Found
How to Migrate Website Content, Images and Metadata Without Losing What Already Works
Move pages with their images, downloads, metadata and internal links—then check what arrived before launch.

Milan Gohil
·
8 min read
OVERVIEW
Here is a short version of what I write.
To migrate website content reliably, inventory everything attached to each page and compare the old and new versions before launch. Include assets and search signals as well as the written content.
Map pages, images, files and their destinations.
Carry over relevant titles, descriptions, alt text and structured data.
Check internal links, canonicals and redirects.
Compare old and new pages and resolve missing items before launch.
When a website is rebuilt, the obvious job is moving the pages. The less obvious job is making sure everything attached to those pages comes with them. That can include:
copy
URLs
images
PDFs
downloads
page titles
meta descriptions
canonicals
structured data
alt text
internal links
author/date information
redirects
A new website can look completely correct while parts of this quietly disappear during migration. Before launch, you need to know what is moving, what is changing, and what should not be lost.

If the migration checks only:
“Is the text on the new page?”
it is incomplete. Google's current migration guidance explicitly recommends transferring hosted images and downloads such as PDFs and including embedded-content URLs in the migration plan.
Want AI to compare the old and new website first?
If you have a staging site, this is a useful first check.
I am migrating this website:
CURRENT WEBSITE:
[OLD WEBSITE URL]NEW / STAGING WEBSITE:
[NEW WEBSITE URL]Act as a website migration QA reviewer.
Compare the old and new websites.
Do not judge the visual design.
For important pages, compare:
1. URL
2. Main page content
3. H1 and important headings
4. Page title
5. Meta description
6. Canonical URL
7. Internal links
8. Images
9. Image alt text
10. PDFs/downloads
11. Structured data where visible
12. Author/date information where relevantClassify each item as:
PRESERVED
CHANGED INTENTIONALLY
MISSING
BROKEN
NEEDS MANUAL REVIEWAlso identify:
- old pages with no obvious new destination
- images/files that appear missing
- links still pointing to old URLs
- metadata that disappeared
- important content substantially reduced
- redirects that should be testedDo not invent Search Console traffic, backlinks or conversions.
Finish with:
- highest-risk migration issues
- pages I should manually check first
- files/assets likely to be forgotten
- pre-launch fixes
- post-launch checks
AI can make the comparison faster. The final migration still needs a proper crawl and manual checks.
Start with one migration inventory
Do not keep one spreadsheet for content, another random list for redirects and a third list of missing PDFs. Create one migration map.

Old URL | New URL | Content | Media | Metadata | Redirect | Status |
|---|---|---|---|---|---|---|
/payroll | /payroll | Keep/update | Move | Preserve/review | — | Ready |
/vat-help | /vat-services | Merge | Move | Rewrite | 301 | Review |
/old-offer | — | Remove | — | — | 404/410 | Ready |
/guide.pdf | /files/guide.pdf | — | Move | — | If URL changes | Test |
For every important page, there should be a clear outcome.
1. Move the useful content—not automatically every old sentence
Content migration does not mean:
Copy old website → paste into new template.
Some pages should stay almost unchanged. Some should be improved. Some should merge. Some should disappear. The content inventory you created before redesign should already classify them. The migration stage simply makes sure those decisions are actually implemented. For example:
OLD
/corporate-tax
/tax-for-business
/business-tax-advice
NEW
/business-tax-advisory
If those pages were intentionally consolidated, the new page should contain the useful information from the old set—not just a short replacement page plus three redirects. Google allows several old pages to redirect to one new destination when their content has genuinely been consolidated there.
2. Check images separately
Images are very easy to lose during a CMS migration. Imagine a manufacturing site has:
product diagrams
installation photography
technical drawings
application examples
The new site may display replacement imagery correctly while the old files still have:
external links
Google Image visibility
references inside old articles
useful filenames/alt text
Google specifically recommends including image and video URLs in a site-move plan. For important images, check: Google: site moves with URL changes.
□ Image migrated
□ New page displays it
□ File isn't broken
□ Alt text still makes sense
□ Old image URL redirects if it changed and needs preserving
□ Internal references use the new file URL
Do not migrate every decorative image simply because it exists. Focus on assets that still serve a purpose.
3. Do not forget PDFs and downloads
This can matter more than many redesign teams expect. Think about a company with:
product specifications
brochures
certificates
menus
application forms
technical manuals
research reports
Customers may have saved those URLs. Other websites may link directly to them. If:
/files/technical-manual-v4.pdf
disappears during the redesign, the new homepage looking better does not help the person following that old link. Google's migration documentation specifically calls out downloads such as PDF documents as content that should be transferred or given an appropriate new location. For every important file decide: Keep same URL Move + redirect Replace with newer file Convert to webpage Remove because it is obsolete
4. Compare page titles and descriptions before and after
A migration can unintentionally replace carefully written page metadata with CMS defaults. For example:
Before
Industrial Pump Repair & Maintenance | ABC Engineering
After
Pump Services
or even:
Untitled Page
That may happen because the new template was never populated correctly. So compare:
page title
meta description
H1
for important pages. Do not blindly preserve everything either. If the old metadata was weak or the page intent changed, improve it intentionally. The key is:
Changed because we decided to change it
rather than:
Changed because nobody noticed it disappeared.
5. Check canonicals when URLs or CMS change
A canonical tells search engines which URL you prefer as the main version when duplicate or similar versions exist. During migration, mistakes can happen such as:
NEW PAGE
example.com/new-service
canonical →
staging.example.com/new-service
or:
canonical →
oldsite.com/service
Google recommends that new URLs use self-referencing canonicals after a site move and that URL-related annotations be updated to the new locations. That is not something a business owner needs to configure personally. It is something worth having on the migration checklist.
6. Structured data can disappear when templates change
Structured data often lives inside the website template rather than in visible copy. A rebuild can therefore remove it without the page looking different. Depending on the website, that might include data for:
products
articles
breadcrumbs
organizations
local businesses
job postings
events
Google uses structured data to better understand page content and, for supported types, determine eligibility for richer search appearances. So if the old site has useful valid structured data: compare it with the new implementation. Do not assume the new CMS recreated it automatically.
7. Update internal links to the new URLs
Suppose:
Old: /services/maintenance
New: /industrial-maintenance
The redirect may work correctly. But the new website should not keep linking internally to:
/services/maintenance
and rely on the redirect every time. Google recommends updating internal links to point directly to new URLs after a migration. So crawl the new website and look for:
links to old URLs
broken links
redirect chains
links to staging URLs
missing contextual links

