Beyond SQL Traceability: Understanding Horizontal vs. Vertical, Technical vs. Functional, and End-to-End Data Lineage

Why simply parsing SQL is not enough — and how a metadata-driven approach creates lineage that engineers can trust and business users can understand.

Summary

Data lineage is essential for trust, governance, and impact analysis — but most lineage diagrams end up being unreadable spaghetti derived from SQL parsing.
This article clarifies the four key perspectives on lineage (horizontal, vertical, technical, functional), explains end-to-end lineage, and shows why a metadata-driven, model-driven platform is the best way to achieve clear, trustworthy lineage for both engineers and business stakeholders.

💡 Not a Medium member? You can read this article for free using this friend link.

What Data Lineage Really Is

Data lineage is the record of how data moves, transforms, and relates as it flows from source to consumption.
It’s not just a technical artifact — it’s a foundation for:

• 🔍 Root cause analysis — trace broken metrics back to upstream data.
• 🔄 Impact analysis — know what will break if you change a column or table.