Work Through 47 On-Page SEO Checks With Evidence, Owners, and Validation

Use the checklist as an issue register, not as a race to make every crawler warning disappear. Start with pages that matter to the business or already have measurable search visibility, then record evidence before changing anything. A medium-severity item can be easier to fix than a critical issue, but effort should not determine severity.

For planning, a within-30-days window for medium-priority work is an operating example rather than a required cadence. Release risk, business value, and the type of defect should determine when a fix is scheduled.

When choosing pages to review first, a historical prioritization example compares pages around positions 11-20 with a page around position 50. This contrast is not evidence that one range will respond faster or produce a specific ranking outcome.

Evidence required: baseline crawl export, live HTML or rendered output, Search Console observations where available, analytics context, page purpose, and the current owner of the affected template or content.

Pass: every checklist item has a documented current state and a reproducible way to verify it. Severity: based on consequence, not issue count. Owner: the team that controls the source of the problem. Corrective action: a testable target state. Validation: rerun the same check after deployment.

If you are auditing more than 20 pages, the existing on-page SEO analyzer can help collect repeatable signals while manual review stays focused on intent, accuracy, and prioritization.

The 47 checks below are organized by category while making every item closable through evidence, pass or fail criteria, severity, ownership, corrective action, and validation.

These 14 checks cover page identity, heading structure, URL clarity, and index eligibility. Many can be quick edits, with 15 minutes per page serving as a historical workflow example, not a service-level expectation.

Title Tags (Items 1-4)

• Item 1: The title accurately names the page's primary topic or query intent without forced repetition. Evidence required: Capture the live title, page purpose, and target query set. Pass: The title is present, specific to the page, and consistent with the visible content. Severity: High when the title is missing or materially misleading; otherwise medium or low. Owner: SEO or content; engineering if a template generates the defect. Corrective action: Rewrite the title to describe the page accurately and update the source template if needed. Validation: Recrawl or inspect the live HTML and confirm the intended title is rendered.
• Item 2: Review title length as a presentation cue; 50-60 characters is a historical screening range, not a fixed search-engine limit. Evidence required: Record the title and how it appears for representative queries and devices. Pass: The title communicates the page clearly without unnecessary truncation risk. Severity: Medium when important meaning is routinely lost; low when the current title remains clear. Owner: SEO or content. Corrective action: Shorten or restructure only when clarity improves. Validation: Verify the live title and observe search-result presentation separately from implementation.
• Item 3: The title is intentionally unique where page purpose is unique. Evidence required: Export titles across the crawl and identify exact or near duplicates. Pass: Duplicate titles are justified by genuinely equivalent page purposes or have been differentiated. Severity: High for broad template duplication that obscures important pages; otherwise medium. Owner: SEO, content, or engineering depending on the source rule. Corrective action: Differentiate the title or correct the template that generated unintended duplication. Validation: Recrawl and confirm the duplicate set is resolved as intended.
• Item 4: The title matches the page's actual content and user intent. Evidence required: Compare title, primary heading, body content, and target query set. Pass: A reviewer can predict the page content accurately from the title. Severity: High when the title promises content the page does not provide; otherwise medium. Owner: Content or SEO. Corrective action: Align title and page content around the same purpose. Validation: Review the published title and page together after deployment.

Header Structure (Items 5-8)

• Item 5: The page has a clear primary H1 that represents the page topic. Evidence required: Capture the rendered heading outline. Pass: The primary heading is present and unambiguous. Severity: Medium for editorial inconsistency; high when a template defect removes the page topic. Owner: Content or engineering. Corrective action: Correct the heading in the CMS or template. Validation: Inspect the rendered page and heading outline.
• Item 6: The H1 uses language that matches the page topic without requiring an exact-match formula. Evidence required: Compare the primary heading with the page purpose and query research. Pass: The heading clearly identifies what the page is about. Severity: Medium when wording creates material ambiguity; otherwise low. Owner: Content or SEO. Corrective action: Rewrite the heading for clarity and intent alignment. Validation: Confirm the live primary heading and review it in page context.
• Item 7: H2 and H3 headings form a logical hierarchy after the H1; avoid jumping to H4 when the structure does not justify it. Evidence required: Export or inspect the rendered heading tree. Pass: Heading levels reflect content hierarchy and are understandable to users. Severity: Medium for template-wide hierarchy defects; low for isolated stylistic differences. Owner: Content or engineering. Corrective action: Reorder heading levels according to the actual information structure. Validation: Inspect the rendered outline after publication.
• Item 8: Subheadings describe real sections rather than acting as containers for repeated keywords. Evidence required: Review every subheading against the paragraph or section beneath it. Pass: Each subheading accurately summarizes the content it introduces. Severity: Medium when headings are misleading or spammy; otherwise low. Owner: Content. Corrective action: Rewrite headings to match the underlying section. Validation: Manual review of the published page.

