When a number in a report looks wrong, someone will eventually ask: where did this data come from? Data lineage is the answer to that question — documented.
Data lineage is the record of where a piece of data originated, how it moved through your systems, what transformations it went through, and where it ended up. It's the audit trail of your data's life.
The Core Idea: Every Data Point Has a History
Think of data lineage like the provenance of an antique: where it came from, who owned it, and what happened to it along the way. For data, provenance matters because errors compound. If a field was miscalculated at step two of a five-step process, every downstream report that uses that field is wrong too. Without lineage, you can't find step two.
Data lineage answers three questions:
- Where did this data originate? Which system, form, file, or person created this record or value?
- How was it transformed? Was it cleaned, aggregated, merged with other data, converted from one format to another?
- Where is it used? Which reports, dashboards, exports, or systems depend on this data?
See exactly what's wrong with your data — try Sohovi free — try Sohovi free.
A Concrete Example
Imagine your weekly sales report shows revenue $12,000 higher than expected. With data lineage, you can trace that revenue figure backward:
- The report pulls from your analytics dashboard
- The dashboard aggregates from your CRM deal stage field
- The CRM was updated by a bulk import last Tuesday
- The import came from a spreadsheet exported from your billing system
- The billing system records a deal as "closed" when a payment is initiated, not confirmed
You just found the root cause in minutes instead of hours — because you can follow the chain.
Why Data Lineage Matters for Data Quality
Data lineage is especially valuable when data quality problems surface. Without it, investigating a wrong number means interviewing everyone who touched the data. With it, you can trace the path and isolate where the error entered.
Impact on trust: Teams that can't explain where their numbers come from lose credibility with stakeholders. Lineage documentation lets analysts say: "This revenue figure comes from X source, transformed by Y rule, and verified at Z step."
Impact on compliance: Regulations like GDPR and CCPA require organizations to know where personal data came from and where it flows. Data lineage is the mechanism for demonstrating that compliance. If a regulator asks "where did this customer record come from and where was it shared?" — lineage documentation provides the answer. In banking specifically, BCBS 239 (the Basel Committee's risk-data-aggregation standard) makes traceable lineage a regulatory requirement for risk reporting, not just a best practice.
Sohovi automatically detects PII in your datasets — emails, phone numbers, SSNs — all processed client-side so your data never leaves the browser.
Impact on change management: When you update a data process — change a formula, update an import mapping, modify a field definition — lineage tells you which downstream reports and systems will be affected. Without it, changes cause unexpected breakages.
Data Lineage at Different Scales
Small business without formal systems: Lineage is often implicit and informal. "The sales numbers come from the spreadsheet Priya exports from the CRM every Monday and pastes into the dashboard." That's lineage — it just isn't documented. When Priya leaves, the lineage knowledge leaves with her.
Teams with multiple data sources: Lineage becomes more complex. Data from three different systems gets combined in a spreadsheet, which feeds a report. Documenting this chain is the difference between "the report is broken" and "the import step failed because the CRM export format changed."
Sohovi finds gaps, duplicates, and format errors in your CRM data — so your team is working from records they can trust.
Larger organizations: Data lineage tools automate the tracking of data flows across systems, creating visual maps of how data moves through pipelines. Enterprise data platforms often have built-in lineage features. For most small businesses, a simple written or diagrammed record is sufficient.
Column-Level vs. Row-Level Lineage
There are two levels at which lineage is typically tracked:
Column-level lineage tracks which source fields contributed to each output column. If a "full_name" field was created by concatenating "first_name" and "last_name," column-level lineage records that relationship.
Row-level lineage tracks the journey of individual records — this specific customer record was imported from file X on date Y, merged with record Z, and then exported to system W.
For most business users, column-level lineage is what matters when investigating report discrepancies. Row-level lineage is more relevant for compliance and individual record audits.
How to Start Documenting Data Lineage Without Enterprise Tools
You don't need specialized software to implement basic data lineage. A structured Google Sheet or Notion document works:
For each key report or data output, document:
- What data sources does it pull from?
- What transformations or calculations are applied?
- Who is responsible for maintaining the process?
- How often is it updated, and when was it last verified?
For each data import or integration:
- Where does the incoming data originate?
- What field mappings apply (source field → destination field)?
- What validation or transformation happens before it's loaded?
Start with your most important report — the one your team uses to make real decisions — and document its lineage from source to output.
Sohovi helps at the first step: when you upload a CSV, it profiles each column so you can see whether the field formats match what your downstream process expects — a quick check for lineage integrity before importing.
Frequently Asked Questions
Q: Is data lineage the same as data provenance? The terms are often used interchangeably. Data provenance focuses specifically on origin — where did this data come from? Data lineage is broader — it covers origin, transformations, and destination. Provenance is a component of lineage.
Q: Do small businesses need data lineage documentation? If you have more than one data source feeding a report or decision, yes. Even simple documentation — "this report comes from that export" — provides value when something breaks or when someone new takes over the process.
Q: How does data lineage relate to data quality? When a data quality problem is found, lineage is how you trace it to its root cause. A missing value, wrong format, or incorrect aggregation can be traced back to where it entered the pipeline — which is where the fix needs to happen.
Q: What tools track data lineage automatically? Enterprise tools like Apache Atlas, Collibra, and Alation track lineage automatically for complex data pipelines. Modern data warehouses (BigQuery, Snowflake, dbt) have built-in lineage features. For smaller teams, documenting lineage manually in a shared document is a practical alternative.
Q: How does GDPR require data lineage? GDPR's right to erasure (Article 17) and data portability (Article 20) require organizations to know where personal data flows so they can delete or export it on request. Without lineage documentation, responding to a data subject request becomes a manual, error-prone search across every system.
Q: What's the difference between data lineage and a data audit trail? A data audit trail tracks changes to specific records over time (who changed what, when). Data lineage tracks how data flows and transforms across systems. An audit trail is about history of changes; lineage is about the path data travels.
Q: How granular does data lineage documentation need to be? Match the depth to the risk. For a critical financial report, document lineage at the column and transformation level. For a weekly newsletter list, documenting the source system and import schedule may be sufficient.
Q: Can data lineage help with debugging broken reports? That's one of its primary uses. When a report breaks or shows unexpected numbers, lineage lets you systematically trace back through each step to find where the error entered the pipeline — instead of guessing.
Data lineage turns "where did this number come from?" from a frustrating investigation into a quick trace. Start by documenting the path of your most important report — source system to final output. That one exercise will reveal more about your data quality than most audits.
