Should you repair or rebuild an inherited WordPress site?
Repair an inherited WordPress site when the foundation is supportable and the defect is bounded. Rebuild when core structure, security, performance or data relationships prevent reliable changes and repeated patches cost more than a planned replacement. Start with access, a backup and a written assessment; do not begin by changing production blindly.
Last reviewed
The visible symptom rarely identifies the root cause
A slow page, broken form or failed update can come from one isolated component or from a site whose architecture makes every change risky. Replacing the theme before tracing the failure may preserve the underlying problem; rebuilding everything can also waste budget when a repair is enough.
Assess before touching production
- Confirm ownership, administrator access, hosting access, repository and current support contacts.
- Create and verify a restorable backup before changes.
- Inventory plugins, theme code, integrations, scheduled jobs, forms and data dependencies.
- Reproduce the reported issue on staging and identify its direct and upstream causes.
- Compare a bounded repair with a staged rebuild, including migration, regression testing and future maintenance.
Choose the smallest option that removes the constraint
A repair is appropriate when the cause is localized, the surrounding system is maintained and acceptance can be tested. A rebuild is justified when structural limits keep recurring or safe changes cannot be made in place. Kaprel’s inherited catalogue was taking around ten seconds on average and fifteen to twenty on its worst pages; rebuilding the foundation brought pages under one second while retaining the media library.
Set a baseline before either path
Record the exact failing journey, error rate, page timings, affected data and recovery steps. After the change, repeat the same checks and monitor regressions; do not compare a controlled staging test with an unrelated production sample.
Common questions
Should a new team update all plugins immediately?
Not before access, compatibility, backup and rollback are understood. Stage updates and test critical journeys.
How do I know an inherited site is safe to work on?
Verify ownership and access, make a restorable backup and inspect the hosting, code and dependencies before production changes.
Does a slow site always need a rebuild?
No. Measure the bottleneck first. Rebuild only when the architecture prevents durable repair or creates unacceptable ongoing risk.
What should an assessment deliver?
A prioritised risk list, reproducible findings, repair options, assumptions and a recommendation tied to cost and operational impact.
Kaprel: a rebuild justified by the measured foundation
Average load changed from around ten seconds to under one; the worst pages had taken fifteen to twenty seconds, and the media library remained intact. The case explains why patching was not the right boundary there.
Get the handoff measured before changing it.
Send the URL. We will write up what we can verify and what remains unknown within two working days.