What is AI-assisted analytics, exactly?

AI-assisted analytics is the operating model where an AI assistant (e.g. Claude) reads your analytics data, filters and summarises it, and helps you investigate trends — but every decision and every change remains in the operator's hands. WebmasterID exposes the analytics surface; Claude reads it through Model Context Protocol (MCP); the operator reviews and acts.

Is the AI making decisions for me?

No. The AI assistant reads data and prepares analysis. It does not deploy code, does not change your subscription, does not modify your sites. Every action is operator-approved. The product is designed around this hard rule.

Why MCP and not a custom API?

Model Context Protocol is an open standard for letting AI assistants subscribe to external data and tools. Using MCP means the same analytics surface works with Claude Desktop, Claude Code, the VS Code Claude extension, and any future MCP client without a per-vendor adapter. It also keeps the integration auditable: every MCP call lands in your workspace audit log.

Do you train any AI models on my data?

No. WebmasterID is a privacy-first analytics platform. Your event data is yours; we never train shared models on it, we don't sell or share it with third parties, and we never send raw analytics to model providers. When you ask Claude to read your data through MCP, the request goes directly from your Claude client to our MCP endpoint — your operator session controls what is read.

How is this different from a generic analytics dashboard?

Generic dashboards are read-mostly UIs you scroll through. AI-assisted analytics is read-and-investigate: the operator asks Claude a question grounded in real data, Claude reads the relevant slice through MCP, and the conversation surfaces patterns you would have missed scrolling. WebmasterID still ships the dashboard UI — AI-assisted is an additional surface, not a replacement.

Which Claude clients are supported?

Any client that speaks MCP today. The /settings/mcp page in the dashboard ships ready-to-paste configuration snippets for Claude Desktop, Claude Code (CLI), and the VS Code Claude extension. The MCP endpoint is workspace-scoped and authenticated by a Bearer API key.

AI-assisted analytics

Read traffic, attribution, and AI visibility with Claude alongside you

WebmasterID is an AI-native analytics platform built for operators who already use Claude (or another MCP client) to investigate, summarise, and explain data. Privacy-first by design, operator-approved by default, no autonomous decisions.

Start free See pricing How MCP works

What it is

Analytics you can talk to — without giving up control

The shift from manual dashboards to AI-assisted workflows is not about replacing operators. It is about giving operators a reading partner: someone who knows the data shape, fetches the slice that matters, and explains what changed. WebmasterID provides the data; Claude reads it through MCP; you decide.

Read analytics data faster

Claude can answer questions like 'which AI assistant sent the most traffic this week' or 'which pages received human referrals from ChatGPT' by fetching the exact slice through MCP, instead of you scrolling through dashboards.

Analyse traffic patterns

Pattern detection on AI crawler activity, human referrals, attribution categories, and event timelines. Claude summarises what changed week-over-week and surfaces what's worth looking at.

Analyse attribution

Understand which sources sent the traffic, which were AI-related, and which landing pages received attributed traffic. Confidence labels per row — Claude can explain the directional limits, not gloss over them.

Analyse AI visibility

See which AI bots crawl your site, which AI assistants send humans to it, and how that changes over time. AI visibility is what AI-native operators ask Claude to monitor first.

Detect anomalies

Sudden drops in traffic, new AI crawler appearances, attribution shifts — Claude can be asked to flag them. The MCP surface is small and deterministic; the analysis happens in conversation with the operator.

Suggest actionable improvements

Claude reads your WebmasterID Agent recommendations through MCP, opens the file referenced in the repo mapping, and prepares a diff. You review and ship. No autonomous deploys.

Why this matters

Modern teams replace fragmented dashboards with AI-assisted workflows

Operators evaluating an analytics platform in 2026 increasingly ask: 'Will it work with Claude?' before asking about specific chart types. WebmasterID is built around that question.

One reading partner across many sites

Instead of switching between dashboards for each project, ask Claude — through MCP — to read the active workspace's data and summarise what changed.

A conversation is faster than a click trail

The dashboard is good for the curated view. The conversation is faster for the ad-hoc investigation: 'why did pageviews on /pricing drop?' is one prompt, not five clicks.

Audit trail by default

Every MCP read is logged in the workspace audit log. You see what was read, when, and by which API key. AI-assisted analytics with full provenance — not a black box.

How MCP fits in

Model Context Protocol is the integration layer

MCP is an open standard. The WebmasterID Agent exposes an MCP server; Claude (or any MCP client) connects with a workspace-scoped Bearer key and reads the allow-listed surface.

