Your analytics expert, inside Claude.
167 audit rules from years of GA4 and GTM client work, and a marketing analyst that answers why numbers moved and which channel pays. Connect your platforms, then ask in plain English. Get the fix, the answer, or a branded spreadsheet, document or deck.
50 free credits. No card. Get started free in your AI app · Try MarStack Intelligence · See how it works
Why did revenue drop in GA4 last month?
MarStack · 62/100 · Needs work
- Critical214 purchase events sent value 0 in the last 30 days.
- HighGoogle Ads is linked but no key event is imported.
Map value to the ecommerce.value variable on the GA4 purchase tag. Want me to prove it on the thank-you page?
- expert rules: 89 GA4, 20 GTM, 34 Meta, 24 Google Ads
- 167
- expert rules: 89 GA4, 20 GTM, 34 Meta, 24 Google Ads
- tools your AI can call
- 45
- tools your AI can call
- Google sign-in for GA4, Tag Manager and Search Console
- 1
- Google sign-in for GA4, Tag Manager and Search Console
- free credits to start, no card
- 50
- free credits to start, no card
Connects to
- GA4Live
- Tag ManagerLive
- Search ConsoleLive
- Meta AdsLive
- Google AdsLive
- BigQueryLive
- LinkedIn AdsComing
- TikTok AdsComing
Try it now
Run a real check on your own site
This is the same check your AI assistant runs through MarStack, without the sign-up. Connect GA4 and GTM after signing up for the full 109-rule audits.
What tracking fires on this page?
Pick a check, enter a page and run it. The real MarStack answer appears here.
The MarStack Agent
Checks the data, explains it, waits for you, then delivers
- 1
Audits and live tests
Check the data first
Before explaining a number, MarStack checks the tracking behind it: scored audits for GA4, GTM, Meta and Google Ads, and a live conversion test that submits your form in a real browser.
ExampleGA4 property audit · 89 checks
64/100
- CriticalConsent Mode default is missing
- HighLead form event is not a key event
- MediumUnassigned traffic is 14%
- 2
Intelligence
Explain it in plain words
What happened, why, and what to do, with the tables behind it. Answers cite past audits and changes, and say when a drop is a tracking problem rather than a business one.
Why did leads drop last week?
ExampleLeads fell 38%, and it is tracking, not demand. Sessions held steady, but generate_lead stopped reaching GA4 the day GTM version 12 was published.
Week Sessions Leads Before 4,820 96 Last week 4,710 60 Next: restore the GA4 tag (MarStack can draft it for approval).
- 3
Agent
Wait for your approval
The agent can only propose changes: negatives, pauses, budgets within 20%, GTM tags for a new page. Each waits in MarStack or Slack with its before and after, and keeps an undo.
ExampleGoogle Ads negative keywords · 123-456-7890
Add 3 negative keywords to Search - Generic
These search terms spent $184 in 30 days with no leads.
What Before After Negatives none free course, jobs, salary Google accepted the change in a dry run.
Waiting for you · expires in 2 days - 4
Delivery
Deliver where you work
Weekly briefs by email and Slack, result tables in Google Sheets and the monthly client deck in Google Slides, on a schedule you write in plain words.
Every Monday 9:00, delivered
Example- Slack #marketingWeekly brief posted, 1 change to approve
- EmailSent to you and the client lead
- Google SheetsSearch terms table, 214 rows
- Google SlidesNovember client deck, 9 slides
A real test
We tested our own lead form. It sent no conversions.
On 5 November 2026 MarStack submitted the contact form on getinlytics.com in a real browser. The form worked and pushed generate_lead to the dataLayer, but no GTM tag forwarded it, so GA4 counted zero leads from that form. A configuration check would have said everything looked fine.
MarStack drafts the missing GA4 tag in a GTM workspace, tests it in preview and publishes only after the account owner approves.