URL Structure (Items 9-11)

• Item 9: The URL is stable, descriptive, and appropriate for the page's role. Evidence required: Record the live URL, internal links, canonical, and redirect behavior. Pass: The URL is understandable and does not require a change solely for keyword inclusion. Severity: High when instability or duplication creates crawl or canonical conflicts; otherwise low. Owner: SEO and engineering. Corrective action: Change only when there is a durable architecture reason, with redirects and internal links updated. Validation: Recrawl the old and new paths and confirm the intended final destination.
• Item 10: Word separators and URL formatting are consistent with the site's established routing conventions. Evidence required: Compare the URL with sibling routes and CMS rules. Pass: The path follows the site's stable convention and resolves cleanly. Severity: Medium for template inconsistency that creates duplicate paths; otherwise low. Owner: Engineering. Corrective action: Normalize routing and redirect unintended variants. Validation: Recrawl route variants and confirm one intended destination.
• Item 11: Dynamic parameters are used only when they serve a real function and their crawl or canonical behavior is intentional. Evidence required: Export parameterized URLs, canonicals, internal links, and indexability. Pass: Parameters do not create unintended indexable duplicates or crawl traps. Severity: High when parameter expansion creates duplicate or uncontrolled URLs; otherwise medium. Owner: Engineering and SEO. Corrective action: Consolidate, canonicalize, block, or retain parameters according to documented intent. Validation: Recrawl representative parameter states and confirm the expected index behavior.

Indexability (Items 12-14)

• Item 12: The intended page is not accidentally blocked from crawling or indexing. Evidence required: Check robots directives, meta robots, response headers, authentication, and live accessibility. Pass: The page is crawlable and indexable when search visibility is intended. Severity: Critical for important pages unintentionally excluded from search. Owner: Engineering or SEO. Corrective action: Remove unintended blocks or document why the page should remain excluded. Validation: Fetch and recrawl the live URL after the change.
• Item 13: The canonical identifies the page that should represent this content. Evidence required: Capture the declared canonical, final response URL, internal links, and duplicate variants. Pass: Canonical intent matches the page's actual preferred destination. Severity: Critical when an important page canonicals elsewhere unintentionally. Owner: SEO and engineering. Corrective action: Correct the canonical source rule and align internal links with the preferred URL. Validation: Recrawl and confirm canonical, response, and internal-link consistency.
• Item 14: The intended page returns a 200 status and is not a soft 404 or part of an unnecessary redirect chain. Evidence required: Record status code, response body, redirect hops, and final destination. Pass: The live page returns the intended content directly when it should be indexable. Severity: Critical when important content is unavailable or misleadingly reported as successful. Owner: Engineering. Corrective action: Restore the intended content or correct the redirect and response behavior. Validation: Request the URL again and recrawl to confirm the final status and content.

These 17 checks require more judgment because they cover intent, content quality, internal navigation, snippet copy, and media. Treat a failure as an evidence-backed content or architecture issue rather than as an automatic ranking explanation.

Content Quality (Items 15-20)

• Item 15: The page answers the intended search task completely enough for its purpose. Evidence required: Record target queries, page purpose, major subtopics, and relevant competing results. Pass: The page addresses the core user need without obvious unsupported gaps. Severity: High for a priority page that materially misses intent. Owner: Content or subject-matter owner. Corrective action: Revise the brief and add or correct only the information needed to satisfy intent. Validation: Editorial review of the published page against the approved brief.
• Item 16: The primary topic is evident early in the page; the first 100 words serve as a historical review checkpoint, not a stuffing rule. Evidence required: Capture the opening copy and target query set. Pass: A reader can identify the page topic quickly without forced repetition. Severity: Medium when the introduction is misleading or unrelated. Owner: Content. Corrective action: Rewrite the opening for clarity and relevance. Validation: Review the live introduction in context.
• Item 17: Related subtopics are included when they are necessary to satisfy the user's task; competing top-10 pages can serve as one research input. Evidence required: List recurring relevant subtopics from query research and competing results. Pass: The page covers necessary subtopics without copying competitors mechanically. Severity: High when a major information gap prevents task completion; otherwise medium. Owner: Content or SEO. Corrective action: Add the missing material with original, accurate explanation. Validation: Compare the revised page with the approved content brief.
• Item 18: Keyword use reads naturally and does not rely on repeated exact-match phrasing. Evidence required: Review body copy, headings, anchors, and repeated terminology. Pass: Language is clear to users and repetition is justified by meaning. Severity: Medium when repetition harms readability or signals low-quality copy. Owner: Content. Corrective action: Edit for natural language and preserve necessary terminology. Validation: Manual read-through of the published copy.
• Item 19: The page gives a direct answer or next step when the query calls for one. Evidence required: Identify the user question and the section intended to answer it. Pass: The answer is easy to find and does not overclaim. Severity: High when the page fails the core user task; otherwise medium. Owner: Content. Corrective action: Add a concise, accurate answer in the appropriate section. Validation: Manual review and rendered-page verification.
• Item 20: Claims are accurate, appropriately qualified, and sourced when the source is available. Evidence required: List material factual claims and their supporting evidence. Pass: Claims can be traced to evidence or are clearly framed as opinion, observation, or uncertainty. Severity: High for unsupported consequential claims. Owner: Content, subject-matter reviewer, or compliance owner as appropriate. Corrective action: Correct, qualify, or remove unsupported statements. Validation: Editorial and evidence review of the live page.

