A Complete List of MCPs For Growth Marketers
Some say MCPs are overvalued. Others say they’re undervalued. There are too many opinions and too much noise. I’ll say this instead:
MCPs are under-used. There’s far more talk than actual usage.
Everyone should try them at least once.
What Is MCP?
Model Context Protocol (MCP). MCP is like a "USB-C port" for AI. It allows your AI assistant to plug directly into your software (GA4, Slack, etc.) to read data and perform tasks for you.
Tips
The GitHub documentations and links below can be intimidating in the beginning. I usually send the link to Claude Code (or another LLMs) and let it guide me step by step, and explain to me when is not clear.
Check your company’s AI data policies before connecting a database like BigQuery.
MCPs For Web Analytics
Google Analytics (Official) — Must Have
https://github.com/googleanalytics/google-analytics-mcp?tab=readme-ov-file
Sample prompts:
“What can the analytics mcp server do?”
“What are the most popular events in my Google Analytics property in the last 180 days?”
Google Tag Manager (Third Party) — Must Have
https://stape.io/solutions/mcp-server-for-gtm
“Set up GA4 e-commerce tracking for my store.”
“Create a variable that returns the local timestamp.”
“Make GA4 fire only when analytics consent is granted.”
Bonus:
Worth trying this GTM skill as well, easy to implement and very powerful:
https://github.com/aimonk2025/google-tag-manager-automation
Amplitude (Official) — Nice to Have
https://amplitude.com/docs/amplitude-ai/amplitude-mcp
“Show me the funnel conversion rate for users acquired via paid social, segmented by device, compared to organic users over the last 60 days.”
“Create an experiment to test the new checkout flow with a 50/50 split.”
Mixpanel (Official) — Nice to Have
https://docs.mixpanel.com/docs/features/mcp
“Create a cohort of users who did purchase in January and check their retention.”
“Show me event patterns for users on iOS vs Android.”
PostHog (Official) — Situational
https://posthog.com/docs/model-context-protocol
“Show me the drop-off in my onboarding funnel for the last 30 days.”
“Which feature flags are currently active and what percentage of users do they affect?”
“Run a SQL query to find users who completed signup but never activated.”
Hotjar (Third Party) — Nice to Have
No official Hotjar MCP exists yet. Best current option is connecting through Zapier’s MCP layer.
“Which pages have the most rage clicks this week?”
“Show me scroll depth data for my pricing page.”
“Summarize heatmap patterns across my top 5 landing pages.”
If You Are Curious About: GA4 vs Amplitude vs Mixpanel vs PostHog
GA4
Google Ads attribution
SEO and organic traffic analysis
A free tool with good-enough reporting
Integration with the Google ecosystem
Amplitude
Deep retention and cohort analysis
A strong experimentation platform
Compass-style behavioral intelligence
Enterprise-grade governance
Mixpanel
Best-in-class funnel analysis
Strong conversion optimization focus
Flexible event-based querying
Growth / PLG team workflows
PostHog
Full control and data ownership
Session replay + heatmaps
Feature flags and A/B tests in one tool
Engineering-driven analytics
Self-hosted for compliance/privacy
MCPs For SEO
Google Search Console (Third Party) — Must Have
https://github.com/ahonn/mcp-server-gsc
“Show me total clicks, impressions, CTR, and average position for the last 28 days.”
“What are my top 20 queries by clicks this month?”
“Show me long-tail queries (4+ words) with decent impressions.”
“Which pages have clickable elements too close together?”
Ahrefs (Official) — Nice to Have
https://docs.ahrefs.com/docs/mcp/reference/introduction
SEMrush — Nice to Have
https://developer.semrush.com/api/basics/semrush-mcp/
Sample prompts for Ahrefs and SEMrush:
“Plot a timeline of monthly organic traffic to ahrefs.com over the past 2 years. Show the major traffic metrics along with the chart as a dashboard.”
“Make a keywords analysis for semrush.com.”
Firecrawl (Official) — Situational
https://docs.firecrawl.dev/mcp-server
“Scrape [competitor page] and extract their meta title, description, and H1s.”
“Search the web for X and return the content of the top results.”
MCPs For Paid Ads
Google Ads (Official) — Must Have
https://github.com/googleads/google-ads-mcp
“Show me campaign performance for the last 30 days — clicks, impressions, CPC, and conversions.”
“Which keywords have the highest CPA in my search campaigns?”
“Which ad groups have a CTR below 1% this month?”
Meta Ads (Third Party) — Must Have
https://github.com/pipeboard-co/meta-ads-mcp
No official Meta MCP exists. Pipeboard is the most widely used third-party implementation.
“Show me ad set performance for the last 7 days segmented by creative type.”
“Which campaigns have the highest cost per lead this month?”
“Compare ROAS across retargeting vs prospecting campaigns.”
