Search Console after a redesign: what to check in the first 30 days
Written by AERVIO, a Montreal studio. Human-reviewed before publish. Contact: hello@aervio.io · +1 438-498-3913.
A redesign is the one project where search visibility can fall without anyone touching a keyword. URLs change, titles get rewritten by a designer who wanted shorter, and a navigation that linked to twelve service pages now links to four. Search Console is where Google tells you what it made of that.
This is the order we check things in during the first 30 days after a relaunch. It is the same list whether the site runs on WordPress or a custom stack. If you want the stack conversation first, read Next.js or WordPress for Quebec teams.
Before launch: export, because the history is the baseline
Search Console keeps roughly 16 months of performance data. That is generous until you need to compare against it. Before launch day, export:
- Performance by page and by query, last 16 months if available
- The Page indexing report (the list of indexed and not-indexed URLs)
- Your current sitemap URL list
- The top 50 landing pages from your analytics
Without this, "we lost traffic" is a feeling. With it, you can say which pages lost impressions, which gained, and which no longer exist. A screenshot of a line chart is not a baseline.
Day 1: confirm the property and the sitemap
Verify that the property matches the real public domain, including the right protocol and the `www` choice. A domain property covers every variant; a URL-prefix property covers only one. Many "we vanished" scares are a team looking at the wrong property.
Then submit the new sitemap and check three things:
- It lists only canonical, indexable URLs that return 200
- It does not list staging hostnames, parameter variants, or the old URLs
- It agrees with `robots.txt` (which should not block anything the sitemap lists)
If the sitemap says "index this" and the page says `noindex`, Google will believe the page. We have seen a staging `noindex` ship to production more than once. It is the first thing we check, not the last.
Days 1–3: inspect the pages that earn money
Use URL Inspection on ten pages, not two hundred: the homepage, each service page, the contact page, and the top three pages by past clicks. For each, confirm:
- The URL is the one you expect (canonical matches)
- The page is indexable
- The rendered HTML contains the text, not an empty shell waiting for JavaScript
"Request indexing" is available for individual URLs. Use it on the handful that matter. It is not a lever for the whole site, and repeating it does not make Google faster.
Week 1: redirects are the project
Every old URL that earned a click or a link should return a single 301 to the closest equivalent new page. "Closest equivalent" is a decision, not a regex. Sending 80 old pages to the homepage tells Google those pages are gone, and in practice it often treats the result as a soft 404.
Check for:
- Redirect chains (old to temp to new) — collapse them to one hop
- Redirect loops, usually from a trailing-slash rule fighting a plugin
- HTTP-to-HTTPS and `www` rules that add a second hop
- Old image and PDF URLs that still receive links
Pull the old URL list from your export and test it. A spreadsheet with three columns — old URL, new URL, status — is the most useful artefact of the whole relaunch.
Weeks 1–2: read the Page indexing report like a diagnosis
Not every "not indexed" row is a problem. The useful skill is telling them apart:
| Status you will see | Usually means | What we do |
|---|---|---|
| Not found (404) | Old URL with no redirect | Add a 301 if the page earned traffic or links; otherwise leave it |
| Page with redirect | Old URL redirecting correctly | Nothing. This is the redirect working |
| Crawled, currently not indexed | Google saw it and was not convinced | Improve the page or merge it into a stronger one |
| Discovered, currently not indexed | Queued, not yet crawled | Wait; check internal links to it |
| Duplicate without user-selected canonical | Two URLs, same content | Set an explicit canonical and fix the internal links |
| Blocked by robots.txt | A rule is in the way | Confirm that is intentional |
Notice what is not on the list: "number of indexed pages went down." After a redesign that merges thin pages, a lower count can be a good result. Count the pages that matter, not the pages that exist.
Weeks 2–4: compare queries, not totals
Total clicks hide the story. Compare the same date range before and after for:
- Branded queries (your name). These should barely move. If they do, something is wrong with titles or indexing of the homepage.
- Non-branded queries tied to money pages. This is where a redesign helps or hurts.
- Queries with many impressions and few clicks. After a redesign this often means a title or description now misses the intent.
Averages in Search Console can mislead when a page gains many long-tail impressions at low positions. Look at the page, then the queries for that page, then the position trend.
If a page lost ground, ask what changed on it. Common culprits: the heading that carried the target phrase became an image, the body copy was cut for "cleaner design", the page lost its internal links, or two pages now compete for the same query.
Core Web Vitals and mobile: watch, do not obsess
Search Console reports Core Web Vitals from field data, which lags. A new template might not show up for weeks. Use it as a trend, and test the money pages directly with PageSpeed Insights or lab tools. Google documents the metrics at web.dev. Fix real regressions on pages that sell; do not chase a green number on the privacy policy.
What we ignore in the first month
- Daily fluctuations. Ignore anything shorter than a week.
- Impressions from queries you never targeted and will never convert
- A handful of 404s from bots probing for `/wp-login.php` on a site that is not WordPress
- Sitemaps showing "Success" with fewer discovered URLs than you submitted, until you have checked whether the missing ones were duplicates
How this fits the rest of the system
Search Console tells you what Google saw. It does not tell you who called. We pair it with form-to-CRM source tracking so you can see which landing page started a conversation — that is the argument in site, CRM, and search as one system. And the pages themselves still need claims a person or an answer engine can quote; that part is in SEO and GEO in Montreal without a fake map pin.
What we will not do
We will not promise your traffic will hold, because it depends on competitors and on Google. We will not buy links, rent lists, or invent a map pin to make a chart look better. We will give you the before and after export, the redirect map, and a short written read at day 30 saying what moved, why we think so, and what we do next.
If you are planning a redesign or are already a few weeks past one and nervous, write to hello@aervio.io with the domain and the date you launched. We are in Montreal, weekdays 9–17 ET, and the line is +1 438-498-3913. We will tell you whether the problem is in Search Console, in the redirects, or in the pages themselves.
Questions
How long until Search Console settles after a redesign?
Expect noise for several weeks. Google recrawls important pages first and the long tail later. If the money pages are indexed, redirected correctly, and queries are stable after about a month, you are normally fine. If they are not, the cause is usually a specific, fixable thing.
Do I need the Change of Address tool?
Only if the domain itself changes, for example old-name.ca to new-name.ca. It does not apply to a redesign on the same domain. In that case your 301 redirects and sitemap do the work.
Can you promise our traffic will not drop?
No. Nobody honest can. We can promise a redirect map, a before and after export, and a clear list of what changed, so a drop has a diagnosis instead of a theory.
Read next
Let's write the next step.