Internal Linking (Items 21-25)

• Item 21: A target of at least 2-3 relevant internal links is an operating example; use relevance and navigation value rather than a fixed quota. Evidence required: Export outbound internal links and review their context. Pass: The page links to useful related destinations where a user would reasonably need them. Severity: Medium when a priority page is structurally isolated; otherwise low. Owner: Content or information architecture owner. Corrective action: Add contextually useful links, not arbitrary count-based links. Validation: Recrawl and manually test each new destination.
• Item 22: Anchor text describes the destination clearly. Evidence required: Export anchor text with source and destination URLs. Pass: A user can anticipate the destination from the anchor and surrounding sentence. Severity: Medium when ambiguous anchors impair navigation; otherwise low. Owner: Content. Corrective action: Rewrite vague anchors where clearer wording improves the path. Validation: Review the live anchor in context.
• Item 23: Relevant high-visibility or well-linked pages connect to this destination when that relationship helps users. Evidence required: Identify likely source pages and inspect their outbound links. Pass: Important destinations are connected from appropriate source pages without forced linking. Severity: High for strategically important orphaned pages; otherwise medium. Owner: Content, SEO, or information architecture owner. Corrective action: Add a contextual link from a genuinely related source page. Validation: Recrawl and confirm the source-to-destination edge.
• Item 24: Indexable pages are not unintentionally orphaned. Evidence required: Compare crawl discoveries with sitemap, analytics, and CMS URL lists. Pass: Every intended search page has at least one meaningful crawlable internal path. Severity: High when an important page is isolated. Owner: Information architecture or content owner. Corrective action: Add an appropriate internal path or document intentional isolation. Validation: Recrawl and confirm discoverability from the intended source.
• Item 25: Breadcrumbs are present when they improve orientation and accurately reflect the hierarchy. Evidence required: Inspect visible navigation and any breadcrumb markup. Pass: Breadcrumbs, when used, match the real site structure and destinations. Severity: Medium for template-wide mismatch; low when breadcrumbs are unnecessary. Owner: Engineering or information architecture owner. Corrective action: Correct the visible hierarchy and any corresponding markup. Validation: Test visible links and validate the rendered structure.

Meta Description (Items 26-28)

• Item 26: Treat 120-155 characters as a historical review range; prioritize clarity over a fixed length. Evidence required: Record the current description and search-result presentation for representative queries. Pass: The description accurately summarizes the page and retains its key message when displayed. Severity: Medium for misleading or missing descriptions on high-impression pages; otherwise low. Owner: Content or SEO. Corrective action: Rewrite for accuracy and useful snippet context. Validation: Confirm the live description and observe snippet behavior separately.
• Item 27: The description states a truthful reason the page may be useful without unsupported promises. Evidence required: Compare description copy with the page's actual content. Pass: Every benefit or claim in the description is supported on the page. Severity: Medium when snippet copy overpromises. Owner: Content or SEO. Corrective action: Align the description with verifiable page value. Validation: Review live metadata against the visible page.
• Item 28: Descriptions are intentionally differentiated where page purposes differ. Evidence required: Export descriptions and identify duplicates. Pass: Duplicates are justified by equivalent intent or have been rewritten. Severity: Medium for broad template duplication; otherwise low. Owner: Content or engineering. Corrective action: Differentiate copy or fix the generating template. Validation: Recrawl and confirm intended uniqueness.

