Quick answer
The data lifecycle, or data cycle, is the path data follows from the decision to collect it to its deletion: plan, collect, process, store, use, share, retain and delete. There is no single standard model; the stages matter because each one needs an owner and a rule, especially the last.
Formula: Delete-by date = collection date + retention days (2026-03-01 + 90 days = 2026-05-30). Rows held = monthly rows × months retained.
What is the data lifecycle?
The data lifecycle, also called the data cycle, is the path a piece of data follows from the decision to collect it to the moment it is deleted. There is no single standard model. The U.S. Geological Survey's Science Data Lifecycle Model has six primary elements (plan, acquire, process, analyze, preserve, publish and share) and three that run through all of them (describe, manage quality, back up and secure). Vendor and textbook versions use five to eight stages, and one widely read definition lists seven: determine the need, generation or capture, preparation, active use, management, long-term storage, and purging and destruction.
The names differ, but every version has the same three jobs: decide what to collect and why, use it, and retire it. The last job is the one most teams forget, which is why this entry spends most of its space on retention and deletion.
How do the stages map to analytics event data?
This table uses eight stages, merging the models above, and applies each to the data a website or app analytics project produces.
| Stage | What happens | Analytics event data example | Control to check |
|---|---|---|---|
| Plan | Decide which questions need which data | A tracking plan lists each event, its properties, an owner and a status | Is every event tied to a question? |
| Collect | Data is captured | Page views, clicks, form submits, purchases from a script, SDK or server call | Is only what the plan needs collected? |
| Process | Data is validated, cleaned, enriched | Dropping bot traffic, checking required properties, naming new events | Do required properties arrive complete? |
| Store | Data is kept in a system | Event rows and daily aggregates | Who can read it, and in which region? |
| Use | Data answers questions | Funnels, retention, attribution, dashboards | Is the metric definition written down? |
| Share | Data leaves the team | A read-only report link, a raw export | Is the audience and the expiry controlled? |
| Retain | Data is kept for a reason and a time | Event rows for a set number of days | Is the number of days chosen, not defaulted? |
| Delete | Data is removed when the time is up | A scheduled deletion job, an erasure request | Is deletion verified and logged? |
Notice that the cross-cutting USGS elements, metadata and quality, are not a stage but a thread: a tracking plan is metadata, and a check on required properties is quality management.
Worked example: how long does each row live, and how many rows is that?
Sample data: a project sets four retention periods in days, using the starting values of the product described below: event rows 365, autocapture rows 90, person profiles 730 and consent receipts 730. A row collected on 1 March 2026 is deleted on:
| Data | Retention | Delete-by date |
|---|---|---|
| Autocapture row | 90 days | 30 May 2026 |
| Event row | 365 days | 1 March 2027 |
| Person profile | 730 days | 29 February 2028 |
| Consent receipt | 730 days | 29 February 2028 |
Retention also sets storage. If the project collects 500,000 event rows a month, the steady-state number of rows it holds is monthly volume × months retained:
- retention of 365 days: 500,000 × 12 = 6,000,000 rows
- 90-day retention: 500,000 × 3 = 1,500,000 rows, which is 75% fewer.
Daily aggregates, which contain no personal data, are not on this clock; that is why a chart from two years ago can still draw while the underlying rows are gone. The practical rule: retention is the dial that decides how much personal data you hold at any moment, so set it on purpose.
What do privacy law and standards ask at each stage?
The GDPR's Article 5 principles line up with the lifecycle stages:
| Principle (Art. 5) | Wording | Stage it governs |
|---|---|---|
| Purpose limitation, 5(1)(b) | "collected for specified, explicit and legitimate purposes" | Plan, Collect |
| Data minimisation, 5(1)(c) | "adequate, relevant and limited to what is necessary" | Collect, Process |
| Storage limitation, 5(1)(e) | identifiable "for no longer than is necessary" | Retain, Delete |
For the deletion stage itself, NIST Special Publication 800-88 Revision 1 (December 2014) is the standard U.S. guide to media sanitization, defined as making access to data on media infeasible for a given level of effort; NIST has since withdrawn it on 26 September 2025 in favour of Revision 2. A database row deleted by a job is one case; the same data in backups, exports and downloaded files is another, and each copy has its own end of life.
How is the data lifecycle different from data lineage and a retention policy?
- Data lifecycle: the stages one dataset passes through, from plan to deletion.
- Data lineage: where a specific number came from and what it was transformed by, which is a record of the Process and Use stages.
- Data retention policy: the rules for the Retain and Delete stages: how long each type of data is kept and how it is removed.
One lifecycle has many lineages (every metric has its own) and one retention policy.
What goes wrong in practice?
- No delete stage. Collection is switched on and retention is left at a default nobody chose.
- One retention for everything. Autocapture rows, profiles and consent receipts need different periods.
- Forgetting copies. An export sent by email outlives the system it came from.
- Treating aggregates as raw data. Aggregates without personal data can be kept; raw rows with identifiers cannot be kept indefinitely.
- Deleting the definition instead of the data. Removing an event definition from a plan does not remove the history already collected.
How does the lifecycle look in TapCub?
Here is the same table in TapCub's menus. TapCub controls the stages inside the product; copies you export or share are yours to track.
- Plan: the tracking plan under All-platform analytics › Data › Tracking plan records the events you intend to collect, with properties, platform, owner and a status: Proposed, Developing, Accepting, Live or Deprecated.
- Collect and process: one script, SDK or server call feeds the project. New events land in the Event inbox and are claimed before they become official definitions; the tracking plan's live check for the last 24 hours shows how many of each event arrived and how complete the required properties are. Bots are counted separately from people.
- Use and share: analyses read the stored events; Configuration › Settings › Public sharing makes a read-only link, optionally password-protected, with selected modules and a limit on how many days are visible; visitor lists and sessions cannot be shared. All-platform analytics › Data › Data export runs an asynchronous raw export, which is not a screenshot of a report.
- Retain: in Configuration › Privacy center the four periods, events, autocapture, person profiles and consent receipts, are set separately. The starting values are 365, 90, 730 and 730 days. Event rows follow the project's data-retention days, capped by the plan.
- Delete: a scheduled job deletes expired detail rows. For one person, create a request in the same Privacy center, choose access, export, restrict or erasure, and find them by email, login ID, device ID or person ID. Restrict and erasure stop new events for that person at once, then wait out a cooling-off period, three days to start with, which you can cancel. Daily aggregates hold no personal data and are not deleted on these timers. Deleting a data source or an event definition does not purge history. Only the owner and admins can open the Privacy center.
Sources
- U.S. Geological Survey, Data Lifecycle: six primary and three cross-cutting elements of the Science Data Lifecycle Model. Checked 2026-11-20.
- Regulation (EU) 2016/679, Article 5, principles relating to processing of personal data: purpose limitation, data minimisation, storage limitation. Checked 2026-11-20.
- NIST, SP 800-88 Rev. 1, Guidelines for Media Sanitization (December 2014; withdrawn 26 September 2025, superseded by Rev. 2). Checked 2026-11-20.
- TechTarget, Data life cycle: a seven-stage definition. Checked 2026-11-20.
Was this explanation clear?
We use your answers to fix and expand this page.
Couldn’t send. Please try again in a minute.
Thanks for the feedback. We’ll use it to improve this page.You’ve already sent feedback on this page. Thanks.

Data Analyst
Explores traffic sources, conversion funnels, and retention, using clear metric definitions and concrete examples to explain what the numbers mean.