Live conversion test · /contact
1 test case failed
- TC3Nothing tracks before a choice
- TC4Accept grants consent
- TC7The form submits
- TC8generate_lead reaches the dataLayer
- TC9GA4 receives generate_lead
- TC13No personal data in tags
The form pushes generate_lead, but no GTM tag sends it to GA4.
How it works
From sign-up to your first audit in five minutes
- 1
Connect your platforms
One Google sign-in covers GA4, Tag Manager and Search Console. Access is read-only, and website checks need no connection at all.
- 2
Create an MCP token
Name it after your AI client. MarStack writes the config with the token already in it, ready to paste.
- 3
Ask in plain English
"Audit my GA4." Answers come ranked by impact, with proof and the exact fix, plus a branded report when you want one.
Why it is different
Config checks say it looks fine. MarStack loads the page and proves it.
- GA4 purchase tag exists
- Trigger: thank-you page
- Measurement ID matches
All a config-only tool can tell you.
- purchase request fired
- value parameter empty
- transaction_id T-88213
Captured from a real page load, with a screenshot. Example result.
More than audits
A marketing analyst on your own data
Audits tell you what is broken. The MarStack Analyst tells you what is happening: it reads GA4, Meta Ads, Search Console, BigQuery and the CRM data you upload, checks the data before it concludes, and ends every answer with what to do next.
Business questions, answered
Why a number moved, which channel or campaign brings deals, whether ad spend pays off, and weekly or monthly reviews, across every platform for the same dates.
- Why did leads drop last week?
- Which channel brings won deals for our spend?
- Why do Meta and GA4 disagree?
Files your team can use
Ask for the analysis as an Excel workbook, a Word or PDF document, or a PowerPoint deck. Numbers are copied from MarStack, never retyped, and agencies can add their own brand.
- Put this in a branded spreadsheet
- Make a monthly review deck
- Send the audit as a PDF
In Claude, or in the portal
Use MarStack in Claude, ChatGPT, Gemini, Cursor or VS Code (paid plans add MarStack Analyst in every one of them), or, on paid plans, ask in MarStack Intelligence in the portal, no AI subscription needed. Quick answers from 2 credits.
- Claude Desktop and Claude Code
- Cursor, VS Code and other MCP clients
- MarStack Intelligence chat
What you can ask
Plain questions, expert answers
“Does the purchase event fire on my thank-you page?”
Loads a page in a real browser and records every tag, pixel, hit and cookie that fires. Hits are blocked, so your analytics stay clean.
inspect_page_tracking5 credits
“What tags does example.com run in GTM?”
Audits the published GTM container behind any website, no access needed: tags, consent settings, measurement IDs and risky Custom HTML.
audit_gtm_public_container10 credits
“Is my cookie banner GDPR compliant?”
Clicks Reject and Accept on the cookie banner in two fresh visits and compares what fires, with a compliance verdict and screenshots.
check_consent_compliance10 credits
“Audit my GA4 property 504011059”
89 expert checks on your GA4 property: data quality, events, ecommerce, links, audiences, PII and consent, scored out of 100 and ranked by impact.
audit_ga4_property10 credits
“Which channel brought the most leads this month?”
Sessions, users, engagement, key events and revenue by channel, campaign, landing page, device or country, compared with the previous period, with data-quality warnings.
get_ga4_performance1 credit
“Show form_start events by landing page for October”
Any GA4 report: your choice of dimensions, metrics, dates and filters, checked against your property so a wrong name gets close matches instead of an error.
run_ga4_report1 credit
“Audit my GTM container GTM-ABC1234”
20 expert checks on the live version of your GTM container, with real tag, trigger and variable names and the fix for each problem.
audit_gtm_container10 credits
“Which queries lost clicks this month?”
Queries, pages, clicks, impressions, CTR and position from Search Console, with the change against the previous period.
get_search_console_data1 credit
“Which pages do buyers visit before their first purchase?”
Answers what standard reports cannot (user paths, funnels, cohorts) with one read-only query on the BigQuery datasets you allow. A free dry run first, a size cap you set, and results with personal data removed.
run_bigquery_sql1 credit
“Where do people drop out between view_item and purchase?”
