Tableau · dashboards · analysis · self-service BI

Tableau as a shared analytics and reporting environment for the business.

I design data sources, dashboards and publishing in Tableau Cloud or Tableau Server. Management, analysts and operational teams get a solution for regular reporting and independent exploration of result drivers.

Implementation does not end with a workbook. It covers performance, refresh, projects, groups, permissions, row-level security, documentation and content-development rules.

Describe the data and audience View the Tableau portfolio
Trusted reusable sourcesshared fields, calculations, refresh and owners
Question-led analysisdashboard, drill-down and exploration without overload
Publishing, access and adoptionprojects, groups, RLS, training and development rules

Common starting point

Tableau runs, but the environment does not scale with the team.

When every workbook has its own source, calculations and permissions, each new dashboard increases maintenance cost instead of extending a shared analytics system.

Workbooks duplicate the same logic

KPI definitions, calculated fields and filters differ between teams and reports.

Dashboards are too slow

Too many sheets, queries, rows and complex calculations block analysis.

Extracts do not refresh reliably

Schedule, credentials and failed-refresh ownership are unclear.

Permissions grow without a model

Access is granted to individuals while roles, groups, projects and RLS remain disconnected.

There is no dev, test and production path

Changes go directly into business reports and are difficult to reproduce.

Self-service means proliferation

Users copy data and workbooks without knowing which assets are trusted.

Analytics product

From a trusted source to secure analysis and action.

I separate the data layer, dashboard design and publishing model, allowing one source to serve multiple workbooks with access controlled at environment level.

01 · Data

Published data sources

Live or extract, shared fields, calculations, metadata and schedules.

02 · Analysis

Workbooks and dashboards

Management and operational views, guided analysis, filters, actions and drill-down.

03 · Publishing

Tableau Cloud or Server

Projects, groups, site roles, permissions, RLS, refresh and ownership.

04 · Use

Audiences and self-service

Viewer, Explorer, Creator, training, trusted assets and change process.

One source, many workbooksless duplicated connection and business logic
Group-managed accesspermissions do not depend on individual exceptions
Analysis has a pathoutcome, cause, detail and action

Data architecture

Live or extract depends on freshness, scale and ownership.

The choice affects performance, database cost, refresh, availability and security and should be made per source.

Freshness and central control

Live connection

Queries reach the source during interaction. It suits current-data needs and a source prepared for the load.

  • performance depends on source and network,
  • permissions may be enforced in the database,
  • each interaction can execute a query.
Performance and controlled snapshot

Tableau extract

Data is materialised and refreshed on schedule, improving analysis speed and reducing source load.

  • refresh requires monitoring,
  • freshness follows the schedule,
  • data volume should match the need.
Consistency across reports

Published data source

A central source shares connection, fields, calculations and refresh instructions with many workbooks.

  • shared data definition,
  • fewer duplicated extracts,
  • an owner for change and quality.

Process

From analytics needs to an environment the team can develop.

  1. 01

    Audience, decisions and current environment

    Sources, Cloud or Server, licences, site roles, projects, reports, performance and collaboration.

    Outcome: scope and architecture
  2. 02

    Data model and sources

    Live versus extract, published sources, relationships, grain, calculations, refresh and owners.

    Outcome: trusted data layer
  3. 03

    Analysis design and prototype

    KPIs, information hierarchy, drill-down, filters, actions, devices and dashboard wireframe.

    Outcome: accepted prototype
  4. 04

    Build and optimisation

    Worksheets, dashboards, calculations, extracts, query reduction and Performance Recorder tests.

    Outcome: test workbook
  5. 05

    Publishing, permissions and RLS

    Projects, groups, site roles, capabilities, row-level security, credentials and schedules.

    Outcome: secure environment
  6. 06

    QA, training and adoption

    Reconciliation, permission tests, documentation, role-based training and change process.

    Outcome: maintainable solution

Use cases

Tableau for reporting, exploration and self-service BI.

