Blog / Case studies /

A review of a content migration after the first month

The migration went live a few weeks ago. Somebody checked the homepage, somebody watched a few charts, and everyone relaxed.

But has anyone compared the old content inventory with the live site? Probably not.

A first-month review catches missed pages, broken journeys and implementation gaps before the new structure becomes “just how it is”. It can’t tell you the long-term SEO result, though. Search engines recrawl at different speeds, demand moves and outside factors play in. So first confirm the migration did what it was meant to, then list what needs more time or evidence.

Revisit the migration’s goals

Open the original plan and list the intended outcomes: URL changes, content consolidation, redesign, CMS changes, redirects, tracking updates, technical improvements. Mark what was delivered and what was deferred. Without that reference, the review drifts into a general chat about traffic instead of checking the work that actually happened.

Separate implementation goals from performance goals. “Redirect every important old URL to a relevant destination” can be checked directly. “Increase qualified organic demand” needs a baseline, an agreed definition and more time. Month one may reveal problems, but it rarely gives enough evidence to credit a business result.

Record changes that happened after launch: emergency fixes, new campaigns, product availability, tracking changes or content edits. These affect interpretation. If a report shows a decline, note other plausible causes before linking it to the migration.

Check the routes people and crawlers use

Crawl a representative set of old and new URLs. Verify that old routes redirect directly to the intended destination, that the new page returns the expected status and that internal links now point to the final URL. Look for chains, loops, irrelevant homepage redirects, 404s and pages that have become orphaned.

Prioritise URLs with past visits, external links, revenue relevance or strong internal importance. The old inventory is probably incomplete, so compare it with analytics, Search Console, backlinks, server logs and the current sitemap. Each source has limits, so note where each URL came from.

Inspect the new pages directly: titles, headings, canonical tags, robots directives, language relationships, structured data and visible content. A redirect can work perfectly while the destination has missing content or the wrong indexation setting. I check templates as well as single examples, because that’s how you spot patterns.

Compare content, not just addresses

For migrated pages, compare the old purpose with the new page. Was a useful section removed? Did several pages merge into one, and does the destination still answer the original questions? A URL map can be technically complete but semantically poor if a specific resource points to a generic category.

Look for missing images, broken embedded content, lost metadata, stale dates, duplicated titles and changes in calls to action. Review content that was intentionally consolidated to ensure that the important information survived in a more coherent form. Do not restore every old paragraph simply because it existed; decide what readers still need.

Ask content owners whether the new structure matches how people use the site. A migration is not successful merely because a crawler sees valid HTML. The navigation, filters, search and conversion paths must also work for real visitors.

Read the data with caution

Compare Search Console clicks and impressions for matched page groups and query themes, over equivalent date ranges where possible. Search Console isn’t a perfect migration tracker, and recent data may be incomplete. Average position moves when the query mix moves, so look at examples rather than one aggregate number.

If tracking changed, check GA4 key events and landing-page data, and verify events fire before comparing outcomes. A missing event may be a measurement problem, not a business decline. And a traffic rise doesn’t prove the migrated experience helps customers if the relevant actions didn’t improve.

Segment by market, device, page type and key path only when the data is sufficient and the segments answer a useful question. Too many slices can turn noise into apparent patterns. State the limits and avoid false precision.

Decide what to fix now

Sort findings by impact, confidence and effort. Fix technical errors that block access, break important routes or misrepresent the page. Investigate data discrepancies that could change a decision. Schedule minor design refinements separately if they do not affect the core journey.

Assign owners and due dates based on actual capacity. I’d rather hand over a short, ranked list than a hundred unprioritised findings, which isn’t a recovery plan. Make clear which issue can be corrected now, which needs more evidence and which is an accepted trade-off from the migration.

After a fix, verify the affected routes and record the change date. Avoid changing content, redirects and metadata simultaneously if the team needs to understand what resolved a problem. Keep a rollback option for changes to routing or templates.

Keep watching the right signals

Set a follow-up review after the first month, with enough time for recrawling and normal operating conditions. Monitor error patterns, indexation of important pages, landing-page behaviour and support feedback. If a major change or release happens, record it so the trend can be interpreted.

Do not wait for a scheduled review if customers cannot complete a key task or important pages are inaccessible. Escalate material problems as they appear. The cadence is for structured learning, not a reason to delay urgent fixes.

Treat the month-one review as a checkpoint

The first month is good for checking implementation and opening a list of questions. It can’t settle every long-term outcome. Separate what you can verify now from what needs more time, and give each open question an owner.

So go back to the migration plan and pick three important page groups. Can you trace each from old URL to new page, confirm the content survived and explain what the performance data can and can’t tell you? Start with any gap in that chain.