Technical problems rarely announce themselves. A stray noindex tag, a sitemap full of redirected URLs, or a JavaScript menu Googlebot cannot follow will quietly cap your rankings no matter how many links you buy. Most audits find the same handful of issues; the hard part is knowing which to fix first.
This checklist covers the areas that matter in the order a crawler meets them: crawlability, indexation, architecture, page experience, structured data, rendering, international targeting, logs and redirects. It ends with a prioritized fix-first table and the three tools you actually need.
Crawlability and indexation
Crawlability
If Googlebot cannot reach a page, nothing else matters.
- robots.txt. Confirm you are not blocking CSS, JavaScript, or directories that hold real pages. A leftover "Disallow: /" from staging is the classic disaster.
- XML sitemap. List only canonical, indexable, 200-status URLs. Submit it in Search Console and check the "Sitemaps" report. Remove redirected, noindexed and 404 URLs; they waste crawl requests.
- Crawl budget. Mostly a concern above tens of thousands of URLs. Check Search Console's "Crawl stats" for requests hitting low-value URLs (filters, calendars, internal search) and block or consolidate them.
- Orphan pages. Compare your crawler's URL list against the sitemap and Search Console's indexed pages. Anything Google knows about but your crawler never found has no internal links; link it or remove it.
Indexation
Search Console's "Pages" report shows what is indexed and why other URLs are not.
- noindex. Crawl and filter for pages with a noindex meta tag or X-Robots-Tag header. Every one should be deliberate; plugin defaults noindex pages more often than you would think.
- Canonicals. Every indexable page needs a self-referencing, absolute canonical that does not point to a redirect. Check "Duplicate, Google chose different canonical than user" in Search Console to see where Google disagrees with you.
- Duplicate content. Look for HTTP and HTTPS, www and non-www, trailing-slash variants, and paginated or sorted copies of the same list. One version is canonical; the rest redirect or canonicalize to it.
- URL parameters. Tracking and sorting parameters create endless duplicates. Canonicalize them to the clean URL and keep parameter URLs out of internal links.
Site architecture and internal links
Internal links tell Google which pages you consider important. Money pages five clicks deep with two internal links will underperform.
- Sort your crawl by depth. Anything important should be within three clicks of the homepage.
- Sort by inlink count. If an old blog post has more internal links than your main service page, fix that.
- Check internal anchor text. "Click here" wastes a signal; the anchor text optimization guide covers the balance.
- Update internal links that point to redirected or 404 URLs.
A page with no internal links is a page you have told Google does not matter.
Page experience
Core Web Vitals
Search Console's Core Web Vitals report groups URLs as "poor," "needs improvement" or "good" for Largest Contentful Paint, Interaction to Next Paint and Cumulative Layout Shift. Fix the "poor" groups first. Usual causes: oversized hero images, render-blocking scripts, late-loading fonts, and ads or embeds that push content around.
Mobile
Google indexes the mobile version of your page. Confirm it has the same content, structured data and internal links as desktop. Tap targets, viewport meta tags and text that needs zooming still cause problems on older templates.
HTTPS
Every page should be HTTPS, with HTTP redirecting in one hop. Check for mixed content (HTTPS pages loading HTTP assets) and confirm the certificate is not expiring.
Structured data
Schema markup does not directly raise rankings, but it makes you eligible for rich results and helps Google connect your pages to entities. Validate with the Rich Results Test and check Search Console's enhancement reports for errors.
| Type | Where | What to check |
|---|---|---|
| Organization | Homepage | Name, logo, URL, sameAs links to social profiles; consistent with your About page |
| Article | Blog posts | headline, author (as a Person with a URL), datePublished, dateModified, image |
| FAQPage | Pages with genuine Q&A | Questions and answers visible on the page; not stuffed with keywords |
| Product | Product pages | name, image, offers with price and availability, aggregateRating only if real reviews exist |
Never mark up content that is not visible on the page; that is a spam policy violation.
Rendering, hreflang and logs
Rendering and JavaScript
Google renders JavaScript, but later and with less budget than a plain HTML crawl. If navigation, product listings or article bodies only appear after a script runs, some of it may never be indexed.
- Use Search Console's URL Inspection tool and compare the rendered HTML with what users see.
- Crawl once with JavaScript rendering off and once on. Big differences in link counts point to client-side navigation.
- Serve primary content and internal links in the initial HTML. Server-side rendering or static generation solves most of this.
- Avoid links that are only onclick handlers with no href; crawlers do not follow them.
International targeting
If you serve multiple countries or languages, hreflang tells Google which version to show. The mistakes are consistent: missing return tags (page A points to B but B does not point back), wrong language or region codes, pointing to redirected URLs, and no x-default entry. Crawl with hreflang validation on and fix every reciprocity error. Never point canonicals across languages.
Log files
Server logs show what Googlebot actually requests. Pull a month of logs, filter to verified Googlebot, and look for three things: URLs it hits often that you do not want indexed, important pages it rarely visits, and 4xx or 5xx responses. Logs are the only reliable way to confirm a crawl budget problem is real.
Redirects and 404s
- Redirect chains and loops. Crawl for chains longer than one hop and collapse them.
- Redirect type. Permanent moves should be 301 or 308, not a 302 left in place.
- 404s with backlinks. Cross-reference Search Console's "Not found" URLs with your backlink data. Any 404 with external links is wasting authority; redirect it to the closest relevant live page. This is the cheapest link win available, and it pairs well with how to build high-quality backlinks.
- Soft 404s. Pages returning 200 with empty templates. Return a real 404 or fill the page.
What to fix first
Sort by impact against effort.
| Issue | Impact | Effort | Priority |
|---|---|---|---|
| Accidental noindex or robots.txt block on important pages | Very high | Low | 1 |
| 404 pages with external backlinks | High | Low | 2 |
| Wrong or missing canonicals causing duplicate indexing | High | Low to medium | 3 |
| Key content or links only rendered by JavaScript | High | Medium to high | 4 |
| Redirect chains and outdated internal links | Medium | Low | 5 |
| Orphaned money pages | Medium | Low | 6 |
| "Poor" Core Web Vitals groups | Medium | Medium | 7 |
| hreflang errors (international sites only) | Medium | Medium | 8 |
| Structured data errors | Low to medium | Low | 9 |
Publishers selling placements on BacklinkCAT should clear the top five before listing. A site with indexation problems passes less value to buyers, and the automatic site health checks on listings will surface some of these issues.
Tools you actually need
- Google Search Console. Free, and the only source of truth for what Google indexed, which canonicals it chose, manual actions and Core Web Vitals field data.
- A crawler. Screaming Frog, Sitebulb or a cloud crawler finds noindex tags, canonical mismatches, redirect chains, orphan pages, depth problems and hreflang errors site-wide.
- PageSpeed Insights. Lab and field data for one URL, with the resources causing LCP and CLS problems named.
Add a backlink tool for the 404 cross-reference; everything else is optional.
Where to start
- Read every "not indexed" reason in Search Console's "Pages" report for pages that should be indexed.
- Run a full crawl and fix priorities one through three from the table above.
- Redirect every 404 that has backlinks.
- Check rendering on one template of each type.
- Schedule the crawl monthly. Templates change, plugins update, and the same issues creep back.