

A WooCommerce backup has a different job than a blog backup. If a blog loses a day, someone rewrites a post from a draft. If a store loses a day, that day of orders is gone: the payments already happened at the gateway, but the orders, addresses, and stock changes that belonged to them do not exist anymore. Nobody can rewrite those from memory. That single difference is what should drive your backup interval, your storage location, and above all whether you have ever tested a restore.
WooCommerce backup scope overview with three columns for database, files, and configuration, plus a restore tested status bar
Most backup advice was written for content sites, where a nightly snapshot is genuinely enough. A store breaks that assumption in two ways.
Your content changes when you decide it changes. Your order data changes when customers decide it changes, which is to say at three in the morning and during the exact hour your host runs its nightly snapshot.
Every order writes to several tables at once: the order record, its line items, its meta, the analytics lookup tables, stock levels on the products involved, sometimes a coupon usage counter and a customer record. Restoring a backup rolls all of that back to the moment it was taken. So the question is never "do we have a backup", it is "how many orders sit between that backup and now".
Refunds, subscription renewals, and status changes count too. An order that was refunded after your last backup comes back as unrefunded, which is a bookkeeping problem rather than a missing row, and harder to notice.
A useful backup plan starts with knowing which parts of your store move fast and which do not. Store data is not one uniform blob.
Where WooCommerce data lives and how fast it changes
| What | Where it lives | How fast it changes |
|---|---|---|
| Orders, legacy storage | wp_posts and wp_postmeta | continuously |
| Orders, HPOS order tables | wp_wc_orders, wp_wc_order_addresses, wp_wc_order_operational_data, wp_wc_orders_meta | continuously |
| Line items | wp_woocommerce_order_items, wp_woocommerce_order_itemmeta | continuously |
| Analytics and lookup tables | wp_wc_order_stats, wp_wc_customer_lookup, wp_wc_order_product_lookup | continuously |
| Customers | wp_users, wp_usermeta | daily |
| Products and stock | posts, postmeta, wp_wc_product_meta_lookup | daily to weekly |
| Settings, gateway config, tax rates | wp_options, wp_woocommerce_tax_rates | rarely |
| Product images, invoices, downloads | wp-content/uploads | weekly |
| Theme and custom plugin code | wp-content/themes, wp-content/plugins | on deploy |
Two practical consequences follow. First, the fast moving data is almost entirely in the database, which is also the smallest part of your site, so backing the database up more often than the files costs very little. Second, if your store still uses legacy order storage, your orders are mixed into the same wp_posts table as your pages, so "just restore the orders" is not a thing you can do without a plan. If you are migrating to HPOS, read our WooCommerce HPOS migration guide before you change how orders are stored and backed up.
Callout
Four things matter, and one commonly backed up category mostly wastes space.
WooCommerce Database Backup
This is the part that actually holds your business. Orders, customers, products, coupons, tax rates, shipping zones, and every setting you have ever clicked live in the database, not in your files. A full database dump including all WooCommerce tables is the non-negotiable core of a store backup.
Back up all tables, not a selection. Custom tables added by extensions (subscriptions, bookings, points, license keys, custom order fields) are easy to leave out of a hand written table list and impossible to reconstruct.
Themes and Custom Plugins
Your child theme, any custom plugin, and the snippets in functions.php are code nobody else has a copy of. If it is not in a Git repository, your backup is the only copy, and you should treat it accordingly.
Off the shelf plugins and themes from a repository or a vendor can be reinstalled, so they are lower priority. Their settings are in the database, which you are already backing up.
wp-content/uploads
is the largest and slowest changing part of your site, and it holds more than pretty pictures:
wp-content/uploads/woocommerce_uploads which is where WooCommerce keeps downloadable products and, in many setups, generated PDF invoices. Losing invoices is a compliance problem in several jurisdictions, not just an inconvenience.Backing up less makes backups faster, which in practice means you run them more often:
Your backup interval is a decision about how many orders you are willing to lose. That is the whole calculation, and it is worth saying out loud, because "nightly" is a habit rather than a decision.
WooCommerce backup frequency versus data loss risk: weekly, nightly, and every four hours plus continuous database compared by orders at risk
Order volume versus backup interval
| Order volume | Database backup | File backup | Worst case loss |
|---|---|---|---|
| Under roughly 10 orders a day | daily | daily to weekly | one day of orders |
| Roughly 10 to 100 orders a day | every 4 to 6 hours | daily | a few hours of orders |
| Over roughly 100 orders a day | continuous or incremental | daily, incremental | minutes |
Low Volume Stores
Medium Volume Stores
High Volume Stores
Whichever tier you are in, back up before you change anything: plugin updates, theme deployments, WooCommerce updates, PHP version changes. A pre-change backup is a different thing from a scheduled one, and it is the one you actually end up using. Scheduling those windows properly is a large part of what a WordPress care plan is for.
