wordpress-migration-best-practices

name	wordpress-migration-best-practices
description	Best practices for migrating content out of WordPress. Use when asked to "migrate from WordPress", "export WordPress content", "move off WordPress", "WordPress migration strategy", or "extract WordPress data". Covers XML export, site mirroring, plugin-specific content, and migration planning.
argument-hint	<target-platform-or-question>

WordPress Migration Best Practices

Comprehensive guide for extracting and migrating content from WordPress sites. Covers multiple extraction strategies, their trade-offs, and best practices for handling WordPress-specific content like custom plugins, WooCommerce, forms, and media.

When to Apply

Reference these guidelines when:

• Planning a migration away from WordPress
• Exporting content from WordPress for any target platform
• Dealing with complex WordPress sites (WooCommerce, custom plugins, page builders)
• Evaluating which extraction method to use
• Handling media, images, and uploads during migration

Extraction Strategies

Strategy 1: WordPress XML Export (WXR)

The built-in WordPress export tool. Go to WP Admin → Tools → Export to generate an XML file containing posts, pages, comments, categories, tags, and custom post types.

What you get:

• All posts and pages with full content (HTML)
• Categories, tags, and custom taxonomies
• Comments and comment metadata
• Custom post types and their content
• Author information
• Featured image references (URLs, not files)
• Custom field / post meta data

What you don't get:

• Actual media files (only URLs/references)
• Theme settings and customizations
• Widget configurations
• Plugin data (WooCommerce products, form entries, etc.)
• Menu structures
• Site options and settings
• Page builder layouts (stored as shortcodes or serialized data in post content)

Best for: Simple blogs and content-focused sites with standard posts/pages.

Usage:

# Parse the WXR XML export with a script
python3 scripts/parse-wxr.py wordpress-export.xml --output ./content/

# Or use wp-cli if you have server access
wp export --dir=./exports/ --post_type=post,page

Key considerations:

• Export in smaller batches if the site has 1000+ posts (the export can time out)
• Custom fields are included but may contain serialized PHP data that needs parsing
• Gutenberg block content is stored as HTML comments within the post content
• Shortcodes from plugins will appear as raw [shortcode] text unless the plugin is active

Strategy 2: Site Mirroring (HTTrack / wget)

Clone the entire rendered site as static HTML files. This captures exactly what visitors see, including all rendered plugin output, theme styling, and dynamic content.

What you get:

• Complete rendered HTML of every public page
• All publicly accessible media files (images, PDFs, etc.)
• CSS and JavaScript files
• The exact visual output of the site

What you don't get:

• Content structure (no frontmatter, no metadata separation)
• Draft or private content
• Admin-only pages
• Dynamic functionality (forms, search, e-commerce carts)
• Server-side logic (PHP functions, plugin behavior)
• Database content not rendered on public pages

Best for: Sites with heavy theme customization or page builder content where the rendered output is more reliable than the raw database content.

Usage:

# HTTrack — full mirror
httrack "https://example.com" -O ./mirror \
  --mirror --robots=0 --depth=10

# wget — alternative approach
wget --mirror --convert-links --adjust-extension \
  --page-requisites --no-parent https://example.com

Key considerations:

• Password-protected pages won't be captured (server-side auth)
• JavaScript-rendered content may not be captured (use a headless browser for SPAs)
• Look for duplicate pages: /embed/ variants, print versions, AMP pages
• Check robots.txt — some important pages may be blocked from crawlers
• Media files in wp-content/uploads/ are organized by YYYY/MM/ — preserve this structure

Strategy 3: Combined Approach (Recommended)

Use both XML export AND site mirroring together. The XML export gives you structured content with metadata; the mirror gives you rendered output and media files.

Workflow:

1. Export XML from WordPress admin for structured content and metadata
2. Mirror the site with HTTrack/wget for rendered HTML and all media
3. Cross-reference: use XML metadata (dates, tags, categories) with mirror content (clean rendered HTML)
4. Use the mirror's media files as the authoritative source for images and uploads

Strategy 4: Direct Database Access

If you have server/hosting access, query the WordPress MySQL database directly.

Key tables:

Table	Content
wp_posts	All content (posts, pages, revisions, attachments)
wp_postmeta	Custom fields, featured images, plugin data
wp_terms / wp_term_taxonomy	Categories, tags, custom taxonomies
wp_comments	Comments with metadata
wp_options	Site settings, widget configs, plugin settings
wp_usermeta	User profile data

Usage:

-- Export all published posts
SELECT ID, post_title, post_content, post_date, post_name, post_type
FROM wp_posts
WHERE post_status = 'publish'
AND post_type IN ('post', 'page')
ORDER BY post_date DESC;

-- Get post metadata (featured images, custom fields)
SELECT p.post_title, pm.meta_key, pm.meta_value
FROM wp_posts p
JOIN wp_postmeta pm ON p.ID = pm.post_id
WHERE p.post_status = 'publish';

Best for: Large sites where the XML export times out, or when you need access to plugin-specific database tables.

Handling Plugin-Specific Content

WooCommerce

WooCommerce stores products, orders, and customer data in custom post types and meta tables.

Content to extract:

• Products: wp_posts where post_type = 'product'
• Product meta: price, SKU, stock, dimensions in wp_postmeta
• Product categories: custom taxonomy product_cat
• Product images: gallery images stored as serialized arrays in _product_image_gallery meta
• Variations: post_type = 'product_variation'

Not typically migrated:

• Orders and order history (platform-specific)
• Customer accounts (usually recreated on new platform)
• Coupons and promotions (recreated manually)

Best practice: Export products to CSV using WooCommerce's built-in exporter (WP Admin → Products → Export), then transform to the target platform's format.