1. Configure once. In the dashboard, go to /settings/mcp and copy the snippet for your Claude client. The page ships ready-to-paste config for Claude Desktop, Claude Code, and the VS Code Claude extension.
2. Restart Claude. The WebmasterID tools appear in the available-tools list. The MCP surface is read-only; there are no write tools today.
3. Ask analytics questions. "Which AI bots fetched /pricing this week?" — Claude reads the relevant slice through MCP and answers. You see the result; the audit log sees the read.
4. Review and act. When Claude proposes a change (a recommendation, a code diff, a copy update), you review and ship. There is no autonomous deploy. The operator is the executor.

Safety model. MCP gives Claude read access and prompt-preparation capability. Code changes, plan changes, and configuration changes are reviewed and deployed by the operator. There is no auto-deploy, no auto-merge, no hidden execution.

Companion tool

Agent-WebmasterID — the free operator companion

Agent-WebmasterID is a free, optional companion tool that runs alongside WebmasterID. It is an investigation helper for operators who want an additional AI workflow layer for monitoring projects and reviewing analytics.

Optional companion
Use WebmasterID without it. Or layer it on for an extra surface that summarises and explains the analytics you already record.
MCP-compatible workflows
Speaks the same MCP surface as Claude. The investigation shape is portable across clients.
Operator-first
Built as an analytics companion, not an autonomous agent. No hidden tracking, no aggressive automation, no black-hat workflows.
Free
Free companion utility. Pricing covers WebmasterID Core + Agent on the dashboard side; the companion does not add a seat charge.

For the conversation about custom integrations, write to info@helperg.com.

What we don't claim

An honest list of what AI-assisted analytics is not

The market is full of overclaim. We try not to add to it.

Not autonomous decision-making
Claude reads and suggests. The operator decides and acts.
Not guaranteed SEO growth
Better visibility into AI traffic does not mean guaranteed rankings. WebmasterID surfaces signal; the work still belongs to the team.
Not guaranteed revenue
Attribution is directional analytics from server-side detection — useful for trends, not a substitute for paid-ad attribution.
Not a fully autonomous business operation
The platform deliberately does not deploy, charge, refund, or message your customers on its own. AI assists the operator; the operator runs the business.

FAQ

AI-assisted analytics, in plain English

What is AI-assisted analytics, exactly?: AI-assisted analytics is the operating model where an AI assistant (e.g. Claude) reads your analytics data, filters and summarises it, and helps you investigate trends — but every decision and every change remains in the operator's hands. WebmasterID exposes the analytics surface; Claude reads it through Model Context Protocol (MCP); the operator reviews and acts.
Is the AI making decisions for me?: No. The AI assistant reads data and prepares analysis. It does not deploy code, does not change your subscription, does not modify your sites. Every action is operator-approved. The product is designed around this hard rule.
Why MCP and not a custom API?: Model Context Protocol is an open standard for letting AI assistants subscribe to external data and tools. Using MCP means the same analytics surface works with Claude Desktop, Claude Code, the VS Code Claude extension, and any future MCP client without a per-vendor adapter. It also keeps the integration auditable: every MCP call lands in your workspace audit log.
Do you train any AI models on my data?: No. WebmasterID is a privacy-first analytics platform. Your event data is yours; we never train shared models on it, we don't sell or share it with third parties, and we never send raw analytics to model providers. When you ask Claude to read your data through MCP, the request goes directly from your Claude client to our MCP endpoint — your operator session controls what is read.
How is this different from a generic analytics dashboard?: Generic dashboards are read-mostly UIs you scroll through. AI-assisted analytics is read-and-investigate: the operator asks Claude a question grounded in real data, Claude reads the relevant slice through MCP, and the conversation surfaces patterns you would have missed scrolling. WebmasterID still ships the dashboard UI — AI-assisted is an additional surface, not a replacement.
Which Claude clients are supported?: Any client that speaks MCP today. The /settings/mcp page in the dashboard ships ready-to-paste configuration snippets for Claude Desktop, Claude Code (CLI), and the VS Code Claude extension. The MCP endpoint is workspace-scoped and authenticated by a Bearer API key.

Where to go from here

Same site, deeper coverage of the parts AI-assisted analytics depends on.

Agent + MCP details

The technical shape of the MCP integration — endpoints, tools, security model.

MCP SEO workflows

Concrete example workflows that use the MCP surface for AI-search-era SEO.

Claude as an SEO operator

How Claude reads recommendations, prepares prompts, and proposes code changes.

AI visibility

The metrics layer Claude reads — AI crawlers + AI assistant referrals.

Attribution explainer

How directional attribution works and how Claude can summarise it.

Multi-site analytics

For agencies and teams running many sites or projects in one workspace fabric.

MCP analytics

The Model Context Protocol surface Claude reads through.

Claude analytics

Claude Desktop, Claude Code, and VS Code Claude — the operator's view.

AI visibility analytics

The AI crawler and AI-assistant signal Claude reads first.

Event Explorer

The investigation surface the MCP server mirrors for AI clients.

Website observability

The end-to-end observability fabric AI-assisted analytics sits on.