Guide · Agent skill
How to use the Analytics Tracking Setup agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The analytics skill turns an AI agent into an analytics-implementation helper. It is aimed at marketers setting up GA4 or Google Tag Manager for the first time, developers who need an event naming convention before instrumenting a site, and analysts trying to work out why events are missing. This guide explains what the skill does, how to install it, how to trigger it, and what it needs before you start.
Analytics Tracking Setupcoreyhaines31/marketingskills · MIT Download the skill Quick start
- 1Download and unzip the skill. Download the ZIP from the directory page and unzip it into ~/.claude/skills/ for personal use or the project's .claude/skills/ folder.
- 2Confirm your access. Make sure you have a Google Analytics 4 property and Google Tag Manager access if you plan to follow the GA4 or GTM steps, plus code access to install gtag.js or GTM snippets.
- 3Describe your decisions. Tell the agent what decisions the data will inform and which conversions matter most.
- 4Ask for a tracking plan. Request a plan with event names, properties, triggers and notes, and agree on naming before implementing.
- 5Implement GA4 or GTM. Ask for the GA4 or Google Tag Manager steps and snippets, then install them in your site or app code.
- 6Check that events fire. Ask the agent to help verify tracking. Use GA4's debugging tools described in the reference files to confirm events arrive.
What the analytics skill makes the agent do
Once installed, the agent works through analytics setup with you. It can design a tracking plan with consistent event names and properties, write GA4 and Google Tag Manager setup steps and code snippets, and help check whether tracking is working.
The skill also enforces careful measurement habits. It asks what decisions the data will inform, computes totals from row-level data rather than eyeballing them, and compares rates only over matching time windows. When a question belongs to another area, it points you to a separate skill: conversion-tracking for ad-platform pixels, conversion tags, CAPI and server-side setup; attribution for attribution models and reconciling conflicting numbers; and ab-testing for A/B test measurement.
- Designs a tracking plan with event names, categories, properties, triggers and notes
- Uses object-action event names such as signup_completed and form_submitted
- Writes GA4 and Google Tag Manager setup steps and data layer snippets
- Helps check whether events are firing and why tracking may be broken
- Routes ad-platform conversions, attribution and A/B tests to other skills
Install the analytics skill
Download the ZIP for the skill from the directory page, then unzip it. You can install it for yourself or for a single project. For a personal install, unzip it into ~/.claude/skills/. For a project install, unzip it into the .claude/skills/ folder inside your project.
If you use the Claude apps instead, upload the skill under Settings → Capabilities → Skills. The skill folder includes SKILL.md, which holds the main instructions, and three reference files: references/event-library.md, references/ga4-implementation.md and references/gtm-implementation.md. It also includes evals/evals.json.
- Personal install: unzip into ~/.claude/skills/
- Project install: unzip into the .claude/skills/ folder in your project
- Claude apps: Settings → Capabilities → Skills, then upload the skill
Trigger the skill with natural requests
You do not need a special command. Describe what you want in plain language and the agent should load the skill when your request matches its description. The skill's description lists trigger phrases such as "set up tracking," "GA4," "Google Analytics," "event tracking," "UTM parameters," "tag manager," "GTM," "tracking plan," "Mixpanel," "Segment," "are my events firing" and "analytics isn't working."
If your question is really about ads, attribution or experiments, ask for that directly. The skill is designed to hand those topics to conversion-tracking, attribution or ab-testing.
- "Set up GA4 for my marketing site. We need to know which demo requests come from paid search."
- "Write a tracking plan for our signup and onboarding flow."
- "Our GTM form tag isn't firing on the contact page. Help me debug it."
- "Give me a UTM naming scheme for our email and social campaigns."
- "How do I measure whether the new pricing page is working?"
Work through a typical setup
Start by telling the agent what decisions the data should support and what your key conversions are. If you have a product marketing context file at .agents/product-marketing.md, .claude/product-marketing.md or the older product-marketing-context.md, the agent reads it first and asks only for what is missing. Then describe your current tools and any privacy or compliance requirements.
