Lorixen.
EN

Recycle Bin or Real Delete? Choosing a Deletion Model for a Small Business App

Short answer: Most small business apps should not delete anything the moment someone clicks the bin icon. A recycle bin (soft delete) that keeps the record hidden for a set number of days, followed by a scheduled permanent purge, protects you from slips and from staff who regret a click on a busy Friday. Use archive for records that finished their life normally, and reserve instant permanent deletion for data you are genuinely obliged or asked to erase. Decide the model per kind of record, not once for the whole app.

Three models in plain terms

Hard delete removes the row for good. The database frees the space and no trace remains apart from backups.

Soft delete keeps the row but marks it, usually with a "deleted at" date and the person who did it. The app hides it from normal screens. It can be restored until a purge job removes it.

Archive is a status, not a deletion. A finished job, an old customer who has not ordered in years, a closed project: still visible in search with a filter, just out of the daily lists.

Comparison

Question Hard delete Recycle bin (soft delete) Archive
Can a mistake be undone from the screen? No, only from a backup Yes, until the purge date Yes, any time
Does it clutter daily lists? No No No, if filtered
Keeps history for reports? No Until purged Yes
Risk of keeping personal data too long Lowest Moderate, needs a purge Highest, needs a policy
Extra build effort Lowest Moderate Low to moderate
Fits Test data, drafts, erasure requests Contacts, quotes, files, notes Customers, jobs, products

How a recycle bin should behave

A good bin is a small feature with several moving parts. Write these into your specification.

  1. A visible Trash view that lists deleted items with the date, who deleted them and a Restore button. Without a screen for it, soft delete is only a developer convenience.
  2. A fixed retention window, such as 30 days, shown to users ("Items are permanently removed 30 days after deletion").
  3. A scheduled purge that really removes rows and their attached files when the window ends. This is the step that gets forgotten, and it is the difference between a bin and hoarding.
  4. Sensible restore rules. What if an invoice number was reused meanwhile, or a customer was merged? Decide whether restore is blocked, renamed or allowed with a warning.
  5. Cascades that make sense. Deleting a customer should not silently wipe their invoices if tax records must be kept. Often the right behaviour is to block deleting a customer with financial history and offer archive instead.

Which records get which treatment

Think record type by record type. As an illustration for a small renovation company's app:

  • Draft quotes and test entries: hard delete after a short confirmation.
  • Site photos and notes: recycle bin, 30 days.
  • Customers with any job history: archive only, never deletable from the normal screen.
  • Invoices and payment records: no delete at all. Void or credit them, which leaves a trail.
  • Staff accounts: deactivate, keep names on past entries.

This is an example, not a rule for your sector. Financial and employment records often carry retention requirements, so check the official guidance in your country, for instance with your tax authority or your accountant, before choosing a purge period.

Erasure requests and the bin

A person asking you to remove their personal data does not want it sitting in a bin for 30 days. Give administrators a separate, deliberate "erase now" action that bypasses the bin, removes the personal fields and keeps only what you are required to keep, for example an invoice with the name replaced by a reference. Privacy regulators publish plain-language guidance on this, and the rules differ by province and country, so use their material and professional advice instead of guessing.

Remember backups too. If a backup holds a deleted record, your policy should say how long backups live and what happens when one is restored.

Pitfalls to catch in testing

  • Search results showing deleted items. Filter the bin out of search, exports and reports by default.
  • Unique fields blocking re-creation. A deleted customer's email still occupies a unique slot, so staff cannot add the same person again. Make uniqueness ignore deleted rows.
  • Files left in storage. Photos and PDFs attached to purged records stay billed and stay exposed unless the purge removes them.
  • Counts that look wrong. Dashboards that count all rows will include the bin.
  • Slow queries. Millions of old hidden rows can drag lists. At that scale, moving old rows to a separate table or using partitioning, as described in the PostgreSQL partitioning documentation, becomes worth discussing.

FAQ

Is soft delete more work than hard delete?

Yes, a little. The cost is in the bin screen, restore rules and the purge job, not the hidden flag itself. Most owners find the first prevented accident covers it.

How long should the bin keep items?

Thirty days is a common starting point because it covers a month-end review. Shorten it for sensitive data and extend it for records that rarely get looked at.

Can staff delete things permanently from the bin?

Allow it for managers only, with a confirmation. Day-to-day staff should only restore.

What if we already hard delete everything?

Add a bin to the record types that hurt most when lost, usually contacts, jobs and documents, then expand. Existing deletions cannot be recovered except from backups.

If you are specifying a custom app and want deletion, restore and retention rules written down properly, send us a short description and we will draft them with you.

Sources

WhatsAppBook a call