Technical and On-Page SEO Checklist

Crawl the full site and export reports for missing alt text, missing image dimensions, broken links, and duplicate titles.

Check that robots.txt is present and blocks crawling of sensitive directories. It controls crawling only: a blocked page can still be indexed, so use noindex or password protection to keep a page out of Google.

Submit XML sitemap to Google Search Console and Bing Webmaster Tools.

Check that the whole site is served over HTTPS and that HTTP traffic redirects to it.

Largest Contentful Paint must load within 2.5 seconds. Optimize images, server response, and render-blocking resources.

Interaction to Next Paint must be 200 milliseconds or less. Optimize event handlers, reduce main thread blocking, and defer non-critical JS.

Cumulative Layout Shift must be 0.1 or less. Set image dimensions, avoid dynamic content injection above the fold.

Give every page a self-referencing canonical tag. It tells Google which URL you prefer when several URLs show the same content.

Create a helpful 404 page with navigation links. Return a real 404 status code for missing pages, not a 200.

Add hreflang tags for pages with versions in other languages or regions. Each version must list itself and every other version, or Google may ignore the tags.

Use the URL Inspection tool in Google Search Console to check the rendered version of a JavaScript-heavy page. Confirm that Googlebot sees your content, not a blank page.

Find and fix redirect chains, where one redirect points to another. Google counts each hop as a separate request, and chains add latency for users. Google says 301s do not cause a loss in PageRank, but advises redirecting straight to the final URL. If a chain cannot be avoided, Google says to keep it ideally at 3 hops or fewer, and always under 5.

Crawl budget is the set of URLs Google can and wants to crawl. Fix soft 404 pages (not-found pages that do not return a 404 status), and keep the server fast and free of errors, because Google crawls less when a server signals overload. Google says noindex does not save crawl budget, since Google still requests the page before dropping it.

Analyze server logs to understand Googlebot crawl patterns, frequency, and wasted crawl budget.

Use the IndexNow protocol to tell participating search engines when a page is added, updated, or deleted.

Generate an image sitemap with <image:image> tags (up to 1,000 per page URL), submit it in Google Search Console, and verify any image hosting domain, including a CDN, in Search Console.

If you use an image CDN such as Cloudinary, ImageKit, Imgix, or Bunny, check whether it converts images to WebP or AVIF automatically. That removes manual conversion work.

Time to first byte (TTFB) measures how fast your server starts to respond. web.dev says most sites should aim for 0.8 seconds or less. A CDN, faster queries, and caching all help.

Add schema markup only where it matches the visible content on the page, and test it with the Rich Results Test. Google says AI Overviews and AI Mode need no special optimizations, so do not add markup just for them.

Give every page a unique title tag that names its main topic. Google sets no length limit for a title element, but the title link is truncated in search results, typically to fit the device width.

Write a short, specific meta description for every page. Google sets no length limit, but long text is truncated in results.

Give each page a clear H1 (the main heading) that names its topic and primary keyword.

Compress images, use WebP or AVIF where you can, and give files descriptive names.

Add descriptive alt text to every informative image. Google calls alt text the most important metadata for an image, and it helps screen reader users too.

Link relevant pages to each other using descriptive anchor text.

Give each page one primary search term and a few related variations.

Add Article or BlogPosting JSON-LD with the properties that apply, such as headline, datePublished, dateModified, and author. Google lists no required properties for Article.

Add Organization JSON-LD with name, logo, URL, contactPoint, and social profile sameAs links. Google says this can influence which logo appears in results and in your knowledge panel.

Add BreadcrumbList JSON-LD to pages that sit in a site hierarchy. The breadcrumb trail shows readers where a page sits, and Google's breadcrumb feature is available on desktop.

Verify each page matches user intent (informational, navigational, commercial, transactional) for its target keywords.

Add author bios with credentials, link to LinkedIn profiles, cite sources, and show first-hand experience. Google's helpful content guidance calls experience, expertise, authoritativeness, and trustworthiness E-E-A-T.

Review all content pages for stale statistics, outdated dates, and deprecated information. Update or remove.

Check which of your target queries already show a featured snippet. Compare the format of the winning result (paragraph, list or table) with yours before you rewrite the section.

Research the People Also Ask (PAA) questions for your keywords and add Q&A sections that answer them on relevant pages.

Check that headings run in order on all pages (H1, then H2, then H3), without skipping levels.

Pages with very little unique text are often thin content. Review these and either expand, consolidate, or noindex them.

Check for multiple pages targeting the same keyword. Consolidate or differentiate to avoid self-competition.

Compare your keyword coverage against top competitors. Identify topics they rank for that you don't cover.

Write each FAQ answer so it stands alone and answers a question a practitioner asks mid-project, not a definition. Google no longer shows FAQ rich results, so write FAQs for readers.

Create one long pillar page on a broad topic and several spoke articles on narrower questions, and link them all with descriptive anchor text.

Add a jump-link table of contents to long pages. It helps readers scan the page and find the section they need.

Review high-converting, high-CPC queries from your paid search campaigns and create organic content targeting the ones you don't yet rank for.

Implement Product JSON-LD with the required properties, such as name. Add AggregateRating only for real ratings, and mark up only what is visible on the page.

Open your key page templates on a real phone. Check that text reads without zooming, nothing scrolls sideways, and menus and forms work. Then run PageSpeed Insights on mobile.

Verify <meta name='viewport' content='width=device-width, initial-scale=1'> is set on every page.