Three approaches cover most stores, and most teams end up using two of them together. Plugin jobs handle routine schedules, host snapshots provide a floor, and manual exports give you a safety net before risky changes.
Backup plugins run inside WordPress, which makes them the easiest option to set up and the most convenient to restore from, because they handle the database, the files, and the URL rewriting in one flow. UpdraftPlus, BlogVault, and Jetpack VaultPress Backup are widely used examples with different models: some push archives to storage you own, others store the backup on their own infrastructure, and some offer incremental capture aimed specifically at stores.
What to check before choosing one, regardless of brand:
The shared weakness is that a plugin runs on the site it is protecting. If PHP is broken, the site is down, or the account is compromised, the tool you need is inside the thing that is broken. That is an argument for a second, independent copy, not against plugins.
Almost every managed host takes automatic backups, and they are genuinely useful: they are outside WordPress, they usually cover the whole account, and they cost you nothing to run. Treat them as your floor rather than your plan.
The limitations are consistent across hosts. Restores are typically whole account or whole site rather than a single table, so recovering three orders from Tuesday means rolling the entire site back to Tuesday. Retention is often short, frequently a week or less, and daily rather than hourly. Backups usually sit in the same provider account as the site, which is fine for a broken update and not fine for a compromised account. And you generally cannot verify what is inside a snapshot before restoring it.
Callout
A manual database dump is the right tool immediately before a risky change, and a useful sanity check that you can produce a complete file at all.
A manual dump is the right tool immediately before a risky change, and a useful sanity check that you can produce a database file at all.
With WP-CLI:
wp db export backup-$(date +%Y%m%d-%H%M%S).sql
wp db export - | gzip > backup-$(date +%Y%m%d-%H%M%S).sql.gz
wp db export orders-$(date +%Y%m%d).sql \ --tables=$(wp db tables '*_wc_order*' '*_woocommerce_order*' --all-tables-with-prefix --format=csv)On a store still using legacy order storage, that targeted export misses most of your order data, because it lives in wp_posts and wp_postmeta with everything else. Export the full database instead.
ls -lh backup-*.sql* && tail -5 backup-*.sqlA complete mysqldump ends with a Dump completed comment. If yours does not, the export was cut short and the file is not a backup.
Through phpMyAdmin, the equivalent is: select the database, open Export, choose the custom method, select all tables, pick SQL with DROP TABLE statements included, enable gzip compression, and download. It works, it is slower, and on large stores it tends to time out, which is exactly when you need it most.
Callout
Callout
The old 3-2-1 rule still holds up: three copies, on two kinds of media, one offsite. In practice, for a WooCommerce store:
A few details make the difference between offsite and merely elsewhere. Use credentials that are not reused anywhere on the site, so a compromised site cannot delete its own archives. Prefer append-only or versioned storage where your provider offers it. And keep a tiered retention rather than a flat window: daily copies for two weeks, weekly for two months, monthly for a year. Slow problems, like a corrupted table or a gradually broken export, only get caught by the older copies.
Encrypt anything that leaves your infrastructure. A WooCommerce database contains names, addresses, email addresses, phone numbers, and order histories, which makes an unencrypted archive in someone's personal cloud drive a data protection problem regardless of whether it is ever touched.
Backup jobs report success for all sorts of files, including archives missing custom tables, dumps that timed out at 80 percent, and jobs that quietly skipped uploads because it grew past a size limit. The green checkmark tells you the job ran, not that the output is usable.
The failures are boring and common: a broken archive nobody opened, an incomplete table list, credentials that expired months ago so the last real offsite copy is from March, or a database dump with no matching file backup. Every one of them is invisible until the day it matters, and discoverable in an hour on any other day.
Do this once a quarter, and after any change to your backup setup:
The homepage loading proves almost nothing. Check the things that quietly depend on state:
Checkout is the most common breakage after a restore, and the reason is structural rather than unlucky: gateway keys, webhook secrets, session tables, and scheduled actions are exactly the things that do not survive a change of environment cleanly. A site that looks perfect can be unable to take a single payment.
Which is why a test order belongs in the recovery checklist, and why the check should not stop the day after. ShopMonitor runs your checkout in a real browser on a schedule and alerts you when a step stops working, so a path that broke during a restore does not sit broken for three days. It verifies the browser checkout and the order received page, not whether a message landed in an inbox, which stays a separate check. It does not create, store, or manage your backups: that stays with your backup tool and your host.
If restore testing keeps sliding off the list, it usually means nobody owns it. Document who runs quarterly restore drills, who signs off on backup job changes, and who verifies checkout after recovery. A WordPress care plan is one place teams record that ownership alongside update windows and recovery targets.
ShopMonitor runs automated checkout and form tests every day and alerts you the moment something breaks. No code changes. No plugins. Setup in 5 minutes.
No credit card required.

About the author
Written by the team at Winning Solutions, the agency behind ShopMonitor. We help WooCommerce stores treat restore testing as unfinished until a real browser checkout has passed, without turning monitoring into a backup product.