Looker Studio · GA4 · Ads · BigQuery

Looker Studio for clear marketing, e-commerce and management reporting.

I design reports that lead the reader from a result to its cause and the next decision. I combine GA4, Google Ads, Search Console, BigQuery and prepared business data into a coherent structure of KPIs, trends and detailed analysis.

Looker Studio is the presentation layer. Report credibility depends on metric definitions, source quality, the data model, access and the way the product is maintained after launch.

Describe the readers and reports See reporting examples
Questions and KPIs before chartsevery screen has a reader, purpose and decision
A source matched to complexitydirect connector, BigQuery table or prepared model
QA, access and ownershipnumbers, filters, credentials and changes have acceptance criteria

The usual starting point

The dashboard exists, but it does not shorten the path to a decision.

The problem is rarely a missing chart. More often it is weak information hierarchy, conflicting definitions, an unstable source or no report owner.

The first screen shows everything

Dozens of KPIs compete for attention, and the reader cannot see which result needs action.

Numbers differ across reports

GA4, Ads, CRM and spreadsheets use different definitions, filters, currencies or time ranges.

The report is too slow

Too many charts, blends and raw BigQuery queries make daily use difficult.

Access depends on one person

Sources use accidental credentials, and the report stops refreshing when the owner changes.

Important analysis remains manual

The team exports data because margin, funnel, segment or the required detail is missing.

Nobody maintains the definitions

Calculated fields, filters and connections change without documentation or impact testing.

Reporting product

The report leads from signal to explanation.

The structure follows the readers, but the logic remains stable: result and deviation first, drivers second, operational detail last.

01 · Signal

Executive overview

Core KPI, trend, target, deviation and concise context.

02 · Cause

Result drivers

Channel, campaign, product, segment, region, funnel stage or cohort.

03 · Action

Operational view

Detail required to optimise, control the issue and assign an owner.

Executive viewresult, target, trend and risk
Marketing and acquisitioncost, leads, sales and channel effectiveness
E-commerce or productfunnel, revenue, margin, retention and segments
Data qualityfreshness, completeness and source discrepancies

Source selection

Not every question needs BigQuery. Not every question fits the GA4 connector.

The source is selected using metric logic, system count, performance, cost, permissions and data reuse.

Simpler reporting

Direct connector

GA4, Google Ads, Search Console or Sheets when one source contains the required metrics and its logic is acceptable.

Good for: a fast start and one team
Shared definitions and scale

BigQuery or a prepared database

Reporting tables for multiple sources, margin, LTV, SaaS, CRM data or better report performance.

Good for: e-commerce, SaaS and multi-source reporting
Process supplement

Spreadsheet or manual input

Targets, budgets, comments, plans and reference data with an explicit owner and update control.

Good for: targets and small supporting tables

A Looker Studio blend can serve a simple case. Critical, reusable or difficult-to-test logic belongs in a prepared data model.

Process

From business questions to a report ready for daily work.

  1. 01

    Readers and decision rhythm

    Management, marketing, sales or e-commerce, usage frequency, decisions, targets and current problems.

    Outcome: reporting brief
  2. 02

    KPIs and data sources

    Definitions, scope, filters, reference system, credentials, freshness and source readiness.

    Outcome: KPI and source map
  3. 03

    Architecture and wireframe

    Page hierarchy, first screen, analysis paths, global filters, reader roles and wireframe.

    Outcome: approved prototype
  4. 04

    Build and optimisation

    Reusable sources, fields, parameters, visualisations, performance and presentation variants.

    Outcome: test report
  5. 05

    QA and business acceptance

    Number reconciliation, filters, ranges, access roles, edge cases and usage scenarios.

    Outcome: acceptance report
  6. 06

    Publication and handover

    Metric documentation, owners, user instructions, training and planned-change backlog.

    Outcome: maintainable report

Use cases

One standard, different questions and levels of detail.

Executive and owner

Management reporting

Target delivery, sales, margin, acquisition cost, trend, risks and decisions required.

