A Review of the Six Stages of the Data Life Cycle#

🌱 Foundations 🔄 The Analysis Process & Data Life Cycle Lesson 010

Important

✨ AI-generated content. This page was written with the assistance of an AI language model and is provided as a learning aid. Despite careful review, it may still contain mistakes, omissions, or out-of-date information. Whether you are new to the topic, a team lead, or a senior practitioner, treat it as a starting point rather than an authoritative reference: read it critically and independently verify anything you act on (code, commands, figures, and factual claims) against official documentation and primary sources before relying on it.

Consolidating the cycle#

This lesson consolidates the life cycle into a compact reference: each stage, its core question, its typical owner, and its characteristic failure — the thing that goes wrong there and surfaces later as an analyst’s headache.

The stages at a glance#

Plan — What data, managed how, by whom, under what rules? Owned by data governance and the teams commissioning collection. Characteristic failure: undefined terms, so two departments capture “customer” differently and their numbers never reconcile.
Capture — How does the data enter? Owned by the systems and people at the point of entry. Characteristic failure: no validation at entry — free-text dates, optional required fields — producing the dirty data Section 4 cleans.
Manage — Where does it live, and who can reach it? Owned by data engineering and IT. Characteristic failure: silos and stale copies, where the warehouse and the team spreadsheet quietly disagree.
Analyze — What does it tell us? Owned by analysts. Characteristic failure: analysing without checking the upstream stages — precise answers from compromised inputs.
Archive — What do we keep, and can we still find it? Characteristic failure: archives nobody can query, so history is technically kept but practically lost.
Destroy — What must go, and did it? Characteristic failure at both extremes: deleting too eagerly (losing the baseline for next year’s comparison) or never deleting (hoarding personal data past its lawful purpose).

Two threads through every stage#

Security and privacy are not a stage; they are obligations at every stage — planned rules, protected capture, controlled access, careful analysis, encrypted archives, verified destruction. Likewise documentation: each stage should leave a record the next stage can rely on, which is exactly the chain of custody the detective lesson demanded.

Using this review#

Two practical habits fall out. When you receive a dataset, walk it backward — who manages it, how was it captured, what did the plan define? — before trusting it. And when an analysis will recur, walk it forward — will the data still exist, unarchived and legal to use, when the next cycle runs? Ten minutes of life-cycle thinking routinely saves days of confused analysis.

Hint