MCPs For Data Warehouses & Databases
BigQuery (Official) — Nice to Have
https://docs.cloud.google.com/bigquery/docs/pre-built-tools-with-mcp-toolbox
Snowflake (Official) — Situational
https://github.com/Snowflake-Labs/mcp
PostgreSQL — Situational
https://github.com/crystaldba/postgres-mcp
Sample prompts for all above:
“Show me user retention cohorts for users who signed up in January.”
“Show me users who added to cart but never checked out.”
“What’s the most common user journey before churn?”
MCPs For Visualization
Looker Studio (Official) — Nice to Have
https://docs.cloud.google.com/looker/docs/connect-ide-to-looker-using-mcp-toolbox
“Find the sessions explore, check what acquisition dimensions exist, then query 30-day conversion rate by channel.”
Tableau (Official) — Situational
https://tableau.github.io/tableau-mcp/docs/intro
“Show me what fields are available in the Web Behavior data source.”
“Which data sources haven’t been refreshed in 30+ days?”
MCPs For CRM & Lifecycle
HubSpot (Official) — Nice to Have
https://developers.hubspot.com/mcp
“Show me all deals that have been in ‘Proposal Sent’ for more than 14 days.”
“Which leads from the last 30 days came from paid social channels?”
“Summarize contact activity for my top 10 open opportunities.”
MCPs For Efficiency & Automation
Google Workspace (Third Party) — Must Have
https://github.com/taylorwilsdon/google_workspace_mcp
The most powerful feature is chaining. LLMs can run end-to-end workflows across Google tools in one go.
Examples:
“Pull data from BigQuery, write it to a Google Sheet, then email the sheet link to my team.”
“Check my Gmail for any data requests, create Tasks for each one, and reply confirming receipt.”
“Create a monthly report Doc from this data, save to Drive, and share with the marketing team.”
Gmail
“Draft an email summarizing this week’s GA4 report and send to my team.”
“Label all emails from our data engineering team as ‘Data Team’.”
Google Drive
“Search for any dashboards or reports in the Marketing folder.”
“Show me files I own that haven’t been opened in 6 months.”
Google Sheets
“Update the ‘Weekly KPIs’ sheet with this week’s numbers.”
“Read the data in Sheet 1 and summarize the trends.”
“Format the header row and add filters to this sheet.”
Google Docs
“Create a weekly analytics report doc from this data.”
“Add a comment to this section flagging a data discrepancy.”
“Draft a data request template doc for stakeholders.”
Google Slides
“Add a new slide with this funnel chart data.”
“Build a QBR deck outline from these performance highlights.”
Google Calendar
“Schedule a weekly analytics review meeting every Monday at 10am.”
“Block 2 hours Friday for dashboard work.”
Google Chat
“Send the weekly metrics summary to the #analytics channel.”
“Post an alert to #data-team that the GA4 export is complete.”
Google Forms
“Show me responses from this week’s user research form.”
“Build a form for teams to submit KPI tracking requests.”
Google Tasks
“Create a task to review the Amplitude funnel analysis by Friday.”
“Add subtasks to my ‘Q1 Analytics Audit’ task.”
Google Contacts
“Find contact info for everyone on the data team.”
“Create a contact group for all analytics stakeholders.”
Slack (Official) — Nice to Have
https://docs.slack.dev/ai/slack-mcp-server/
“Find all threads I haven’t replied to this week.”
“Alert #data-team when GSC rankings drop significantly. (with google search console mcp)”
Notion (Official) — Nice to Have
https://github.com/makenotion/notion-mcp-server
“Create a new page summarizing this week’s analytics highlights.”
“Find all pages tagged ‘Q1 Analytics’ in my workspace.”
Linear (Official) — Situational
“Find all bugs related to the analytics dashboard.”
“Create an issue: ‘Fix GA4 funnel tracking broken on mobile’ — assign to me, priority high.”
“Which issues have been in ‘In Progress’ for more than 2 weeks?”
Zapier (Official) — Nice to Have
Zapier helps to automate your workflows across different tools.
Use Cases
Automated Alerts & Notifications
Connect GA4/Amplitude anomalies → Slack alert to your team
Send email when a KPI drops below threshold
Notify Slack when a new Looker Studio report is shared
Push BigQuery query results to a channel on a schedule
Reporting Automation
Schedule weekly GA4 summary → auto-send to email list
Export Amplitude cohort data → append to Google Sheets automatically
Trigger a Looker report → save PDF to Google Drive every Monday
Pull data from BigQuery → populate a Notion dashboard
Data Pipeline Glue
Form submission (Typeform/Webflow) → send event to Amplitude
New CRM deal (HubSpot/Salesforce) → log conversion event in GA4
App event triggers → create row in BigQuery or Google Sheets
Sync user properties between your CRM and Amplitude
Cross-Tool Workflows
New Jira ticket for data request → notify analyst on Slack
Google Sheets data update → trigger BigQuery load job
Amplitude flag for churned user → create task in Asana/Notion
GA4 goal completion → add lead to HubSpot CRM
Enjoy exploring the MCPs! If any of them are difficult to configure, leave a comment and I’ll add more detailed instructions.