Marketing

Performance and acquisition

Cost, campaigns, leads, conversions, ROAS, CAC and the quality of acquired traffic or customers.

Commerce

E-commerce

Funnel, orders, revenue, discounts, returns, products, segments and profitability.

Product and SaaS

Activation and retention

Trial, activation, feature use, cohorts, MRR, churn and the path to paid subscription.

Content and SEO

Visibility and content

Search Console, landing pages, engagement, conversion and content contribution.

Analytics

Quality control

Source freshness, volume, missing data, deviations and integration status.

Tangible outcome

The first screen should show whether the result needs action.

This is a sample executive-page specification. Final scope depends on the business model and agreed sources.

Report pageExecutive overviewmonth / week
ResultKPI, target and deviation

Current value compared with plan and previous period.

TrendChange over time

Seasonality, turning point and the moment the deviation began.

DriversWhat influenced the result

Channel, product, segment, region or funnel stage.

NavigationMove to detail

A link to the operational view that preserves context and filters.

Acceptance criteria

The report must be trustworthy, responsive and safe to share.

Exact tolerances, roles and expected freshness depend on the sources and are agreed before the build.

KPIs match their definitionsvalues have a reference source, tolerance and documented differences
Filters preserve contextperiod, segment, channel and controls affect the correct charts
The report loads in acceptable timecharts, blends, sources and queries are optimised
Credentials match the audienceowner, viewer or service-account access fits the sharing model
Freshness is understoodthe reader knows when each source updates and what is not yet included
The report has an ownerdocumentation covers metrics, sources, permissions and change procedure

Honest limitations

What Looker Studio does not solve by itself.

  • It is not a data warehouse.Complex joins, history and multi-source logic should be built before the reporting layer.
  • It does not create shared KPI definitions.A calculated field does not replace a decision about metric meaning and source of truth.
  • A blend is not always a stable model.Complex joins, different grains and repeated use are better handled in the database.
  • BigQuery creates query costs.Report refreshes may run queries, so smaller reporting tables and cost control are needed.
  • Credentials require a deliberate choice.Owner credentials simplify sharing but extend data access through the report.
  • Every connector has limits.Freshness, quotas, field coverage and support depend on the source and provider.
Krzysztof Surowiecki

Analytics and data communication

I connect the data model, information design and reader needs.

I lead the project from questions and metric definitions through build, QA and training. The report becomes the final part of the analytics process—not a separate decorative layer.

20+years of experience
150+analytics and data projects
50+clients

Figures cover my broader analytics and data practice, not only Looker Studio implementations.

Before we talk

Frequently asked questions.

Can a report use GA4 directly?

Yes, when one source and connector metrics cover the scope. Event-level data, custom sessions, CRM, margin or multiple systems often require a prepared BigQuery table.

When do I need BigQuery?

When the report joins sources, requires custom KPI logic, long history, e-commerce or SaaS data, or performs poorly on raw connectors and blends.

Can data be restricted by reader?

Yes. The method depends on the source and credentials. Options include viewer credentials or email-based filtering when the model supports it.

Are Looker Studio and Looker the same product?

No. Looker Studio is a reporting and visualisation tool based on data sources and connectors. Looker is a broader BI platform with a semantic modelling layer and a different implementation model.

Will the report work well on a phone?

I design for the important resolutions and usage pattern. Looker Studio is not a mobile app, so small-screen layouts need fewer charts and stronger prioritisation.

What happens after publication?

I provide documentation and training. We can also arrange source monitoring, periodic QA, report development or ongoing analytics support.

First step

Describe the readers, sources and three questions the report should answer.

I will ask about KPI definitions, decision frequency, sharing model and the readiness of the reporting data.

Discuss the report See reporting examples

Prepare the project

Check the data and possible BI options.

GA4 configuration auditorassess the source before building the reportRun the auditor BigQuery calculatorestimate source costs for reportingEstimate cost BI tools comparisonLooker Studio, Power BI, Tableau and alternativesCompare tools