Make buttons and links large enough to tap with a thumb, with space between them. Small tap targets are hard to use on a phone.

Judge LCP, INP and CLS separately for mobile and desktop. web.dev suggests using the 75th percentile of page loads for each.

Use a body font size that is easy to read on a phone. Search Console flags pages that do not use legible font sizes.

Test hamburger menu, sticky headers, and bottom navigation on real devices. Make every page reachable in a few taps.

Claim your Google Business Profile if you serve customers in person or in a service area. The next items cover completing and maintaining it.

Check that your name, address, and phone number (NAP) are identical across all directories, citations, and social profiles.

Fill every Google Business Profile (GBP) field: hours, services, attributes, products, description, and Q&A.

Upload several high-quality photos (exterior, interior, team, products).

Publish Google Posts weekly with offers, events, or updates. Keep the schedule steady.

Set up a system to ask customers for reviews, and reply to every review, positive and negative, promptly.

Build citations on Yelp, BBB, industry-specific directories, and local business associations. Consistency is critical.

Choose the single most specific primary category in Google Business Profile. Whitespark's 2026 local pack ranking factors list scores the primary category highest.

Audit backlinks with Ahrefs or Google Search Console. Disavow only if you have many spammy or low-quality links you cannot get removed. Google calls the disavow tool advanced and says most sites do not need it.

Use Ahrefs/SEMrush to find sites linking to competitors but not you. Target those domains for outreach.

Create original research, data studies, or tools that naturally attract backlinks from industry publications.

Contribute expert articles to industry publications with relevant audiences. Google treats paid guest-post links that pass ranking credit, and links with optimized anchor text in guest posts, as link spam.

Monitor unlinked brand mentions and request link additions. Use Google Alerts or Ahrefs Alerts.

Check that backlink anchor text varies: brand name, bare URL, generic words, and keywords. Optimized anchor text on links placed on other sites can count as link spam under Google's spam policies.

Track new and lost referring domains each month.

Check that your organization has accurate, consistent profiles on sites such as Wikipedia, Wikidata, and Crunchbase.

Answer relevant media requests on platforms like Source of Sources, Featured, or Help a B2B Writer to earn editorial backlinks.

Find dead links on relevant high-authority pages and offer your equivalent content as the replacement to the site owner.

Ask ChatGPT, Perplexity and Gemini several versions of your brand and product questions. Log every factual error, then publish corrective content on the pages those answers rely on.

Test your top keywords each month on ChatGPT, Perplexity, Claude, and Google AI Overviews. Log which sources get cited and in what format.

Run an accessibility audit (Lighthouse, axe). Fix critical issues: alt text, ARIA labels, keyboard navigation, focus management.

Make every page reachable within a few clicks of the homepage. Pages buried deep in the site are harder for visitors and crawlers to reach.

Pre-build your top ~500 pages by traffic and use on-demand fallback ('blocking') for the rest. Keeps build times under a couple of minutes at 100K+ pages instead of 30+ minute full rebuilds.

Before generation, build an intent matrix (e.g. 'San Francisco Plumbers' = local, 'Plumbing Tools' = commercial). Use it to prevent templates from cannibalizing each other.

Generate a small cohort and measure performance before scaling to thousands. If it underperforms after a few months, fix intent match and data quality before expanding.

Start with 10-20 pages, then ramp to 50-100 per week, so a quality or indexing problem shows before it covers thousands of pages.

Score every generated page on a 100-point scale (data completeness, uniqueness, engagement, technical health). As a rule of thumb, auto-noindex weak pages and flag borderline ones for human review. The exact cutoff is your own quality bar, not a Google standard. A common starting point is to noindex below ~70 and review 70-79.

Measure unique word count across sibling pages and consider noindexing very thin variants to avoid thin-content issues. There's no official Google minimum; a threshold like ~500 unique words is a practical rule of thumb, not an established standard.

Use automated scoring to measure content deviation between templated pages, and merge or consolidate near-duplicate siblings into their parent hub. A target like ~30%+ unique content is a useful internal heuristic rather than a documented Google threshold.

Flag pages with stale data, broken feeds, or outdated API responses before they publish. Prevents wasting crawl budget on dead pages.

Track indexed vs submitted pages in GSC monthly. A low indexation rate is a signal that Google sees limited value in many pages. Fix it via better TTFB, internal linking, and less sitemap spam. A target like ~70%+ is a practical benchmark to aim for, not an official Google threshold.

Create separate, descriptively named sitemaps (sitemap-cities.xml, sitemap-products.xml). Split into many files at scale. Only lastmod matters, so ignore priority and changefreq.

After ~6 months with zero traffic: enrich with unique data, consolidate into a parent, 301 to a sibling, or 410 delete. Don't let ghost pages accumulate and waste crawl budget.

Every child page (e.g. 'Plumbers in San Francisco') must link up to its parent ('Plumbers in California') to create crawlable pathways and authority flow.

Link each page to nearby or related variants (SF → Oakland, San Jose). Creates relevance signals and prevents orphaned content.

Link definitions to comparison pages, comparisons to product directories, and back again to build semantic relationships across template types.

Add the right schema for each template: Product + AggregateRating for reviews, FAQPage for Q&A, LocalBusiness (with lat/long) for locations, BreadcrumbList for hierarchy.

Use a GSC Page regex like ^https://example\.com/(city|state|product)/ to isolate each template and track impressions, CTR, and position separately from editorial content.

Use GA4 regex filters matching your URL structure to compare conversion rates between templates and identify which patterns earn the most revenue.