DField SolutionsMérnöki stúdió · Budapest
Loading · Töltődik
Skip to content
Back to glossaryTERM

Soft delete vs hard delete

Related service Websites, web apps & online shops

DEFINITION

Soft delete: the row stays in the table, only a deleted_at timestamp or is_deleted flag is set, and every query filters by it. Hard delete: DELETE FROM truly removes the row and the associated audit trail or backup is the only memory of the past. Soft delete is convenient: reversible, undo-able, easy to audit. On the GDPR side it is a trap. An erasure request under Article 17 (right to erasure) means a soft-deleted row that still holds personal data is not erased, and if your background jobs, analytics, ML feature store, or backups keep reading it, you have not honoured the request. Fix: two-stage deletion. A user-facing delete sets soft delete (30-day undo window), and a scheduled job after 30 days hard deletes the PII (or anonymises it and keeps the numeric columns for statistics).

RELATED TERMS06
  • Answer Engine Optimization (AEO)→

    We shape content so AI answer engines (ChatGPT, Perplexity, Google AI overviews) cite YOU: clear claims, structure, sources, schema. The successor discipline to classic SEO in 2026.

  • Zero-Click Search→

    Searches resolved right on the results page (AI overview, snippet) with no click to any site. Rising fast: you win by being the cited source, not just by ranking.

  • Featured Snippet→

    The boxed answer Google lifts to the top ("position zero"). Win it with a crisp, directly-phrased answer placed right under a question heading.

  • People Also Ask→

    Google's expanding box of related questions. Each one is a content opportunity: answer them explicitly to capture the box and the long-tail traffic.

  • Crawl Budget→

    How many pages a search engine will crawl on your site per visit. It matters for large sites: waste it on junk URLs and your important pages get crawled late.

  • Canonical URL→

    The <link rel="canonical"> that tells search engines which URL is the master when duplicates or params exist. Prevents ranking signals from being diluted across near-identical pages.

MENTIONED IN THE BLOG08