A laptop in an office showing AnalyticsLiv Intelligence

AI analytics for GA4 and BigQuery

AnalyticsLiv Intelligence is AI analytics software for GA4 and BigQuery. It reads your data all day and tells you what changed and why, in plain language. Two more modules measure what your media spend causes (MMM) and group your customers by how they behave (AudienceIQ).

Dashboards show numbers and wait for you to ask. Intelligence tells you what changed and why.

How does AI analytics for GA4 and BigQuery work?

The Insight Engine reads your data, spots what changed and sends you the finding. You stop hunting through reports.

  1. Connect your data, read-only

    GA4 and BigQuery, with Google Ads and DV360 for context.

  2. The Insight Engine spots what changed

    It explains why it matters and ranks it by revenue impact.

  3. Findings come to you

    In Slack and email, or in the in-product feed.

  4. Ask in plain language

    Ask your own questions and get answers.

Anomaly detection with context

Not just “sessions dropped”, but where, for whom, since when, and where to look first.

ContinuousSegment-awareSlack & email

Example · Insight EngineLive
Checkout completion · returning mobile users
Tue

MobileReturningShipping step

Plain-language insights, ranked

Findings arrive as sentences a marketing head can act on, ranked by what acting is worth.

The so-what, spelled outRanked by impact

Example · Insight feedLive
#1 by impactOpportunity
Returning visitors convert at 2.4× the first-visit rate. The remarketing audience is underfunded.
#2Anomaly
Checkout completion fell 14% for returning mobile users since Tuesday.
#3Opportunity
Paid social CAC improved 9% week-on-week.

Ask your data anything

“Why did conversions dip last Tuesday?” Ask in plain words and see the numbers and the reasoning.

Natural languageGA4 + BigQueryReasoning shown

Example · Ask your data
You asked: “Why did revenue dip last weekend?”
A feed disapproval paused 60 Shopping SKUs for 31 hours.

Example screens with sample data.

Read-only · live in a dayGA4 + BigQueryGoogle Ads & DV360 contextDelivered to Slack & email
A professional working on a laptop showing analytics charts and graphs
Reads like
an analyst.
See it on your own dataFeed live on your GA4 · one MMM run walked end to end
Book a demo

What does the MMM module measure?

What each channel really causes, not what last-click credits it with. It uses total spend and revenue over months, with no cookies or consent needed.

ChannelAttribution's storyModel estimate (example)
YouTubeNo last click, no credit12% ±2 underfunded, room to grow
TV & offlineinvisible to tracking9% ±4 measured from your CSV history
Awareness halovanishes between cookiesFolded in, carried inside each channel's response curve
Basecredited to media42%, what the brand earns on its own

Example figures. Budgets should follow the model’s column.

Two MMM reports: summary and optimisation

Google AdsDV360MetaGA4 · optional controlCSV · TV, print & offline
Channel levelCampaign level
Example · summary_report.pdf

The Summary Report

What your history shows, with the evidence.

  • How well the model fits your history (R², MAPE, wMAPE)
  • Contribution and ROI by channel, base included, and where returns diminish
Example · optimisation_report.pdf

The Optimisation Report

The same budget, split differently.

3.8 → 4.2+$80K · same budget
  • A new split of the same budget, within limits you set
  • What the shift is worth. The new plan never has a lower ROI than today’s

Revenue split by what drove it (example)

Base: what the brand earns without media. Example figures. Yours is modelled from your own history.

Configure

Add spend, your KPI and the ROI you expect. Defaults are there if you need them.

Model runs

The Bayesian model runs in about ten minutes.

Reports arrive

Both reports land in your inbox, configuration attached.

Bayesian · every estimate shows its confidenceNo cookies, no user data~10-minute runsPriors you control

AudienceIQ: customer segmentation inside Intelligence

AudienceIQ groups your customers by how they behave and explains each group in a plain rule. Then it gets the groups ready for the ad platforms you already use.

Common questions about Intelligence

Ask us your own question

AI analytics software for GA4 and BigQuery. It reads your data all day, tells you what changed and why, and answers your questions in plain language. It also includes MMM and AudienceIQ customer segmentation.

Yes, through the AudienceIQ module. It groups customers by behaviour, explains each group in a plain rule, and prepares lists for GA4, Google Ads, DV360 or Meta. Input is GA4 data via the BigQuery export, or a CSV or Excel file.

It runs as a module inside Intelligence, on the same workspace and data, so you can demo and run them together. It is also sold on its own: see our marketing mix modeling page.

Read-only access to GA4, and optionally BigQuery, Google Ads and DV360. No new warehouse, live within a day, nothing written back. If GA4 needs fixing first, our GA4 and Tag Manager team can do it.

Dashboards wait for you to ask. Intelligence finds what changed and why, and sends it as a plain sentence ranked by revenue impact, in Slack, email or the in-product feed.

Spend and revenue history from Google Ads, DV360 and Meta, with GA4 as an optional control and CSV for offline media. Around eighteen months is enough to start.

It only uses totals of spend and revenue: no cookies, no user data, no consent needed. So it also measures channels tracking can’t see.

Insight Engine connections are live within a day. A full MMM run takes about ten minutes once your history is connected.

See Intelligence on your own GA4 data

Book a 30-minute demo on your own data. The insight feed reads your GA4 live, and we walk through one MMM run end to end.

A specialist replies by email to set up the demo on your own GA4 data.

Prefer to talk? Call +91 83205 76622, message us on WhatsApp or email [email protected].

Please enter your name.

Please enter a valid email address.

Please add a line about what you need.

Guides on AI and GA4 data from our team

All articles