Skip to the indexed answers
Digital Marketing Index

usdigitalmarketingindex.com SEO

Which SEO indicators should be reviewed during the first month after a CMS migration?

Five groups, each with an owner and a rule for when to act: whether pages can be crawled and are being indexed, whether redirects and status codes behave as mapped, whether the new templates still emit the right tags and content, how clicks, impressions, and queries compare with a recorded baseline, and how conversions from organic search compare. The first month has a particular difficulty. Some movement is expected while search engines reprocess the site, and some movement is a defect. Google’s site move guidance says to expect temporary fluctuation, and that for medium-sized sites it can take a few weeks or more for new URLs to replace old ones (Google’s guidance on site moves with URL changes). The monitoring plan exists to tell those apart: defects show up as specific failures on specific URLs, and they can be found and fixed while the transition runs its course.

Daily crawl and index checks

In the first week, check daily. Technical faults introduced at launch do the most damage the longer they stand.

Indicator Source Looking for
robots.txt Fetch the live file Not the staging version; no site-wide disallow
<!–INLINECODE0–> Crawl of key templates; URL Inspection No staging <!–INLINECODE1–> left on live pages
Server errors Server logs; crawl stats report No rise in 5xx
Response time Monitoring; crawl stats Stable
Crawl requests Crawl stats report Activity present; the guidance notes Google may crawl a moved site more heavily for a while
Pages indexed Page indexing report New URLs rising; old URLs moving to redirected
Reasons for not indexed Page indexing report No growth in blocked, <!–INLINECODE2–>, soft 404, or redirect error
Sitemap status Sitemaps report Read successfully; URLs discovered
Rendering URL Inspection, live test, on one URL per template Content and links present in rendered HTML
Canonical selected URL Inspection Google's choice equals the declared canonical

Search Console reports are not real time. Server logs and direct requests show the present state, so use them for daily checks and the reports for trends.

Traffic and query baselines

A comparison needs a baseline, recorded before the migration.

Baseline to hold:

Data Period Granularity
Clicks, impressions, CTR, average position 16 months if available, at least the same weeks of the previous year By page, by query, by device, by country
Organic sessions and conversions Same By landing page
Indexed page count At launch By section
Crawl of the old site Before launch Every URL with title, headings, canonical, status, inlinks
Top pages and top queries lists Before launch The priority set to watch
Rich result reports Before launch Valid item counts by type

Reading traffic after launch:

  • Compare with the same weekdays, and with the same period last year, to separate the migration from seasonality.
  • Where URLs changed, sum old and new URLs for each page, using the redirect map. Looking at either alone shows a fall or a rise that is only the handover.
  • Segment by template or section. A sitewide average hides a single broken template.
  • Watch the priority pages individually.
  • Separate branded from non-branded queries. Branded queries can hold steady while everything else falls.
  • Check query coverage per page: has a page stopped appearing for queries it used to appear for?

Errors and redirects

Indicator Source Looking for
Redirect test Scripted requests for every URL in the redirect map One hop, correct target, final 200; run daily in week one
404s Server logs, grouped by URL and referrer Old URLs missing from the map; broken internal links
Soft 404s Page indexing report Redirects to irrelevant pages; empty templates
Redirect chains and loops Crawl None
Internal links to redirected URLs Crawl Declining to zero
Canonical tags Crawl New URLs; no staging host
Titles and headings Crawl compared with the old crawl No unintended mass changes
Structured data Rich result reports; Rich Results Test on templates Valid item counts steady
hreflang Crawl Reciprocal; new URLs
Images Crawl No broken images; alt text present
Mixed content and certificates Crawl; browser console None
Page experience measurements Field data reports No regression as data accumulates

The log of 404s repays reading in the first days. Each frequently requested missing URL is an old address someone forgot, and adding it to the map fixes it.

Escalation triggers

A four-week monitoring schedule:

Week Frequency Owner Observation Intervention rule
1 Daily Developer Status codes, robots.txt, <!–INLINECODE3–>, server errors, redirect test Any site-wide block or failing redirect rule: fix the same day
1 Daily SEO lead 404 log; URL Inspection on each template Missing mapped URLs added within a day; rendering fault raised at once
1 Twice SEO lead Page indexing and sitemap reports Rising exclusion reason investigated
2 Every two days Developer Server errors, response time, redirect test As week 1
2 Twice SEO lead Full crawl compared with baseline: canonicals, titles, links, structured data Template defect logged and scheduled within the week
2 Weekly Analyst Clicks and conversions by section against baseline Section falling well beyond the site average: page-level review
3 Twice SEO lead Indexing progress; old URLs still indexed; rich result reports Stalled indexing of a section: check internal links, sitemap, and rendering for it
3 Weekly Analyst Priority pages and queries A priority page absent for its main queries: compare its content and links with the old version
4 Weekly SEO lead Full crawl; redirect test; 404 log Residual issues closed or accepted in writing
4 Once All owners Month review against baseline Decide what continues into ongoing monitoring

Escalation triggers, defined before launch so that nobody has to argue thresholds during an incident:

Trigger Class Action
robots.txt blocking the site, or <!–INLINECODE4–> on live templates Regression Immediate fix; highest priority
Sustained 5xx errors Regression Immediate; infrastructure owner
Redirects failing or pointing to wrong targets Regression Same day
A template rendering without its main content or links Regression Same day
Canonicals naming old or staging URLs Regression Within days
Structured data items dropping to zero for a type Regression Within days
Old URLs gradually leaving the index while new ones enter Expected Monitor
Day-to-day movement in positions Expected Monitor
Combined old-plus-new clicks somewhat below baseline in the first weeks, with indexing progressing Expected Monitor; review weekly
Combined clicks falling steadily with indexing complete Needs investigation Page-level comparison with baseline
One section far worse than the rest Needs investigation Template and content comparison for that section

How to tell a transition from a regression:

Sign Transition Regression
Scope Across the site, proportionate Concentrated in a template, section, or URL pattern
Technical checks All pass A check fails
Indexing Progressing Stalled or reversing
Direction Stabilizing Worsening
Cause None found after checking A specific fault identified

The thresholds for “well beyond” and “somewhat below” should be set by the team against the site’s own normal variation, taken from the baseline. This guide gives no universal percentage, because none is documented.

Keep a dated log of every finding and fix. At the month’s end it is the record of what the migration cost, what was repaired, and what remains.

Most of what this schedule checks should have been specified before the build began. What belongs in that specification is covered in the guide to which SEO requirements belong in the brief for a planned website redesign.

The order of work: record the baseline, check technical health daily in the first week, compare traffic with old and new URLs combined, and act at once on anything that is a failed check and patiently on anything that is only movement.

Link trail

Reader comments on this answer

Add your comment to this answer

Your email address will not be published. Required fields are marked *