Skip to main content

WISLR Launches AI Channel Analytics: First-Party Reporting and Attribution for AI & LLM Traffic

WISLR AI Channel Analytics is now generally available. Server-level capture of AI training crawls, conversation citations, real-user referrals, and AI-attributed sales across ChatGPT, Gemini, Claude, Perplexity, and every other LLM bot. First 30 days free, then from $200/month, live same day. Use cases for ecommerce, publishers, B2B lead gen, agencies, and site migrations.

Composite of four WISLR AI Channel Analytics dashboard views: LLM-attributed revenue by month, a year-long daily activity heatmap, momentum growth scores for leads, referrals, citations, and training, and top days for training crawls by AI bot
Composite of four WISLR AI Channel Analytics dashboard views: LLM-attributed revenue by month, a year-long daily activity heatmap, momentum growth scores for leads, referrals, citations, and training, and top days for training crawls by AI bot

AI Channel Analytics Is Now Generally Available

WISLR AI Channel Analytics is live. It is a server-level reporting and attribution dashboard for the AI and LLM traffic that browser-based analytics tools measure poorly or miss entirely. It captures every request hitting your domain at the edge, classifies AI activity by verified bot fingerprint, and reports training crawls, conversation citations, real-user referrals, and AI-attributed sales in one live view across ChatGPT, Gemini, Claude, Perplexity, and every other LLM bot.

The first 30 days are free on every plan. After that, pricing starts at $200 per month for sites up to 1,000,000 monthly sessions, with no setup fee. Deployment happens the same day, and on most sites behind an edge CDN service the dashboard is live in minutes. The full product page is at AI Channel Analytics.

Free for 30 days
Then a flat subscription starting at $200 per month
WISLR AI Channel Analytics
2.5x to 5x
More AI-referred sessions captured at the server level than GA4 and other browser-based tools report
WISLR device-by-device testing
100%
First-party data, captured at the server level with no JavaScript dependency and no cookie consent gaps
Edge request capture
Same day
Setup and deployment, including bot fingerprinting and revenue attribution, live in minutes on most sites behind an edge CDN service
Deployment

Why We Built It

We built this because the tools most teams rely on cannot see most of the AI channel, and the reasons are baked into how those tools work. Browser-based analytics tools, GA4 included, only log a request when a JavaScript tag fires in a real browser. Two of the most important AI behaviors never open a browser at all. Training crawlers read your pages server to server. Citation fetches happen when an AI assistant pulls a page mid-conversation to answer a live question, and the answer renders inside the chat. Neither one fires a tag, so a browser-based report never sees them.

The third behavior, a real user clicking through from an AI answer, is partially visible but badly undercounted. Our device-by-device testing across ChatGPT, Gemini, and Claude found that mobile AI apps render outbound links in isolated WebViews that strip the referrer, and Gemini and Claude pass no attribution signal on most platforms. The result is a 2.5x to 5x undercount of AI-referred sessions. The full methodology is in LLM Traffic Is a Blind Spot in Your Analytics.

The other tool category, AI-visibility and Share of Voice platforms, works from the opposite direction. Those tools query AI engines with a sample basket of prompts and report an estimated share of AI answers. That approach is useful for competitive benchmarking, but the number is an estimate built from prompts an analyst picked, and it never touches your site. A score tells you AI mentions your brand; it cannot tell you which pages AI reads, who it sends to your site, or what those visitors buy.

AI Channel Analytics fills the space between the two with verified, first-party measurement of what the AI channel actually does on your site, starting at the first crawl and following it all the way to revenue.


Four Signals, Not One Number

The dashboard is organized around the signal model we published in LLM Traffic Monitoring: The Three Signals, extended at launch with a fourth signal that closes the loop to revenue.

Signal 1
Training crawls
GPTBot, ClaudeBot, PerplexityBot, Google-Extended, Bytespider, and CCBot read your pages on a schedule to feed the next model. This is the signal that decides what AI knows about your brand.
Signal 2
Conversation citations
ChatGPT-User, Claude-User, and Perplexity-User fetch pages mid-chat to answer real-user questions. It is the closest thing to a vote of confidence AI produces, and it never appears in GA4.
Signal 3
Real user referrals
Visitors who clicked an AI citation and landed on your site. These show up partially in browser-based tools, but they are undercounted by 2.5x to 5x because of mobile WebView and missing referrer behavior.
Signal 4 · New at launch
Sales and leads
AI-referred visitors who become revenue: orders, form fills, and booked calls, matched back to their AI source by our attribution engine. This is the signal that shows whether the channel is actually paying off.