Next, ask for a tracking plan. The skill suggests a table with event name, category, properties, trigger and notes. After you agree on the plan, ask for the GA4 or Google Tag Manager steps. Finally, use the agent to check whether events arrive as expected.
- Name the decisions and key conversions first
- Build the tracking plan before writing any code
- Map internal business names to provider event names explicitly
- Use GA4's recommended event names where they exist, such as sign_up
gtag('event', 'sign_up', {
'method': 'email',
'plan': 'free'
});Follow naming and UTM conventions
Consistent names matter more than clever ones. The skill recommends lowercase names with underscores, specific names such as cta_hero_clicked instead of button_clicked, and context placed in properties rather than in the event name. Keep provider-defined names for automatic or recommended events, because naming a custom event does not make the provider collect it automatically.
For UTM parameters, the skill recommends lowercase values, hyphens between words, one spelling per source, and no personal data. Keep allowed values in a shared sheet and build links from it.
- Lowercase everything: GA4 treats LinkedIn and linkedin as two sources
- Use hyphens between words and one separator for the whole company
- One spelling per source, such as linkedin rather than li
- Never put emails, names or phone numbers in parameters
dataLayer.push({
'event': 'form_submitted',
'form_name': 'contact',
'form_location': 'footer'
});Use careful measurement habits when checking results
When the agent reports numbers, it should derive totals, rates and deltas from row-level data with a script or spreadsheet. It should cross-check headline figures a second way and report segment numbers for the segment in question rather than burying them in a group total.
Rate metrics must divide numerator and denominator over the same closed window. If you divide a full month of spend by a partial month of conversions, the result is wrong. Cohorts still inside their observation window should be labelled as pending, not counted as failures.
- Compute totals from row-level data, not by scanning rows
- Cross-check headline figures a second, independent way
- Restate rates over matching closed windows and say which weeks they cover
- Label cohorts still inside their window as pending
What people use it for
- Setting up GA4 and Google Tag Manager on a marketing site for the first time
- Writing a tracking plan with consistent event names before developers add tags
- Debugging why a form submission or click event is missing in reports
- Creating a UTM naming scheme shared across email, social and paid campaigns
- Auditing a product onboarding funnel to see which steps are tracked
Frequently asked questions
Is the analytics skill free?
The skill is published under the MIT licence, which is stated on the directory page. Check the licence file in the download for the full terms. GA4 and Google Tag Manager are separate Google products with their own terms.
What do I need before using it?
For GA4 or GTM steps you need a Google Analytics 4 property and Google Tag Manager access, plus website or app code access to install gtag.js or GTM snippets. You also need an agent that supports skills, such as Claude.
Does the skill run scripts or use the network?
According to the directory listing, the skill does not run scripts and does not use the network. It consists of instruction files and reference documents that the agent reads.
How do I install the analytics skill?
Download the ZIP, then unzip it into ~/.claude/skills/ for personal use or into a project's .claude/skills/ folder. In the Claude apps, you can instead upload it under Settings → Capabilities → Skills.
Can it set up Google Ads conversions or Meta pixels?
Not as its main job. Ad-platform pixels, conversion tags, CAPI and server-side setup belong to the separate conversion-tracking skill, which the analytics skill points you to.
Will it compare attribution models or fix mismatched numbers between tools?
No. Attribution models, multi-touch, MMM, incrementality and reconciling numbers across tools are routed to the attribution skill. The analytics skill alone will not settle those disputes.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you are setting up GA4 or Google Tag Manager for the first time, need a naming convention and tracking plan before developers instrument events, or need help finding why events are missing. Pick something else for ad-platform conversion setup, attribution analysis or A/B test measurement, and get legal review before collecting regulated or privacy-sensitive data, since the skill does not check consent or compliance.
Download the skill