Before you start
- Verified Search Console property
- HTTPS site with stable canonical host
Separate crawl, index, and rank
Crawling discovers URLs; indexing stores eligible URLs in Google’s index; ranking orders indexed URLs for a query.
Fixing robots.txt helps crawl; noindex removes index eligibility; ranking changes need content, links, and system-level quality signals.
Official: https://developers.google.com/search/docs/fundamentals/how-search-works
Read the ranking systems guide once per quarter
Google’s ranking systems guide lists systems such as helpful content, link analysis, and spam fighting — names change over time.
Use it to interpret Search Console messages and public updates instead of third-party “checklists”.
Official: https://developers.google.com/search/docs/appearance/ranking-systems-guide
When traffic moves, segment before you rewrite
In Performance, filter by page, query, country, and device. A sitewide dip often traces to a template or host change, not every article.
Export a 28-day compare window and note deploy dates, plugin updates, and hreflang edits on the same timeline.
Improve substance before micro-optimizing titles
Thin category pages, duplicate faceted URLs, and auto-generated tag archives rarely recover from title-tag shuffling alone.
Merge or noindex low-value archives; strengthen pages that already get impressions but poor CTR.
Document outbound and UGC link policies
Paid links, excessive reciprocal schemes, and spammy comment links can trigger link-related systems.
Use rel="nofollow" or rel="sponsored" where appropriate; moderate comments; audit footer partner links.
Official: https://developers.google.com/search/docs/essentials/spam-policies