Management

Executive cockpit

Outcome, target, trend, profitability, risk and key drivers.

Sales and e‑commerce

Sales performance

Revenue, margin, products, customers, regions, channels and transaction detail.

Operations

Operational analytics

Process, quality, inventory, lead time, SLA, anomalies and ownership.

Customers

Customer analytics

Segments, cohorts, retention, LTV, behaviour and cross-channel journeys.

Analysts

Guided exploration

Controlled exploration with parameters, actions and drill-down without source duplication.

Organisation

Managed self-service

Trusted sources, projects, roles, templates, publishing standards and creator training.

Tangible outcome

A sales dashboard connects the outcome with exploration.

Illustrative structure. The final workbook follows the data model, audience and decision rhythm.

WorkbookSales & Margin Cockpitmanagement + sales
OverviewRevenue, margin and target

Trend, variance, context and primary alerts.

DriversProduct, channel, customer and region

Contribution analysis and filtering to the source of change.

ExplorationActions and drill-down

From KPI to segment and transaction detail.

AccessRLS and audience groups

Salesperson sees own scope, manager the region, management the whole.

Acceptance criteria

The environment must be credible, fast and secure.

Performance, freshness and permission thresholds follow the data, platform and use.

KPIs reconcile with sourcesmetrics have definitions, systems of record and tolerances
Workbooks respond in acceptable timesheets, queries, calculations and rendering are optimised
Refresh is observableschedule, last success, errors and owner are known
Permissions pass scenariosViewer, Explorer, Creator and groups have correct capabilities
RLS does not expose out-of-scope dataeach role sees only the correct rows and aggregates
Content has owners and documentationsources, workbooks, projects and changes have accountability

Honest limitations

What Tableau does not solve automatically.

  • It does not replace a warehouse and data model.Complex integration, history and definitions should precede the workbook layer.
  • A published source needs governance.Consistency requires an owner and a controlled change process.
  • Live does not automatically mean fast.Performance depends on source, network, queries, interactions and workbook design.
  • An extract needs a refresh process.Schedule, credentials, errors and dependencies require monitoring.
  • RLS needs security testing.User filters, data policies and database security have different risks and costs.
  • Self-service without standards creates chaos.Licences and web authoring do not replace trusted sources, templates and training.
Krzysztof Surowiecki

Visualisation and business analytics

I connect data design, visual analysis and the team's way of working.

I lead work from question and source definitions through workbooks, publishing and training, combining technical quality with clear narrative and adoption.

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

Figures refer to the broader analytics and data practice, not only Tableau implementations.

Before we talk

Frequently asked questions.

Tableau Cloud or Tableau Server?

The choice depends on infrastructure responsibility, network, security, integrations, administration and company policies. Dashboard design works on both; publishing and maintenance differ.

When should we use live and when extract?

Live suits current-data needs and a well-prepared source. Extract often improves speed and reduces source load but requires scheduled refresh and monitoring.

Can data be restricted by user?

Yes. Options include user filters, data policies and database-enforced security. The choice follows risk, scale and data governance.

Should shared data sources be published?

Yes, when multiple workbooks share data and definitions. It reduces duplicated connections, extracts and calculations but needs an owner and versioned change process.

Can existing dashboards be optimised?

Yes. I assess sources, queries, calculations, sheet count, rendering and Performance Recorder output, then prioritise redesign.

How should a team prepare for self-service BI?

It needs trusted sources, roles, projects, publishing rules, templates and training matched to Viewer, Explorer and Creator.

First step

Describe the Tableau environment, sources and three decisions the dashboards should support.

I will follow up on Cloud or Server, audiences, licences, projects, refresh, RLS, performance and data readiness.

Discuss Tableau implementation View management reporting

Prepare the project

Review the portfolio and BI options.

Tableau Public portfoliovisualisation and analysis examplesView portfolio BI tools comparisonTableau, Power BI, Looker Studio and alternativesCompare tools Marketing data hubsources, warehouse and reportingExplore resources