# Lineage stops at the BI layer because each BI tool hides its logic in its own format and needs its own connector

GUIDE, read 10 Nov 2026

**The short answer.** Lineage usually ends at a dashboard because BI logic lives in LookML, DAX, workbooks and proprietary APIs that a SQL parser cannot read. DataHub, Euno, Foundational and Select Star document field-level BI lineage; 11 other tools reach BI partially, and OpenLineage's core spec shows none found on 10 Nov 2026.

## Most graphs end at a dashboard, not at a field

Trace a wrong number back from a dashboard tile and you meet the same wall in most tools. The warehouse side is detailed: tables, columns, dbt models. The BI side is a box labeled with a dashboard name. Between the box and the tile sits the part you needed, which is the calculated field, the filter and the measure.

Foundational and Select Star are built to go further, and the rest of this guide shows where each tool stops.

## BI logic lives in formats a SQL parser cannot read

A warehouse tool can read query logs. A BI tool does not run its logic in the warehouse in a form you can log. Looker keeps logic in LookML. Power BI keeps measures in DAX and data shaping in M-Query. Tableau keeps calculations inside workbooks and, for prep work, in Prep flows. A tool that wants field lineage must read those formats or call the BI vendor's metadata API.

Foundational says it supports Power BI including M-Query and DAX, Tableau including Prep Flow, and Looker including LookML. Coalesce Catalog says it maps Looker column-to-field links from LookML and Tableau column-to-field links through the Metadata API. Select Star says it reverse-engineers logic in BI dashboards.

Each of these is a separate piece of engineering per BI product. That is why coverage is a list of named tools, not a switch.

There are two ways to bridge the gap. One is to parse the BI tool's definitions and call its APIs, which is what the tools above say they do. The other is to register the BI object as an external asset and draw an edge to it. Databricks documents the second: external assets such as Salesforce, MySQL, Tableau and Power BI can be registered in Unity Catalog and linked into lineage. That gives you an object to point at. It does not tell you which field inside the workbook reads which column.

The difference shows up on the day a field is wrong. An asset-level edge tells you which dashboard to open. A field-level edge tells you which tile, and which calculation inside it, to inspect.

## Each BI tool needs its own connector, and the limits are documented

The documented limits are specific. They are worth reading before you assume your BI tool is covered.

| Tool | Documented BI limit | Where it applies |
| --- | --- | --- |
| Atlan | BI lineage depends on a BI crawler and works only for supported warehouse-to-BI pairs, which vary by BI tool | Atlan docs |
| Collibra | Power BI lineage excludes DirectQuery and live connections; Qlik metadata is ingested but technical lineage is not yet available | Collibra docs |
| Secoda | Column-level lineage to BI fields is documented only for Looker and Tableau among visualization tools | Secoda docs |
| Monte Carlo | Documents Tableau field-level lineage and says field lineage with other BI tools is still being developed | Monte Carlo docs |
| Coalesce Catalog | Sigma and Omni have asset-level lineage only, with no column or field-level lineage | Coalesce docs |
| Databricks | External assets such as Tableau and Power BI can be registered in Unity Catalog and linked into lineage | Databricks docs |
| Snowflake | BI collection through Horizon metadata connectors for Tableau and Power BI is in Preview | Snowflake docs |

## Which tools document field-level BI lineage

Four tools carry a yes in the BI field column of the matrix: DataHub, Euno, Foundational and Select Star. DataHub's lineage guide shows Snowflake tables orchestrated by dbt feeding Looker views and explores, and supports chart-to-dataset and chart-to-dashboard edges. Foundational lists Power BI, Tableau, Looker, Sigma and QuickSight. Select Star claims cross-platform column-level lineage tracking BI dashboards back to source systems. Euno advertises column-level lineage from data sources through to BI dashboards and AI agents.

Eleven tools are marked partial: Alation, Atlan, Coalesce Catalog, Collibra, Databricks, dbt, Manta, Monte Carlo, Secoda, SYNQ and Snowflake. Partial means the BI hop is documented for some BI tools only, or only down to the report or dashboard.

