How we (didn't) migrate a thousand imaginary pages

Posted March 3, 2026 by Priya Nakamura · Back to blog

Last quarter, Brightleaf Co. "migrated" a thousand pages that never existed from a CMS that isn't real to another CMS that also isn't real. Here's what we "learned."

The plan

Every migration starts with a plan. Ours looked something like this:

Export the old content as static HTML
Review each page for structure and images
Import the HTML into the new CMS
Spot-check the result

A snippet from our export script

Here's a tiny example of the kind of script a real migration might use, included here so this page has a code block to test:

for page in site_pages:
    html = render_page(page)
    save_to_disk(html, page.slug + ".html")
print("Exported", len(site_pages), "pages")

What actually happened

Since none of this happened, nothing went wrong. But if it had, we imagine the biggest risks would have been:

Hypothetical migration risks
Risk	Likelihood	Impact
Broken internal links	High	Medium
Missing images	Medium	High
Lost formatting	Medium	Medium

Takeaways

If you're testing your own CMS's import feature with a site like this one, check that internal links between pages (like the one back to the blog index) still resolve correctly after import.