Multi-page find and replace in WordPress: updating an old phone number on forty pages by chatting

The phone number changed, the office moved, you dropped a service, prices were updated. The information is not in one place but in forty pages, and nobody knows which forty. A stock WordPress install does not search inside content bodies; you need a plugin or the database.

Tecrube’s Website tool connects to your WordPress site and splits the job in two: find first, replace second. The split itself is the safety.

Search lists the posts and pages that contain the phrase, scanning the body rather than the title. Do not replace before you see that list; the phrase appears where you do not expect. The classic case: while replacing an old number you find it in a customer review or a years-old announcement, and you may not want to edit historical text.

Tell Tecrube “find the old number starting with 0212 across all pages” and the result arrives with page name, location and the surrounding sentence. You tick which ones change.

Keep the phrase long enough: a four-digit fragment matches in other contexts too; search the full form and run spaced and unspaced variants separately. Account for formatting: if part of the text is bold or an HTML tag sits in the middle, the string is not one piece; if search misses it, check the page source.

Take a backup: a bulk change is not a one-click undo. WordPress keeps revisions, but reverting forty pages individually takes longer than the replacement did. Tecrube stores a snapshot of the affected pages before applying.

A bulk operation can partially fail: twenty-eight of thirty pages update and two are refused. Calling it done without reading the result means those two carry stale information for months.

The reasons are mundane: the page is locked by another user, the post type is protected, or the content lives inside a page builder’s own data structure. That last one is sneaky; the page looks right in the editor while the content is stored elsewhere. Tecrube lists updated and skipped pages with reasons at the end; skipped ones are handled individually.

When the update reports success but the page still shows the old text, the problem is caching, not content. Check in order: the cache plugin (some hosts run their own layer and you purge from the host panel), the CDN (purge the relevant paths) and the browser (confirm in a private window).

Following that order is faster than re-running the replacement because it “did not apply”, and it avoids the risk of a double change.

Internal links left in the body after a domain move, old figures still in copy after a price update, links to a social account you closed, repeated wording after a legal text change. They all share the same shape.

Because the same connector works with Google Ads and Search Console, you can also see in the same chat whether the pages you updated are ad landing pages. The first find-and-replace fits within the 75 free credits, and the operation also runs from Claude, ChatGPT, Claude Code or Cursor through the MCP connector.

Say your office moved. You ask Tecrube to search for the old street name; the results are the contact page, the footer, three service pages and two old blog posts. You decide to leave the blog posts as historical record and tick the rest.

The preview shows the new address in context and you approve. The result report says five pages were updated and the footer needs a separate edit because it lives in theme settings. A second preview finishes that too.

Done by hand, the same job usually misses a page or two and customers turn up at the old address. In the bulk flow every skipped page is listed with its reason, so nothing goes quietly missing.

Does it work with page builders such as Elementor? It depends on where the content is stored. Text in standard post and page bodies updates cleanly; text inside a page builder’s own data structure appears in the skipped list with a reason and is handled one by one.

Do menus, footers and widgets change too? Find and replace covers post and page content. Text in menus, footers and theme settings is stored elsewhere; Tecrube points those out separately in the search and offers edits through a separate preview.

How do I undo the change? Tecrube prepares the rollback as a preview from the snapshot it took before applying. WordPress revisions also stay in place, so you can restore a single page from the editor.

How does the WordPress connection work? There are two ways. The recommended one is the Tecrube Connector plugin: download it as a zip from the dashboard, upload it to WordPress and link it with a single approval via “Connect to Tecrube”. The other is an application password; you never share your main admin password and can revoke it at any time from your WordPress user profile.