A weekly Search Console routine that actually changes decisions
Most teams open Google Search Console, admire charts, and leave without changing a single template, title, or sitemap rule. A useful weekly routine is short, opinionated, and decision-oriented: every section ends with “do nothing,” “investigate,” or “ship a fix.” This checklist is built for that outcome—not for dashboard tourism.
1. Philosophy: decisions over dashboards
Search Console is a feedback loop. Impressions and clicks tell you how searchers meet your pages. Indexing reports tell you whether technical and quality systems are fighting your publishing plans. Enhancement reports tell you whether structured data still matches reality. None of that matters unless it changes backlog priority.
A healthy weekly habit produces at most a handful of tickets: rewrite three titles, remove filter URLs from the sitemap, fix a noindex accident on a section template, refresh one guide that sits on page two for a query you can win. If your weekly output is only screenshots, the routine is too vague.
Pair this checklist with the mental models in our practical SEO guide. When fixes involve robots, canonicals, or sitemaps, use the deeper playbooks on robots vs noindex vs canonical and sitemap include/exclude rules.
2. Before you start: windows, notes, owners
Use the same property each week ( Domains property preferred when you control DNS). Compare a recent 28-day window to the prior 28 days for direction, then drill into 7-day views only for anomalies. Record notes in a shared doc with four columns: observation, hypothesis, decision, owner/due date.
Define severity up front:
- Ship this week: accidental noindex on key templates, sitemap advertising error URLs, sharp non-brand click collapses tied to title changes you can reverse, security or manual action alerts.
- Plan this sprint: query clusters with strong impressions and weak CTR, thin pages accumulating “crawled not indexed,” schema invalidity on money templates.
- Monitor: small fluctuations, single-URL oddities without pattern, seasonal shifts you already expect.
Assign one owner for technical tickets and one for editorial tickets so the weekly review does not end as an orphaned Notion page.
3. The 45–60 minute weekly block
- 5 minutes — Alerts & manual actions: Security issues, manual actions, and email alerts first. Anything here preempts optimization theater.
- 10 minutes — Performance overview: Clicks, impressions, CTR, average position for 28 days vs prior 28. Note brand vs non-brand if you segment.
- 15 minutes — Queries & pages: Export or filter for opportunities and problems (details below).
- 15 minutes — Indexing & sitemaps: Look for new reason codes spiking; spot-check sitemap processing.
- 10 minutes — Experience / enhancements / inspections: Core Web Vitals trends if available, rich result errors, 2–3 URL inspections on changed templates.
- 5 minutes — Log decisions: Write the tickets. If nothing qualifies as ship/plan, explicitly write “no action—monitor X.”
Protect the time box. Deep forensic crawls belong outside the weekly block once a pattern is confirmed.
4. Performance: queries and pages that deserve action
Query views that force decisions
- High impressions, low CTR: Candidate for title/description tests. Draft alternatives with the title tag generator and meta description generator, then edit for honesty—do not promise what the page lacks.
- Strong position, weak clicks: Often a SERP feature or unattractive snippet. Inspect the live result intent match before rewriting the whole article.
- Rising impressions, flat clicks: New visibility; decide whether to improve the page now or wait for a fuller data week.
- Brand queries collapsing: Check site availability, homepage changes, and reputation issues before SEO tweaks.
Page views that force decisions
- Top pages losing clicks week over week: Diff recent deploys; confirm canonical and index status with inspection.
- New URLs with impressions but no clicks: Snippet and intent alignment first; content depth second.
- Cannibalizing URLs: Two pages trading queries—decide a primary, merge or retarget the secondary, align internal links and sitemap membership.
Do not chase average position as a vanity KPI. Position without clicks or conversions is a research clue, not a trophy. Use the keyword research tool and content brief generator when the decision is “create or rewrite,” not when the decision is “fix a template bug.”
5. Indexing: separate emergencies from noise
Indexing reports mix healthy exclusions with problems. Your weekly job is to detect change and pattern—not to “make every URL green.”
Treat as emergencies
- Valid pages suddenly marked excluded by noindex on templates that should rank.
- Sitemap spike of submitted URLs marked not found or redirect.
- Large drop in indexed pages correlated with a release.
Treat as project work
- Growth in “Crawled – currently not indexed” on thin taxonomies—improve or noindex/consolidate per policy.
- Duplicate without user-selected canonical—fix conflicting signals; see the canonical guidance in this directive guide.
- Soft 404 patterns on empty category or search URLs—fix status codes and sitemap inclusion (include/exclude article).
Treat as monitor
- Alternate page with proper canonical (when that is intended).
- Blocked by robots.txt on admin paths you deliberately Disallowed.
- Excluded by noindex on thank-you templates you deliberately marked.
When you change robots or sitemaps, validate files with the robots.txt generator, XML sitemap validator, and SEO audit checker before you assume Search Console will “just catch up.”
6. Experience and enhancements
Core Web Vitals (when field data exists) should influence engineering priority if money templates are Poor—especially mobile LCP on article or product templates. A weekly glance is enough to see trend direction; optimization work happens in engineering sprints with real measurements.
For enhancements, open the types you actually implemented (Article, FAQ, Product, Breadcrumbs). New invalid items after a release are actionable the same week. Philosophy for markup quality lives in schema: helpful vs spammy; use the schema markup generator only after you confirm visible parity.
URL Inspection is a scalpel: use it on the exact templates you changed, not on fifty random URLs. Request indexing sparingly for important new or fixed URLs after you verified the live HTML.
7. Worked examples of real decisions
Example A: CTR opportunity → title ship
Observation: Query “xml sitemap include exclude” shows solid impressions on your guide; CTR looks weak relative to neighboring queries on the same site.
Hypothesis: Title emphasizes brand voice but buries the include/exclude promise.
Decision: Ship a clearer title/description this week; annotate the date; review in 2–3 weeks. No full rewrite yet.
Example B: Indexing spike → sitemap cleanup
Observation: “Discovered – currently not indexed” jumps after an ecommerce release.
Hypothesis: New filter URLs entered the product sitemap.
Decision: Engineering removes facet URLs from sitemap generation; submit cleaned index; monitor count next week. Editorial does nothing.
Example C: Accidental noindex
Observation: Blog section URLs excluded by noindex; traffic down on blog.
Hypothesis: Staging robots meta leaked into production config.
Decision: Emergency fix same day; inspect sample URLs; request indexing for top posts after verification. Add a deploy checklist item.
Example D: Cannibalization
Observation: Two articles alternately appear for the same non-brand query.
Hypothesis: Overlapping intent; internal links point both ways.
Decision: Pick a primary; 301 or retarget the secondary; update internal links and sitemap; adjust titles so intents diverge if both must live.
Example E: Explicit no-action
Observation: Impressions dip mid-week then recover; no release correlation.
Decision: Monitor. Write it down so next week’s reviewer does not re-litigate the same blip.
8. Copy-paste weekly log template
Week of: YYYY-MM-DD
Property: example.com (Domain)
Window: last 28 days vs previous 28
Alerts / manual actions:
-
Performance notes (brand / non-brand):
-
Query decisions:
- [ ] Ship title/meta:
- [ ] Plan content refresh:
- [ ] Monitor:
Page / cannibalization decisions:
-
Indexing / sitemap decisions:
-
Enhancements / CWV:
-
Tickets filed (owner, due):
-
Explicit no-actions:
- Keep logs lightweight. The point is institutional memory: why you changed a title, why you ignored a dip, why a sitemap rule exists. Browse more practical articles in the blog, and use technical SEO tools when the weekly review turns into implementation work.