Each signal comes from a different AI behavior, hides from browser-based analytics in a different way, and calls for different work. Rolling them into one traffic number buries all of that, which is why the dashboard keeps them separate.


What Ships in the Dashboard

The launch dashboard includes five headline report views, each answering a question teams actually bring to the data.

  • Momentum. A growth score per signal over the trailing period, so you can see at a glance whether training, citations, referrals, and revenue are accelerating or stalling.
  • Top Days for Training. The days AI training bots crawl you hardest, per engine, so publishing and update schedules can line up with when models are actually reading.
  • LLM-Attributed Revenue. Revenue matched to AI sources, sliced by vendor: OpenAI, Gemini, Copilot, Perplexity, Anthropic, and xAI, each segment traceable to the orders behind it.
  • Top Referral Destinations. The pages where AI chat referrals land most over the last 30 days, with training, citation, and referral counts per page.
  • Daily Activity. A year-long heatmap of leads, referrals, citations, and training crawls, one square per day, showing how AI-channel activity trends over time.

Behind the headline views sit the working reports: AI bot crawl coverage by engine and page, a fetched-content leaderboard, conversion funnels by AI source, revenue attribution with time-to-purchase by platform, product and buyer-level detail, content freshness for AI training, and pages crawled but never cited. Each view is designed to leave you with something to do next, not just a number to watch.


How the Data Is Captured

The numbers come from three mechanisms, all of them running at the server level.

  • Server-level request logging. Every HTTP request hitting your domain is captured at the edge, including requests from AI bots that never execute JavaScript. There is no tag to fire and no dependence on cookie consent.
  • AI bot fingerprinting. User-agent matching plus verified IP range checks classify every bot by platform with high confidence, separating GPTBot from ChatGPT-User from an impersonator.
  • Order and lead attribution. AI-referred visits are matched to order confirmations and form fills for verified revenue attribution, with probabilistic matching covering the rest.

Deployment depends on how your site is served. Sites running their own edge CDN service activate directly, usually within minutes. Sites on platform-managed edges, Shopify, Wix, Squarespace, and BigCommerce, can be enabled with an added edge service. The site tester on the product page checks your domain and tells you which path applies before you commit to anything.


Use Cases

The same four signals read differently depending on what your site sells. These are the patterns we built the launch dashboard around.

Ecommerce: prove the AI channel pays, then feed it

A store sees ChatGPT-referred sessions climbing but GA4 shows the channel as a rounding error. Server-level capture surfaces the real referral volume, and order attribution ties purchases back to their AI source. From there you can get into the working questions: which products AI is promoting, which AI platform sends buyers versus browsers, and how time-to-purchase differs by engine so email cadences and retargeting windows can match how each platform's shoppers behave. On one mid-market store we measured during the pilot, AI-attributed revenue reached six figures over a five-month window while the browser-based report showed a fraction of it.

Publishers and content teams: see what AI is reading and quoting

Training coverage shows which sections of the site each AI engine reads and which it ignores. The fetched-content leaderboard shows which articles AI assistants pull mid-conversation to answer real questions. The gap between the two, pages crawled but never cited, is the editorial to-do list: content AI has absorbed but does not consider quotable, usually fixable with clearer titles, structured data, and FAQ-style copy.

B2B and service businesses: attribute the leads AI sends

For sites where the conversion is a form fill or a booked call, the sales-and-leads signal matches those conversions back to their AI source. Buyers increasingly research a vendor with an AI assistant, then arrive with no referrer and get logged as direct traffic. Server-level attribution reclassifies that journey, so the channel that actually sourced the lead gets credit in the pipeline review.

Agencies: report a channel your clients cannot see anywhere else

Client reporting on AI visibility usually stops at a Share of Voice score. A live dashboard of training crawls, citations, referrals, and attributed revenue turns the AI channel into a line item with numbers behind it, and gives the content and technical recommendations you already make a measurable before-and-after.

Site migrations and replatforms: protect AI coverage through the change

Redirects and template changes that preserve Google rankings can still break AI crawl patterns. Watching per-engine crawl coverage before, during, and after a migration shows immediately whether GPTBot, ClaudeBot, and Google-Extended followed the move or dropped sections of the site, while there is still time to fix it.