Image Optimization (Items 29-31)

• Item 29: Informative images have alt text that conveys their relevant purpose; decorative images are handled appropriately. Evidence required: Export image alt attributes and inspect the images in context. Pass: Alt text is accurate and useful for accessibility without keyword stuffing. Severity: High for accessibility-critical content; otherwise medium. Owner: Content, design, or engineering. Corrective action: Add or correct alt text according to the image's function. Validation: Inspect the rendered image and accessibility tree.
• Item 30: Image file names are maintainable and descriptive when naming is part of the asset workflow; avoid opaque names such as IMG_4821.jpg when a clearer name is practical. Evidence required: Review asset names and delivery paths. Pass: Names follow the site's asset convention without breaking URLs unnecessarily. Severity: Low unless asset handling creates operational or crawl problems. Owner: Content operations or engineering. Corrective action: Rename only when the benefit outweighs migration risk and references can be updated safely. Validation: Confirm the image loads and all references resolve after any change.
• Item 31: Image weight and dimensions are appropriate for how the image is rendered. Evidence required: Collect transfer size, intrinsic dimensions, rendered dimensions, and loading behavior. Pass: Images are delivered efficiently without harming necessary visual quality. Severity: High when oversized assets materially affect page performance; otherwise medium. Owner: Engineering or design. Corrective action: Resize, compress, or change delivery settings based on measured waste. Validation: Retest the page and compare the same performance evidence.

Use this tier after higher-severity crawl, index, content, and navigation defects are understood. Do not skip a missing H1 or a canonical conflict merely because structured data or performance tools surface more visible warnings. See the related on-page SEO mistake guide for common prioritization errors.

Structured Data and Schema (Items 32-36)

• Item 32: Structured data, when used, matches the visible content and a supported content type. Evidence required: Capture the live markup, visible content, and current documentation for the selected type. Pass: Markup is valid, truthful, and consistent with what users can see. Severity: High for misleading markup; medium for invalid markup that the site relies on operationally. Owner: SEO or engineering. Corrective action: Remove unsupported properties or correct the markup to reflect visible content. Validation: Validate the live markup and inspect the rendered page.
• Item 33: Structured data validates without material errors for the intended implementation. Evidence required: Save validator output and the source markup. Pass: No unresolved error contradicts the intended supported type. Severity: Medium unless the markup is misleading or breaks a critical integration. Owner: Engineering or SEO. Corrective action: Correct the specific invalid property or type. Validation: Rerun validation against the live URL.
• Item 34: Question-and-answer markup is not added merely because a page contains questions; any structured data used must reflect visible content and current supported eligibility. Evidence required: Compare visible questions and answers with the markup and current documentation. Pass: Markup, if present, is accurate and justified; absence of markup is not treated as a failure by itself. Severity: Medium for misleading or stale markup; low when no markup is needed. Owner: SEO or engineering. Corrective action: Remove unsupported markup or align it with visible content. Validation: Validate the live page and markup after deployment.
• Item 35: Breadcrumb markup, when present, matches the visible breadcrumb path. Evidence required: Compare structured breadcrumb items with rendered navigation. Pass: Labels and destinations agree with the visible hierarchy. Severity: Medium for template-wide divergence; otherwise low. Owner: Engineering or information architecture owner. Corrective action: Align visible and structured breadcrumb paths. Validation: Test each link and validate the markup.
• Item 36: Organization or WebSite markup is used only where it accurately represents the site and follows current documentation. Evidence required: Inspect the markup source and the visible site identity information. Pass: Claims and properties in the markup are accurate and supported. Severity: High for fabricated identity information; otherwise medium or low. Owner: Engineering or SEO. Corrective action: Correct or remove inaccurate properties. Validation: Validate the live markup and compare with visible site information.

Core Web Vitals and Page Experience (Items 37-41)