Google recommends server-side permanent redirects such as 301 or 308 for permanent moves and warns against redirecting unrelated old pages to an irrelevant destination such as the homepage.
8. Check dates and authors when migrating articles
Blog migrations often preserve the body text but lose:
original publication date
updated date
author
categories
featured image
related internal links
That can create a strange result:
A six-year-old article suddenly appears to have been “published yesterday.”
If the information is genuinely being updated, show that properly. If it is simply being migrated, do not manufacture a new publication history. For article structured data, Google supports properties such as datePublished, dateModified, author, headline and images. Google: article structured data.
9. Check the new website before switching it live
Do not wait until after launch to discover the missing pieces. Take the most important pages and compare:

Check | Old | New | Correct? |
|---|---|---|---|
URL | — | — | — |
Main content | — | — | — |
Title | — | — | — |
H1 | — | — | — |
Meta description | — | — | — |
Canonical | — | — | — |
Images | — | — | — |
PDFs | — | — | — |
Internal links | — | — | — |
Schema | — | — | — |
CTA/form | — | — | — |
Start with:
highest organic-traffic pages
strongest lead/sales pages
pages with backlinks
important products/services
frequently used resources
You do not have to inspect 5,000 pages manually. Use a crawler for scale and manually inspect the pages where mistakes would matter most.
10. Make sure staging restrictions do not reach production
It is normal for a staging site to be hidden from search engines. What is not normal is launching the website with those settings still active. Google lists forgotten noindex directives and robots.txt blocks among common migration mistakes that can stop new pages from being indexed. Before launch check: Google: site moves with URL changes.
□ noindex removed where appropriate
□ robots.txt correct
□ canonicals point to production
□ staging URLs are not inside content
□ sitemap uses production URLs
Small details. Large possible consequences.
What should the business owner ask before approving migration?
You do not need to inspect the HTML yourself. Ask whoever is handling the migration:
Have all existing URLs been inventoried?
Which pages are being removed or merged?
Are important images and PDFs included?
Are titles, descriptions and canonicals being checked?
Are internal links being updated?
Is existing structured data being preserved or rebuilt where appropriate?
Have redirects been tested?
Has the new website been crawled before launch?
Have staging/noindex settings been checked?
Who checks Search Console after launch?
If the answer is:
“We are just copying the content into the new CMS.”
there is probably more to review.
Migration is complete when the important things still work
The goal is not to preserve every byte of the old website. The goal is to avoid losing useful things accidentally. Your migration can intentionally:
rewrite content
change images
merge pages
replace PDFs
improve metadata
restructure the website
But every meaningful change should have a reason. The simplest rule is:
Before Migration: Know what exists.
During Migration: Know what changes.
Before Launch: Compare old vs new.
After Launch: Check that users and search engines can still reach what matters.
That is much safer than discovering after launch that the website looked complete—but half of what made the old site useful never came with it.
Sources and further reading
All Blogs
All Blogs & Articles
© 2026 Webshunya Studio®. All rights reserved.







