Data Stack Index / v 02.06
Verified 2026·07·03
Send a correction
§ Report · Published 2026·07·03

The state of data observability, 2026.

What 24 independently verified tools — every data quality, catalog, and lineage product in this index — reveal about pricing opacity, the consolidation wave, and open-source economics. Every number below is derived from the open dataset at build time; nothing is hand-maintained or vendor-supplied.

67% of vendors publish no pricing
4 acquisitions since Jan 2025
9/24 open-source licensed
8/24 speak OpenLineage at all
01
Pricing transparency

The industry's quiet price tag.

16 of 24 tools (67%) publish no pricing at all. Among the 8 that do, 5 are free open-source projects — leaving 2 commercial products with a verifiable dollar figure, starting between USD 750 and USD 799. A free tier exists on 12 of 24 (50%).

Quality & testing 7/11 sales-led
Catalog & discovery 7/10 sales-led
Lineage & metadata 2/3 sales-led

Sales-led = no dollar figure anywhere on the vendor's site, verified per tool. Per-tool detail on the vertical index and each tool's pricing page.

02
Consolidation

The acquisition wave.

6 of 24 tools (25%) have been acquired — 4 of them since January 2025, as warehouse and platform vendors buy their way into metadata. The buyers are the platforms themselves: Snowflake, Atlassian, Datadog, Cloudera, IBM. Standalone tools are becoming platform features.

03
Open source

OSS economics.

9 of 24 tools are open-source licensed and 1 more source-available; 9 can be self-hosted at no license cost. But only 5 are free-forever community projects — 4 run an open-core model where the paid cloud carries the roadmap. The cautionary tale is maintenance: community projects without a commercial engine can stall (Amundsen's last release predates this report by nearly two years), which is why migration paths out of stalled catalogs are among the best-documented in the category.

Open-source licensed 9/24
Self-hostable, no license cost 9/24
Open-core (paid cloud on top) 4/24
Free forever, community-run 5/24
04
Capability table stakes

What "standard" actually means.

dbt integration is now table stakes — 12 of 24 tools integrate natively and only 3 have nothing. Column-level lineage ships in 12 of the 19 tools that do lineage at all. The open standard, though, still hasn't won: 1 tool is OpenLineage-native, 7 merely consume it, and 16 of 24 don't speak it at all.

Native dbt integration 12/24
Column-level lineage 12/24
Any OpenLineage support 8/24
SaaS-only (no self-host path) 5/24

Capability values come from each tool's structured spec — see the concepts glossary for definitions and the methodology for the scoring rubric.

05
Use the data

Cite it, check it, take it.

Every statistic on this page regenerates from the tool dataset on every deploy — the numbers can't drift from the source. The full dataset is downloadable as JSON or CSV from the open data page, with per-tool verification dates. When citing, link this page; the URL is stable and year-stamped.

Spot a number that doesn't match a vendor's current reality? That's a data correction, and those are the ones we fix fastest — send a correction.

About this report.

Compiled from 24 tool entries verified by hand against public sources — vendor docs, pricing pages, repositories, and press — with no vendor-supplied content and no paid placement. Most recent verification: 2026·07·03. Selection criteria live on the coverage page.