A BI platform is often chosen on the strength of a demo: polished charts, a familiar interface, the promise of "everything in one place". The daily work is about something else: how the platform prepares data, where the model lives, who sees what, and what happens at large data volumes.
I put five platforms into one map across ten criteria: Power BI, Tableau, Qlik, Looker and SAP Analytics Cloud. The map plays on its own, criterion by criterion. Pause it where your own question starts.
Open the map on a separate page
How to read the map
- Ruby spheres are the platforms. Glass beads on the orbit are the ten criteria.
- The active criterion moves to the centre and connects to each platform in turn. At that moment the line for that platform appears in the list.
- The key point in each line is in white, the rest is detail.
- Below the map you will find pause, navigation between criteria and a format switch: wide 16:9 or vertical 9:16.
Where to start
Comparing features makes sense after three questions about your own situation.
- Where the data lives and who prepares it. With a SQL warehouse and data engineers, the "Data preparation" and "Modelling and calculations" rows read differently than when everything starts in Excel.
- Who will build and maintain the model. The "Entry barrier" row shows where each platform gets hard: DAX, scripts, LookML or planning models.
- What is non-negotiable for you. On-premises deployment, row-level access restrictions, budgeting and planning. A requirement like that rules options out faster than comparing charts.
The comparison as a table
The same content without animation. Rows are criteria, columns are platforms.
| Selection criterion | Microsoft Power BI | Tableau | Qlik Sense / Qlik Cloud Analytics | Google Looker | SAP Analytics Cloud |
|---|---|---|---|---|---|
| Main strength | Combines data preparation, modelling and reporting at a moderate starting price | Flexible visual analysis and fine-grained control over charts | Exploring relationships in data through the associative model | Unified metric definitions in a centralised model | Combines analytics and corporate planning |
| Best suited for | Finance and operations analysts, and organisations that run on Microsoft | Analysts for whom visual data exploration is the core of the job | Teams with many sources that need to explore how they relate | Teams with a SQL warehouse and mature data engineering | Companies on SAP with budgeting and planning tasks |
| Entry barrier | Relatively low after Excel. Complexity grows with DAX and modelling | Basic charts are quick to learn. Complex calculations take practice | The basic interface is approachable, complex solutions need scripts and Set Analysis | Higher for the model developer. Business users work through prepared Explores | Moderate for reports, higher for models and planning processes |
| Data preparation | Power Query for cleaning, combining and transforming data | Tableau Prep for visual cleaning and building preparation flows | Data Manager and Load Script for loading and transformations | Most transformations usually run in the warehouse through SQL. LookML describes the analytical model | Import, data preparation and live connections. Tight integration with SAP |
| Modelling and calculations | Semantic models, table relationships, DAX measures | Relationships, calculated fields, LOD and table calculations | Associative model, expressions, Set Analysis | LookML models, SQL, version control through Git | Analytic and planning models, plan versions, Data Actions |
| Visuals and interactivity | A solid all-round set, plus additional visuals. Custom design takes extra work | Especially convenient for custom charts, maps and visual exploration | Strong at interaction through value selections and analysis of associated data | Convenient for standardised dashboards. Less freedom for bespoke design | Geared towards corporate dashboards, planning tables and management scenarios |
| Large data volumes | Import, DirectQuery, Direct Lake. The result depends on the model and capacity | Extract or live connection. The result depends on the source and query structure | What matters is app size, the model and available capacity | Computation is mostly done by the SQL warehouse. Its speed and cost become critical | Live or import. Performance depends on the model and the source system |
| Access control | Content permissions, data restrictions at row and object level | Content permissions, user filters and row-level access policies | Permissions on spaces and apps, Section Access | Roles, user attributes, access filters | Roles, teams, access to objects and model data |
| Cloud and on-premises | Cloud service or on-premises Power BI Report Server with a different feature set | Tableau Cloud or your own Tableau Server | Qlik Cloud or Qlik Sense Enterprise on Windows | Looker (Google Cloud core) runs in Google Cloud. Looker (original) has a customer-hosted option | Cloud platform that can connect to on-premises sources |
| What complicates the work most | Complex DAX, a poor model, licensing for report distribution | Complex calculations and maintaining a large number of separate reports | Scripts, the specifics of the associative model and app maintenance | The need for a prepared warehouse and LookML developers | Complexity of implementation, administration and planning models |
Sources
The "Data preparation" row is based on vendor documentation:
- Power BI Desktop, Microsoft Power Platform
- About Tableau Prep, Tableau
- Loading and transforming data with scripting, Qlik Sense on Windows Help
- Introduction to LookML, Google Cloud Documentation
- SAP Analytics Cloud Connectivity Guidelines, SAP Help Portal
Platform capabilities change with every release. Before you decide, check the points that are critical for you against the current documentation.
More on the approach: Business Intelligence and Data Architecture.
Which criteria decided the choice of BI platform in your team?
