the new leading indicator
Before an engine can include you, it has to read you. Which AI bots fetch which pages, and how often, is the earliest signal of AI visibility you can measure - and it lives in your server traffic, not in a SERP.
Four workflows that make AI visibility operational.
Crawl-map your content
Filter AI crawler traffic by URL to see which sections retrieval bots actually fetch. Heavily crawled clusters are what AI engines treat you as an authority on; ignored clusters are where visibility work starts.
Separate Google from AI
One view splits Googlebot from GPTBot, ClaudeBot, and PerplexityBot, so crawl-budget conversations stop conflating indexing you want with scraping you may not.
Verify robots.txt decisions
Ship a disallow rule, then watch the log to see who actually honored it. The built-in probe confirms whether blocks are enforced or being bypassed by a cache.
Track the answers themselves
Paid plans ask the major engines your buyers' questions and record whether you are mentioned, cited, and how you compare - AI Visibility picks up where position tracking ends.
Not all crawling is the crawling you want.
On many WordPress sites, bot requests now rival or exceed human ones. Some of that is Googlebot keeping your index fresh - crawling you want. Some is AI training crawlers reading your archive for a model you will never hear from again. Treating those as one “bot traffic” number hides the decision you actually face: welcome the crawls that send humans back, and meter or block the ones that don't. The bot directory documents each crawler's operator, purpose, and robots token, so every disallow line you ship is a decision, not a guess.
The monthly AI brief: four numbers that survive scrutiny.
AI crawl volume, by operator
Which AI companies read the site this month and how the volume trends. This is the slide that lands hardest with stakeholders, because it is their own server data - not an industry statistic.
Most-fetched content
The pages retrieval bots pull hardest, read as a machine-written verdict on topical authority. Movement here predicts answer inclusion before any visibility metric moves.
Mention and citation rate
For the tracked buyer prompts: how often the brand appears in answers, how often it is cited, and the share-of-voice delta against the named competitor. Trend, not snapshot.
AI-referred sessions and quality
Attributed sessions from assistants with engagement against site average - presented honestly as a floor, since referrer stripping hides part of the channel. Quality per session is the number that justifies the work.
SEO team FAQ.
Does blocking AI training bots hurt my Google rankings?
Not if you block the right tokens. GPTBot, ClaudeBot, and CCBot have nothing to do with Google Search indexing, and Google-Extended is specifically designed to control AI training use without affecting Googlebot. Viz keeps search engine crawlers in a separate, never-blockable category, so an AI block cannot touch your SEO by construction.
What is Google-Extended and should I block it?
Google-Extended is a robots.txt control token: disallowing it tells Google not to use your content for its AI training, while Googlebot and normal indexing continue untouched. Whether to block it is a business decision - it trades possible inclusion in AI features against keeping your content out of model training. The useful first step is measuring how much Google AI-related crawling you actually receive.
How do I report AI visibility to a client or executive?
With crawl evidence instead of anecdotes: which AI engines read the site, which pages they fetch most, how volume trends month over month, and - where referrers survive - how many humans AI assistants send back. That turns "we think AI matters" into a chart with dates on it.
Can I see whether an AI engine crawled a specific page?
Yes. The request log records every request with URL, user agent, time, and classification, so filtering to a URL shows every bot that fetched it and when. That is useful both for proving reach and for investigating why a page appears - or fails to appear - in AI answers.
Is answer engine optimization just SEO with a new name?
The fundamentals carry over - crawlability, authority, content that answers real questions - but the measurement and the win condition changed. SEO optimizes for position on a page the user scans; AEO optimizes for inclusion in an answer the user trusts. Positions come in tens, inclusions in twos and threes, and there is no impression report from the engines. The skills transfer; the scoreboard is new, and it starts with data only your server has.