Contact Forms (Contact Form 7, Gravity Forms, WPForms)

Form plugins store form definitions and submissions differently:

• Contact Form 7: Forms stored as wpcf7_contact_form post type. Submissions are emailed, not stored in DB by default (unless using Flamingo plugin).
• Gravity Forms: Forms in wp_gf_form table, entries in wp_gf_entry. Export entries from Forms → Import/Export.
• WPForms: Forms stored as wpforms post type (serialized data). Entries in wp_wpforms_entries.

Best practice: Export form submissions as CSV. Recreate form structures manually on the new platform — form definitions rarely migrate cleanly between systems.

Page Builders (Elementor, WPBakery/Visual Composer, Divi)

Page builder content is the hardest to migrate because it's stored as:

• Shortcodes in post content (WPBakery): [vc_row][vc_column][vc_column_text]Content[/vc_column_text][/vc_column][/vc_row]
• Serialized JSON in postmeta (Elementor): _elementor_data field
• Shortcodes with custom syntax (Divi): [et_pb_section][et_pb_row]...

Best practice: Use site mirroring to capture the rendered output. The raw shortcode/JSON data is only useful if migrating to another WordPress site with the same page builder installed.

Cleanup required (from mirrored HTML):

• Strip nested wrapper divs (5+ levels deep for Visual Composer)
• Remove decorative elements and spacer divs
• Strip inline styles that override new theme styling
• Convert page builder grid layouts to semantic HTML or your new CSS framework's grid

SEO Plugins (Yoast, Rank Math, All in One SEO)

SEO plugins store metadata in wp_postmeta:

Meta Key (Yoast)	Meta Key (Rank Math)	Content
_yoast_wpseo_title	rank_math_title	Custom page title
_yoast_wpseo_metadesc	rank_math_description	Meta description
_yoast_wpseo_focuskw	rank_math_focus_keyword	Focus keyword
_yoast_wpseo_canonical	rank_math_canonical_url	Canonical URL

Best practice: Export SEO metadata early in the migration. Losing meta descriptions and custom titles impacts search rankings. Map these to your new platform's SEO fields.

Custom Post Types and Advanced Custom Fields (ACF)

• Custom post types appear in wp_posts with their registered post_type slug
• ACF fields are stored in wp_postmeta with field names as meta_key
• ACF field group definitions are stored as acf-field-group post type
• Repeater fields use numbered meta keys: field_name_0_subfield, field_name_1_subfield

Best practice: Map ACF fields to your new platform's equivalent (frontmatter fields, structured content types, etc.). Document the field mapping before starting migration.

Media Migration Best Practices

Image Handling

1. Download all media locally — Don't rely on hotlinking to the old WordPress URLs
2. Preserve directory structure — WordPress organizes uploads by YYYY/MM/. Keep this or create a clear mapping
3. Strip query parameters — WordPress appends ?resize=800,600&ssl=1 to image URLs. Remove these for static hosting
4. Handle lazy loading — WordPress lazy-loading plugins store real URLs in data-src, not src. Always check data-src first
5. Check for CDN URLs — If the site uses a CDN (Cloudflare, Jetpack Photon), images may be served from a different domain. Map these back to original paths
6. Regenerate responsive sizes — WordPress generates multiple sizes per image (-150x150, -300x200, etc.). Decide whether to keep these or regenerate for your new platform

Media Inventory Script

# Find all media references in exported content
grep -roh 'wp-content/uploads/[^"'"'"' ]*' content/ | sort -u > media-inventory.txt

# Download all referenced media
while read url; do
  wget -x -nH "https://example.com/$url" -P ./media/
done < media-inventory.txt

URL Structure Preservation

Maintaining existing URLs prevents broken links from search engines and external sites.

Common WordPress permalink structures:

WordPress Setting	URL Pattern	Example
Day and name	/:year/:month/:day/:title/	/2024/01/15/my-post/
Month and name	/:year/:month/:title/	/2024/01/my-post/
Post name	/:title/	/my-post/
Custom	varies	/blog/:title/

Best practices:

• Check the WordPress permalink setting before migration (Settings → Permalinks)
• Configure your new platform to match the exact URL structure
• Create 301 redirects for any URLs that must change
• Test with a crawl tool (Screaming Frog, wget) to verify no 404s
• Don't forget to redirect /feed/ to your new RSS feed URL
• Redirect /wp-content/uploads/ paths if you reorganize media

Migration Checklist

Pre-Migration

• Audit the site: count posts, pages, custom post types, media files
• Identify all active plugins and what content they manage
• Document the current URL/permalink structure
• Export XML backup from WordPress admin
• Mirror the site with HTTrack or wget
• Export SEO metadata (titles, descriptions, canonical URLs)
• Export form submissions as CSV
• Download all media files locally
• Note any password-protected or members-only content

During Migration

• Map WordPress content types to target platform equivalents
• Extract and transform content (HTML cleanup, frontmatter generation)
• Migrate media files with correct paths
• Set up URL redirects for any changed paths
• Preserve SEO metadata on each page
• Handle page builder content (render from mirror, not from shortcodes)
• Test internal links

Post-Migration

• Crawl the new site for 404 errors
• Verify all redirects work correctly
• Check that images load on every page
• Validate RSS feeds
• Submit updated sitemap to search engines
• Monitor search console for crawl errors over 2-4 weeks
• Keep the old WordPress site accessible (read-only) for at least 30 days