OpenLineage is the one tool with none found on 10 Nov 2026. Its core spec and integrations pages describe jobs and datasets, and no page read documents BI-field or metric lineage.

| Tool | Column | BI field | Metric |
| --- | --- | --- | --- |
| DataHub | yes | yes | none found |
| Euno | yes | yes | partial |
| Foundational | yes | yes | none found |
| Select Star | yes | yes | partial |
| Alation | yes | partial | none found |
| Atlan | yes | partial | none found |
| Coalesce Catalog | yes | partial | partial |
| Collibra | yes | partial | partial |
| Databricks | yes | partial | none found |
| dbt | yes | partial | partial |
| Manta (IBM) | yes | partial | none found |
| Monte Carlo | yes | partial | none found |
| Secoda | yes | partial | partial |
| Snowflake | yes | partial | partial |
| SYNQ | yes | partial | none found |
| OpenLineage | yes | none found | none found |

## Metric lineage is thinner still

A metric is a defined measure, such as revenue, that sits between a column and a dashboard tile. No tool in the matrix carries a plain yes for metric lineage. Seven are marked partial: Coalesce Catalog, Collibra, dbt, Euno, Secoda, Select Star and Snowflake. The evidence behind partial is varied. Coalesce Catalog says it can answer which warehouse column feeds a report metric. Secoda's evidence shows dbt metrics imported as dictionary terms and Power BI measure names, with no metric-level lineage beyond that. Snowflake describes semantic views that define business metrics once for reuse.

Ask for your own metric, not the vendor's demo metric. Write down its definition first, so you can tell whether the tool's answer matches it. Pick one whose definition lives in a BI workbook rather than in the warehouse, since that is the case the evidence covers least. Coalesce Catalog's page shows how narrow the claim can be: it says it can answer which warehouse column feeds a report metric.

## How the scores treat the BI gap

Upstream weights depth of lineage at 25 of 100 points. Euno's total is 67.0, first of 16, ahead of DataHub at 62.5 and Secoda at 61.5. Euno is cloud SaaS only and publishes no dollar prices; a team that needs lineage inside its own network, or a full catalog around it, will pair it with another tool.

Foundational (58.0) and Select Star (46.5) also reach the field. Select Star's evidence file records no published pricing, no stated refresh frequency and undocumented per-user MCP scoping. See [the matrix](/matrix/) for the grid and [how we score](/method/) for the weights.

## A test for your own BI tool

The table above is a starting point, not a verdict. Vendors update connectors every month, and a limit recorded on 10 Nov 2026 may be gone by the time you read this. Check the vendor's current documentation for your exact BI product and its connection mode, since Collibra's Power BI exclusion of DirectQuery and live connections shows that the mode can matter as much as the product.

Pick one calculated field in your busiest dashboard, ideally one that has been wrong before. Write down the three warehouse columns it depends on. Ask each candidate to show the path from the field back to the raw table, and to show the reverse: change the warehouse column, then list the fields affected. If the tool stops at the workbook or dashboard, you have found where it ends for your stack.

## Sources

- Foundational: docs.foundational.io/en/articles/15082464-supported-languages-and-platforms.md: https://docs.foundational.io/en/articles/15082464-supported-languages-and-platforms.md (read 10 Nov 2026)
- Coalesce Catalog: docs.coalesce.io/docs/catalog/navigate/lineage.md: https://docs.coalesce.io/docs/catalog/navigate/lineage.md (read 10 Nov 2026)
- Select Star: www.selectstar.com: https://www.selectstar.com/ (read 10 Nov 2026)
- Euno: euno.ai: https://euno.ai/ (read 10 Nov 2026)
- Atlan: docs.atlan.com/product/capabilities/lineage/best-practices/capture-and-troublesh: https://docs.atlan.com/product/capabilities/lineage/best-practices/capture-and-troubleshoot-lineage (read 10 Nov 2026)
- Collibra: productresources.collibra.com/docs/collibra/latest/Content/CollibraDataLineage/r: https://productresources.collibra.com/docs/collibra/latest/Content/CollibraDataLineage/ref_technical-lineage-supported-data-sources.htm (read 10 Nov 2026)
- Secoda: docs.secoda.co/features/data-lineage.md: https://docs.secoda.co/features/data-lineage.md (read 10 Nov 2026)
- Monte Carlo: docs.getmontecarlo.com/docs/field-lineage: https://docs.getmontecarlo.com/docs/field-lineage (read 10 Nov 2026)
- Databricks: docs.databricks.com/aws/en/data-governance/unity-catalog/data-lineage: https://docs.databricks.com/aws/en/data-governance/unity-catalog/data-lineage (read 10 Nov 2026)
- Snowflake: docs.snowflake.com/en/user-guide/snowflake-horizon/horizon-context.md: https://docs.snowflake.com/en/user-guide/snowflake-horizon/horizon-context.md (read 10 Nov 2026)
- DataHub: docs.datahub.com/docs/features/feature-guides/lineage: https://docs.datahub.com/docs/features/feature-guides/lineage (read 10 Nov 2026)
- OpenLineage: openlineage.io/docs: https://openlineage.io/docs/ (read 10 Nov 2026)

Last reviewed 10 Nov 2026. Canonical: https://datalineagecompare.com/guides/why-lineage-stops-at-bi/
