WordPress backup recovery for a site that came back incomplete
Restoring an older backup, or moving a site to a new host, can bring WordPress back with gaps: a homepage without its content, missing menus, images with missing files or wrong paths, plugin settings that never came across, or a database table that had to be recreated. Before changing anything, we find out what the backups still contain. Then we recover what exists, rebuild only the parts you approve and test the changes on staging before the live site is updated.
Norml recovers WordPress sites that a backup restore or a host move left incomplete. The work covers a comparison of the backups with the restored site, recovery on staging, tests before release, and then the updates you approve and a handover record of the code, hosting and licenses.
What recovery and modernization cover
Every stage ends with something you can check: a list, a test result or a record. The diagnosis can be agreed on its own, before any recovery work.
A safe copy before any changeWe copy the current files and database and keep the original backup archives untouched, so every later step can be compared against them or undone. The work itself happens on a staging copy.
01
Backups compared with the restored siteEach backup archive, database export and uploads folder is compared with the restored installation: page and homepage content, page-builder data, menus, media files, theme and plugin files, plugin settings and custom tables. We trace each gap to its cause where the evidence allows, for example a table left out of an export or old file paths stored in the database.
02
A recovery plan you approveThe findings are sorted into three lists: what the backups can restore, what has to be rebuilt and what is lost. Each rebuild item comes with the simplest replacement we can suggest, and none starts without your approval.
03
Content, menus, media and settings put backRecovered homepage content goes back into the editor or page builder it was made with. Menus and internal links are rebuilt, media files found in the backups are put back, and file paths stored in the database are corrected after a dry run shows what will change. Theme and plugin settings, shortcodes and custom post types are checked one by one.
04
Database tables and the code that uses themA table that had to be recreated is checked against the plugin or theme code that reads and writes it: its columns, indexes and expected data, and the features that depend on it. Where an older export still holds the data, it is imported and its row counts are compared with that export.
05
The new hosting brought in lineOn the new host we check the site address saved in the database, server paths in configuration files, the PHP version and extensions, permalink rewrite rules, file permissions, scheduled tasks, outgoing email, SSL and DNS. Settings that still point at the old server are corrected and recorded.
06
Tests, release and a way backBefore release we check key pages on desktop and mobile, navigation, media, forms, plugin features and integrations, PHP error logs, broken links and browser console errors. A fresh backup is taken and the rollback is agreed before the live site changes. You receive a list of the tests passed and the issues still open.
07
Updates in priority order, then a handover recordOnce the site works, we list improvements in priority order for your approval: core, plugin and theme updates, removing plugins nothing uses, security hardening, and analytics and conversion tracking. The handover record covers the code, database, hosting requirements, plugins and licenses, integrations, access and known issues.
Shows our staged method on a live WordPress site: we replaced a nutrition platform's Elementor build with a custom theme one part at a time, checked each release against the old version and removed deprecated Google Tag Manager tags.
Shows our work on an existing WordPress codebase: a New York coworking brand's site carried a decade of changes. We added Git, QA, a CI/CD pipeline and a staging environment and began refactoring its code in stages.
Shows a staged rebuild of a live store: over an 18-month partnership we rebuilt an instant photography store's site piece by piece, and the new version went live gradually while we automated review collection, customer support and warehouse fulfillment.
Every backup you haveAll backup archives and database exports, with the date of each and the host or plugin that made it. Older copies count too: content missing from one backup can survive in another.
01
Access to the current and previous hostingInvitations to the WordPress admin and the new hosting account, with its database tools or SSH, created for each person on our team so you can remove them at any time. If the previous hosting account is still open, access to it as well, since files can survive there.
02
Evidence of the site before the restoreScreenshots, old page addresses, a sitemap, analytics or Search Console access, and any copies in web archives. Where the backups are silent, these show what the recovered site should contain.
03
Notes from earlier workWhat the previous developer restored or changed, the theme and custom plugin code or its repository, and licenses for paid plugins and page builders.
04
What happens to the site nextWhether the site will be kept, passed to another team or sold, so the updates and the handover record fit that plan.
05
One person who approvesA named person who signs off the recovery plan, each rebuild item and every release.
06
How a recovery runs
[Preserve]
Nothing is changed until the current site and every backup are copied and stored.
Step 01
[Compare and diagnose]
We run the comparison, trace the gaps and send you the lists of what can be restored, what needs rebuilding and what is lost.
Step 02
[Agree the plan]
You choose the recovery order and approve any rebuild items. Items you decline are recorded on the lost list with a note.
Step 03
[Recover on staging]
Approved items are restored on a staging copy and checked against the backups and the evidence of the old site.
Step 04
[Test and release]
Release follows once the agreed tests pass on staging and the backup and rollback are in place. The live site is checked again afterwards.
Step 05
[Update and hand over]
Approved updates go out in small, tested releases. The handover record, test results and open issues go to the site owner, and our access is removed unless further work is agreed.
No account layer between you and them. These are the Norml specialists this work runs through.
Let's chat
“I've spent 10 years working across development and marketing. I started as a developer, but design, SEO and advertising were already part of the job. That taught me to treat a website as one system: the business, the content, the design and the code have to work together. At Norml, I lead the architecture behind our work and stay involved wherever strategy, design, development and search meet.”
Yes. The comparison and the three lists are a separate first stage. Afterwards you decide whether we carry out the recovery, your own developer does, or it waits.
What happens to content that no backup contains?
It goes on the lost list together with whatever shows it existed, such as a screenshot or an archived copy of the page. Rebuilding it from that evidence is an option you can approve item by item, and some items may not be worth replacing.
How do you check that recovered content is right?
Each recovered page is compared with its copy in the backup and with screenshots or archived versions of the old site, where they exist. Any difference is listed for your review, and the release waits for your sign-off on the agreed test list.
Can the site stay on its new host?
Yes, if the host provides what the site needs. When it does not, for example a PHP version, an extension or enough memory for the plugins, we explain the options, such as a different plan or another host, before recovery continues.
Will the site keep its URLs and search visibility?
Recovered pages keep their original addresses. We check them against the old sitemap, and against analytics or Search Console data where you have it. An address that cannot come back is redirected to its closest replacement. Rankings also depend on things outside the site, so they are not promised.
Will you redesign the site while recovering it?
No. Recovery puts the original build back. A new design, new features or a different theme are separate items, each proposed with its reason and started only after you approve it.
Do you use AI tools in this work?
Yes, for repetitive parts: tracing which templates and plugins use a table or shortcode, comparing database exports, listing media references that point to missing files and drafting test scripts. A developer reviews every finding and every change, and changes are tested on staging before release.
What if the backups contain malicious code?
If the comparison finds signs of a compromise, such as unknown admin accounts or core files that differ from the official release, we stop and agree on a cleanup first. Restoring from the same backups would otherwise bring that code back.
We plan to sell the site. What does the handover give a buyer?
A buyer gets the same handover record, written so their technical reviewer can check each item on the live site, along with the test results and the issues still open. Moving the site and its accounts to the new owner after the sale is a separate handover project. Valuing the business and financial due diligence are not part of our work.
What falls outside this work?
Ongoing maintenance after the handover can be agreed as its own arrangement. Performance optimization and SEO work beyond keeping existing addresses are scoped separately. Hosting fees and paid plugin licenses are not included.
See how we work on live websites
Each case study covers the starting point, the work we did and what changed afterwards.