Content freshness: keep what AI has on file accurate

Whatever was on a page the last time a training bot fetched it is what AI tools have on file. The freshness report lists the URLs training bots crawl most and the date each engine last read them, so pricing, specs, and policy pages can be reviewed on the schedule that actually matters, which is the model's schedule rather than the CMS's.

Pricing

The first 30 days are free on every plan. After that, a single flat monthly subscription scales with monthly session volume. Every plan includes the live dashboard and edge data capture, and there is no setup fee.

Monthly sessions Price per month
Up to 1,000,000 $200
Up to 3,000,000 $250
Up to 6,000,000 $550
Up to 10,000,000 $750

For teams that want a partner to act on what the dashboard surfaces, a separate Senior Strategist engagement is available as a quarterly retainer starting at $6,250. The dashboard stands on its own; the strategist engagement is for teams that also want help interpreting the data and shipping the changes it points to.


How It Fits Next to the Tools You Already Have

AI Channel Analytics does not replace your analytics stack, and it is not another Share of Voice score.

Keep GA4, or whichever browser-based tool you run, for the channels it was built to measure. AI Channel Analytics covers the channel where the browser-based model structurally breaks: it sees the crawls and citation fetches that never open a browser, and it recovers the referral and revenue volume that WebViews and stripped referrers hide.

Keep a Share of Voice tool if off-site competitive benchmarking matters to you. It answers a different question: how often AI mentions your brand across a sampled set of prompts. AI Channel Analytics answers what happened on your site, with a verified request log rather than an estimate. The two work fine side by side, but only one of them can put a dollar figure on the channel. The KPI framework behind the dashboard is laid out in AI Performance Metrics: The Seven KPIs and the measurement model in LLM Traffic Monitoring: The Three Signals.


Getting Started

Setup happens the same day, and the free 30 days start as soon as data starts flowing.

  1. Test your site. The tester on the product page checks whether your domain runs on its own edge CDN service (direct activation) or a platform-managed edge like Shopify, Wix, Squarespace, or BigCommerce (enabled with an added edge service).
  2. Schedule setup. A short call covers deployment and confirms what the dashboard will track for your site.
  3. Watch the first signals land. Training crawls usually appear within hours; citations, referrals, and attributed revenue fill in as the log builds.
See AI Channel Analytics

Frequently Asked Questions

What is WISLR AI Channel Analytics?

AI Channel Analytics is a server-level reporting and attribution dashboard for AI and LLM traffic. It captures every request hitting your domain at the edge, classifies AI bots by user agent and verified IP range, and reports four signals in one live view: training crawls, conversation citations, real-user referrals, and AI-attributed sales and leads across ChatGPT, Gemini, Claude, Perplexity, and every other LLM bot.

How is AI Channel Analytics different from GA4 and other browser-based analytics?

Browser-based tools only log a request when a JavaScript tag fires in a real browser. AI training crawls and citation fetches never open a browser, so those two signals are invisible to GA4, Adobe, Mixpanel, and every tool built on the same model. Real-user referrals from AI are partially visible but undercounted by 2.5x to 5x because mobile AI apps strip referrers. WISLR captures at the server level, so every request is recorded whether or not a browser was involved.

How is AI Channel Analytics different from AI-visibility and Share of Voice tools?

Share of Voice tools query AI engines with a sample basket of prompts and report an estimated share of AI answers. That is useful directional benchmarking, but it is probabilistic and stops at the answer. AI Channel Analytics measures what actually happened on your site: which pages AI read, which pages it fetched to answer live questions, which visitors it sent, and which of those visitors became revenue.

How much does AI Channel Analytics cost?

The first 30 days are free on every plan. After that, pricing is a flat monthly subscription that scales with monthly session volume: $200 per month for sites up to 1,000,000 sessions, $250 up to 3,000,000, $550 up to 6,000,000, and $750 up to 10,000,000. Every plan includes the live dashboard and edge data capture, with no setup fee.

How long does setup take and what does my site need?

Setup and deployment happen the same day, and the dashboard is typically live in minutes. Sites on their own edge CDN service can be activated directly. Sites on platform-managed edges such as Shopify, Wix, Squarespace, and BigCommerce can be enabled with an added edge service. The site tester on the AI Channel Analytics page checks your domain and tells you which path applies.

Does AI Channel Analytics replace GA4?

No. It runs alongside whatever analytics stack you already use. GA4 remains useful for the channels it was built to measure. AI Channel Analytics covers the AI channel specifically, where the browser-based model breaks down, and gives you the training, citation, referral, and revenue data those tools cannot see.