Give your store a maintenance routine

Make maintenance a routine, not a surprise

Once a store is live, someone needs to own its health, orders, and recovery. AI can help interpret evidence and propose changes. You or your service provider decide what to run.

The routines below are an operating template for a supported live release. They do not turn monitoring, email, backups, or payment integrations on in the current local sample.

Each business day: look at the shop and orders

1. Open the homepage, one product, and the cart on a phone or narrow browser window. Check image loading and visible errors.
2. Review orders that need action and any payment-provider alerts. Match uncertain payments to their original order before taking action.
3. Check whether the most recent scheduled backup succeeded and whether failure alerts have an owner. If scheduling has not been configured, this check cannot be marked complete.
4. Read customer support requests in your normal support tool. Do not export their details to an AI chat for routine diagnostics.

Browsing a cart is not a reason to submit a new paid order every day. Use the release’s documented, noncharging health checks where available.

Each week: inspect changes and capacity

• Review errors since the last change, using redacted summaries.
• Check disk usage, media growth, and hosting/provider usage against your budget alerts.
• Confirm that backup files are available outside the running server and old copies are retained according to your recovery plan.
• Review who still has Admin, server, and provider access.
• Look at dependency/security notices without installing everything automatically.

Keep a log with: date, check, result, action, and owner. “No action needed” is a useful result when it is based on a check.

Each month: prove recovery and review costs

Practise a restore in a separate environment using the documented procedure. For the current sample, use the database restore rehearsal. For a live shop, also account for media, configuration, and provider records; a database-only rehearsal is incomplete disaster recovery.

Review hosting, storage, email, AI, and payment fees. Cancel unused services only after checking what depends on them. Keep your recovery contact and account recovery methods current.

Use this AI maintenance brief

Help me perform the documented maintenance review for my store version.
Start read-only. Report storefront availability, failed-job/error counts,
disk usage, latest backup age, latest restore-test result, and recent
changes only where these checks exist. Mark missing checks as unknown.
Use summaries; do not print settings, credentials, database rows, or
customer messages. Do not place orders, modify payments, restart services,
upgrade packages, or clean up files without a separate reviewed plan.
Return: issue, impact, evidence, smallest next step, and responsible person.

Decide when to get help

A broken checkout, unexplained payment mismatch, inaccessible Admin, failed restore, or suspected credential exposure deserves prompt attention. Use the troubleshooting lesson to preserve evidence. Pause affected operations through your approved incident process if continuing could lose orders.

An optional Store Care plan adds a scheduled monthly health and backup review and a defined allowance for small fixes or AI-change review. It is not continuous monitoring, emergency response, or running your shop for you. Daily order handling remains your responsibility. Compare the maintenance scope.

Ready when

Your calendar has an owner for daily, weekly, and monthly work; you know where the log and backups live; and you know whom to contact when a check fails. Marking this lesson read only records learning progress, not a maintenance check.