Browse all patterns

Backup & recovery checklist

Why backups matter

  1. Because you want to
Most data lives "in the cloud". That means: "on someone else's server." Accidental deletion, a compromised account, a vendor outage, or an employee leaving with the wrong access can all cause data loss. Backups are not about distrust — they're about making sure that for every system that matters, you have a copy somewhere you own.
  1. Because you have to
An expired backup that falls short of the legal minimum is a compliance risk, not just a data risk:
Legal retention periods (Netherlands)
  • Financial records: 7 years (wettelijke bewaarplicht)
  • HR / employment records: 5–7 years after end of employment
  • Donor / personal data: per your privacy policy + AVG/GDPR; retain only as long as necessary, but ensure you can reproduce records if audited

Step 1: Know what you're protecting

Before setting up backups, map your critical data. You can't protect what you haven't inventoried.
List your critical systems and data sources one-off
Which systems hold data you can't function without? E.g. donor database, email, financial records, website, document storage.
Classify by impact one-off
For each system: what happens if it's unavailable for 1 hour / 1 day / 1 week? This sets your recovery priority.
Identify the data owner for each system one-off
One named person responsible per system.

Step 2: Set up backups

The 3-2-1 rule is the classic backup standard: 3 copies of important data, on 2 different media types, with 1 copy offsite or in a separate cloud account. For a 5–15 person org running entirely on SaaS tools, that's overkill.
Our recommendation: keep at least one copy of each critical dataset somewhere you control, outside the original vendor — something you could fall back on if the vendor disappeared tomorrow. In practice: an annual export to a restricted Drive folder, or monthly for donor data.

✳️ One-off checks

Google Workspace: Enable a third-party backup tool one-off
Google has no native backup tool. Vault (Business Starter+) handles compliance retention but can't restore a mailbox to a point in time. For real backup and restore of Drive and Gmail, use Backupify or Spanning. Without one, your only safety net is the 30-day deletion window in the Admin console.
Google Workspace: Confirm shared drives are in scope one-off
Personal drives and shared drives are separate — check both are covered.
Website: Confirm your hosting provider takes daily backups one-off
Most managed WordPress hosts (Kinsta, WP Engine, SiteGround) include daily backups — verify this and know how to trigger a restore. If using Webflow or another SaaS CMS, check the platform's versioning and export options: these vary significantly.
CRM/donor database: Check whether your CRM supports export one-off
Most SaaS CRMs (HubSpot, Salesforce, Donorbox, etc.) allow CSV or JSON exports. If your organisation uses a custom-built donor database, confirm automated database backups are configured on the hosting provider: daily frequency, at least 30 days retention, stored separately from the primary server.
Slack: Check your export options and set expectations one-off
Slack is not a system of record. On free and Pro plans, message history is limited and not fully exportable. On Business+ and above, admins can export all messages. Either way: decisions and outputs belong in Slite or Asana, not Slack.

🔄 Recurring back-ups

CRM/donor database: Run a recurring export monthly
Even with automated backups, a manual export to a restricted Drive folder adds a layer of protection — and gives you something portable if you ever switch CRMs.
Financial records: Export administration annually (at year-end)
At year-end, either your accountant pulls a full export as part of close (confirm this with them and verify you have access), or do it yourself. Required for the 7-year bewaarplicht.
Remote: Download contracts and payslips annually
Remote manages its own infrastructure backups. Your responsibility: download employment contracts and payslips to a restricted Drive folder once a year. These records must be retained for 5–7 years after end of employment (see legal retention periods above).
Slite: Export the full workspace annually
Slite is your primary knowledge base — losing it would be disruptive. Request a backup of your workspace content. Store the export in a restricted Drive folder.
Website/CMS: Take a manual backup before any major change
Before updating plugins, changing themes, or migrating: take a snapshot.
Remote: Download contracts and payslips at offboarding
Always download and store a departing employee's contract and payslips before closing their account in Remote. Make sure this is captured on the offboarding checklist so it actually gets done.

No action required

  • Asana: Operational, not archival. Deleted projects and tasks can be recovered within 30 days via Trash. Anything worth keeping long-term should be saved to Slite or Drive at the time of creation.
  • Exception: when you use Asana as your CRM/donor database⁠
  • Claude: Conversations are not a system of record. Any output worth keeping should be saved to Slite, Drive, or Asana at the time of creation. Claude has no persistent storage to back up.

Step 3: Test your recovery

Untested backups are not backups. Most organisations discover their backups don't work when they need them. Don't skip this section.
Do a test restore for each critical system annually
Pick one file, one record, or one snapshot — and actually restore it.
Time the restore annually
How long did it take? That's your Recovery Time Objective (RTO) — the realistic answer to "how long will we be down?"
Document the restore procedure after first successful test
Write it in plain language: where backups are stored, who has access, what tool to use. Keep a copy somewhere accessible even if your main systems are down.

Step 4: Access & security of backups

Restrict access to backup files one-off
Backups containing personal data must be limited to authorised staff. Same access rules as the original data.
Store backup credentials in your password manager one-off
Proton Pass, not a shared doc or personal notes.
Check that backups are encrypted one-off
Most cloud providers encrypt at rest by default — verify. For manual exports in Drive: restrict sharing, no "Anyone with link".
Configure emergency access one-off
Set up an admin recovery contact in your password manager. Document who holds it and how to invoke it.
Include backup access in your offboarding procedure one-off
When someone leaves, revoke their access to backup storage and the password manager before closing their account.

Accompanying tasks

15 tasks to put this pattern in place and keep it running. Untick anything you've already done. In Asana they all sit under one task, "Backup & recovery checklist".

Log in to create in Asana No Moral Fabric account?

Setup, one-off

Maintenance, recurring

7 Jul 2026 by Eva

15 accompanying tasks

10 to set it up, 5 to keep it running.

Log in to create in Asana See the tasks