Funnels, landing page paths, returning-user cohorts and events that stopped firing, from your GA4 export in BigQuery. Prebuilt, reviewed queries with a free dry run and a size cap you set.
query_ga4_events1 credit
“Which channels get credit for purchases under each attribution model?”
Reviewed BigQuery SQL for the questions that matter most: first touch, last touch, linear and position-based attribution on your GA4 export, Meta results vs GA4 vs CRM by campaign, leads to won deals by source, and cost per won deal and ROAS on CRM revenue.
run_attribution_query1 credit
“Why does Meta report more leads than GA4?”
Meta says 40 sales, GA4 says 25: finds out why. Reads both platforms for the same dates, the latest audits and, if you like, your thank-you page in a real browser, then ranks the causes by evidence with the fix for each.
diagnose_platform_mismatch2 credits
“Which source brought the most won deals this quarter?”
Leads, deals, win rate and won value from your CRM by source, campaign, stage or month, compared with the period before. Connect HubSpot, Salesforce or Zoho CRM (read-only, refreshed daily) or upload a CSV from any CRM. Names, emails and phones are never stored.
get_crm_data1 credit
“Which channel brings the most won deals for our ad spend?”
GA4, Meta Ads, Search Console and your CRM for the same dates in one table per channel or campaign: sessions, key events, ad spend, search clicks, deals and won value, with cost per deal and return on ad spend. Each platform's own conversions side by side, never added together.
get_marketing_overview1 credit
“Audit my Meta ad account 1234567890”
34 expert checks on your Meta ad account: pixel and Conversions API, match quality, optimisation events, Learning Limited, rejected ads, frequency, UTMs, plus a live check that the pixel waits for consent.
audit_meta_ads10 credits
“Which Meta campaigns had the best ROAS last month?”
Spend, clicks, CTR, CPC, purchases or leads, cost per result and ROAS from Meta Ads, by campaign, ad set or ad.
get_meta_performance1 credit
“Set up tracking for my new landing page”
Scans a new landing page and your live GTM container, then drafts the tags and triggers it needs. Tested in GTM preview first; nothing is published until you approve.
propose_gtm_setup10 credits
“Does my contact form conversion really reach GA4?”
Tests your conversion tracking end to end: opens the page in a real browser, answers the cookie banner, fills and submits the form, and checks that GA4, Google Ads and Meta receive the conversion.
run_live_test8 credits
“Which search terms spent money with no leads last month?”
Cost, clicks, conversions, value and cost per conversion from Google Ads by campaign, ad group, keyword or search term.
get_google_ads_performance1 credit
“Audit my Google Ads account”
24 expert checks on your Google Ads account over the last 30 days: conversion tracking, wasted search terms, budgets, bidding and ad strength, ranked by impact.
audit_google_ads10 credits
Fewer tokens, more answers
Up to 6x fewer AI tokens than raw Google data
A raw GA4 or Search Console connector hands your AI long blocks of API data and leaves it to do the maths. MarStack hands over a finished table with the changes already worked out, so each answer uses less of your AI plan's usage limit.
- More questions before you hit your usage limit
- Room in the conversation for follow-ups
- Your data already connected: no exports, no copy and paste
Traffic by channel, last 28 days vs the 28 before5.0x fewer
527 tokens2,635 tokensDaily traffic trend, 28 days3.6x fewer
1,050 tokens3,813 tokensTop 25 landing pages vs the period before5.9x fewer
1,531 tokens8,960 tokensTop 100 Google search queries vs the period before2.9x fewer
5,107 tokens14,973 tokens
Read-only unless you approve
MarStack reads settings and reports. A change happens only after a person approves it, with an undo.
No analytics pollution
Browser checks block every tracking hit, so tests never add visits.
Encrypted access
Google tokens are encrypted at rest and refreshed automatically.
Your tokens, your control
One token per AI client. Revoke any of them in a click.
Ask your first question in five minutes
50 free credits cover about five audits in your AI client. No card needed.