Tableau data blending is a method of combining data from two different data sources in a single view without a database-level join. Instead of merging data in the database, Tableau queries each source independently and combines results at the visualization layer. This guide explains how blending works, when it is appropriate, and its key limitations.
Tableau data blending is a method of combining data from two or more different data sources in a single visualization without joining them in the database. Where a join occurs inside the database — combining tables before data is returned to Tableau — data blending occurs at the Tableau layer: Tableau queries each data source independently, then combines the results in the visualization using a left join at aggregated level.
How Data Blending Works
When you blend two data sources in Tableau, you designate one as the primary data source (the one that drives the view's dimensions) and one or more as secondary data sources (the ones being blended in). Tableau executes one query per data source, independently. The results are then combined in memory at the granularity of the blended dimensions — the linking fields that are shared between the sources.
The linking field is the column that matches records between the primary and secondary source — analogous to a join key, but applied after the queries have already executed rather than inside the database. Tableau performs a left join at aggregated level: every row from the primary source appears in the result; matching secondary aggregates are brought in where a linking field match exists.
When Data Blending is Necessary
Data blending was originally introduced to handle a specific architectural constraint: combining data from two Tableau data sources that cannot be joined at the database level because they live in different systems.
**Cross-database analysis:** Combining data from a Snowflake data warehouse (operational transaction data) with data from a Google Sheets document (manually maintained targets or mappings) in one view. The two sources are in different systems with no shared query layer — database-level join is impossible. Blending is the alternative.
**Aggregated target vs. actuals:** A common use case is comparing actuals (from the data warehouse, at transaction grain) with targets (from a planning tool or spreadsheet, at aggregate grain). If targets are only available at monthly or regional level, blending at that level works well — the secondary source contributes aggregated targets, not row-level data.
**Supplemental reference data:** Adding external benchmark data, market data, or reference values maintained outside the warehouse to enrich a dashboard built from warehouse data.
Key Limitations of Data Blending
Data blending has significant limitations that make it a workaround rather than a preferred pattern:
**Left join only at aggregated level:** Blending always performs a left join after aggregation. You cannot get an inner join or right join. You cannot blend at row level (pre-aggregation). If you need non-aggregated row-level combination, you need a database join, not blending.
**Calculated fields across blended sources are restricted:** In secondary sources, only aggregated measures can be used in calculations — you cannot reference individual dimension members or use row-level functions from the secondary source. Calculated fields across sources are limited to aggregated operations.
**Limited LOD expression support:** LOD expressions in secondary sources are restricted. FIXED LODs referencing dimensions not in the view are particularly problematic in blended views.
**Performance:** Blending executes multiple queries and combines results in Tableau's memory, rather than letting the database perform a single optimized join. For large datasets, blending is significantly slower than a database-level join.
**Not available in Tableau Prep:** Data blending is a Tableau Desktop/Server visualization feature, not a data preparation feature. Tableau Prep uses database-level joins.
Data Blending vs. Data Joining
For data in the same database or queryable by the same connection, a join is always preferable to blending. A join executes in the database, leverages the database optimizer, uses the full power of SQL, and avoids the limitations above. Use blending only when the data sources genuinely cannot be joined — they live in different systems, or one source is a file rather than a database.
The recommended pattern: if you find yourself using data blending frequently for the same combination of sources, consolidate those sources into the data warehouse using an EL pipeline. The short-term convenience of blending becomes a long-term performance and governance liability at scale.
Blending Best Practices
When blending is genuinely necessary:
Set the linking field explicitly (right-click the field in the secondary source, set to Linked) rather than relying on Tableau's automatic linking by field name — automatic linking can produce unexpected results.
Keep the secondary source simple: blend at the highest granularity the secondary data supports, minimize calculated fields in the secondary source, and pre-aggregate in the secondary source's query rather than relying on Tableau's aggregation.
Test blend performance early: if the secondary source is large and blending is slow, consider materializing the combination as a view or table in the warehouse rather than blending dynamically.
Our Tableau consulting services covers data source architecture, blending optimization, and migration of blends to database-level joins. Contact us to discuss your Tableau data architecture.
A former Microsoft data architect audits your data foundation, identifies your top priorities, and sends you a written plan. Free. No pitch.
Book a Call →