• Item 37: Largest Contentful Paint is reviewed against a reference threshold of 2.5 seconds on mobile, not as a guarantee of ranking change. Evidence required: Collect field data where available and controlled lab diagnostics for the same URL. Pass: The page meets the adopted performance target or has a documented remediation plan. Severity: High when poor loading blocks user tasks; otherwise medium. Owner: Engineering. Corrective action: Optimize the measured bottleneck, such as server response, render-blocking resources, or hero media. Validation: Retest with the same field or lab method after deployment.
• Item 38: Cumulative Layout Shift is reviewed against a threshold of 0.1 as a stability reference. Evidence required: Capture field data and identify elements contributing to layout movement. Pass: Unexpected shifts are within the team's adopted target. Severity: High when movement disrupts interaction; otherwise medium. Owner: Engineering or design. Corrective action: Reserve space, stabilize fonts, or fix the specific shifting component. Validation: Retest the same page and interaction path.
• Item 39: Interaction to Next Paint is reviewed against a threshold of 200ms as a responsiveness reference. Evidence required: Collect field data where available and profile slow interactions. Pass: Priority interactions meet the adopted responsiveness target. Severity: High when interaction delay impairs task completion; otherwise medium. Owner: Engineering. Corrective action: Reduce main-thread work or optimize the measured interaction path. Validation: Retest the same interaction after deployment.
• Item 40: The page works across intended mobile viewports without unintended horizontal scrolling or blocked controls. Evidence required: Test representative devices and viewport sizes. Pass: Content and controls remain usable without layout breakage. Severity: High when users cannot complete core tasks; otherwise medium. Owner: Engineering or design. Corrective action: Correct responsive layout, overflow, or component sizing. Validation: Repeat the same device and viewport tests.
• Item 41: Interstitials and overlays do not prevent users from accessing primary content or controls. Evidence required: Document triggers, timing, viewport behavior, and dismiss controls. Pass: Overlays are usable, dismissible, and do not obstruct the primary task unnecessarily. Severity: High when content or controls become inaccessible; otherwise medium. Owner: Product, design, or engineering. Corrective action: Change trigger, size, placement, or dismissal behavior. Validation: Retest the interaction on representative devices.

Advanced Content Signals (Items 42-47)

• Item 42: Outbound links are used when they help users verify or continue from the information on the page. Evidence required: Review external destinations, anchor context, and relevance. Pass: Links are accurate, working, and editorially justified. Severity: High for harmful or misleading destinations; otherwise low. Owner: Content. Corrective action: Remove, replace, or update links that no longer support the page. Validation: Test each changed destination and review its context.
• Item 43: Freshness information is shown when a date materially helps users understand the currency of the content. Evidence required: Identify claims or instructions that can become outdated and compare them with the visible date context. Pass: Dates are accurate and not refreshed cosmetically without substantive review. Severity: Medium for time-sensitive guidance; low for timeless content. Owner: Content. Corrective action: Update the content and date only when a substantive review occurred. Validation: Review the live date and the changed sections together.
• Item 44: Author or reviewer information is included when it helps users evaluate responsibility and expertise, without fabricating credentials. Evidence required: Capture byline, reviewer, role, and any claims of expertise. Pass: Identity and credentials shown on the page are accurate and supportable. Severity: High for fabricated credentials; otherwise medium. Owner: Editorial or compliance owner. Corrective action: Correct or remove unsupported identity or expertise claims. Validation: Verify the live byline and supporting profile information.
• Item 45: For long pages, greater than 1,500 words is a historical reference point where jump navigation may help; use actual page complexity rather than a fixed word threshold. Evidence required: Review page length, section count, user tasks, and current navigation. Pass: Users can reach major sections efficiently when the page's structure warrants it. Severity: Medium for difficult navigation on high-use pages; otherwise low. Owner: Content, design, or engineering. Corrective action: Add or improve a table of contents or jump links when useful. Validation: Test links and keyboard navigation on the live page.
• Item 46: Reading level and terminology fit the intended audience and task. Evidence required: Review a representative sample with the audience and subject matter in mind. Pass: Language is understandable without removing necessary technical precision. Severity: High when ambiguity creates material misunderstanding; otherwise medium. Owner: Content or subject-matter owner. Corrective action: Rewrite unclear passages while preserving accuracy. Validation: Editorial review of the published revision.
• Item 47: The page passes a manual quality review covering accuracy, usefulness, clarity, originality, and technical consistency. Evidence required: Use the completed issue record, final live page, and unresolved-risk notes. Pass: No unresolved critical defect remains and any accepted limitations are documented. Severity: Critical if material inaccuracies or search-access defects remain; otherwise according to the open issue. Owner: Editorial owner with SEO and engineering sign-off where relevant. Corrective action: Resolve remaining defects or document an explicit acceptance decision. Validation: Repeat the manual review on the live page and record closure.

After all 47 checks have been reviewed, use tooling to maintain the baseline where repeatable signals can be automated. For sites with 50+ pages, the existing on-page analyzer can collect recurring checks while owners keep manual review for judgment-heavy decisions.

Not all 47 findings deserve the same response. Use severity to describe consequence and effort to describe implementation cost. Keeping those dimensions separate prevents an easy cosmetic edit from displacing a harder but more consequential fix.

High Impact, Low Effort

• Title or heading corrections can be fast when the issue is isolated; about 10 minutes per page and a 2-4 week observation window are historical examples. Do not treat that window as a promised ranking response.
• An H1 correction can be simple when the CMS exposes the field, but template defects may require engineering ownership.
• Internal-link additions can be low effort when a relevant source page already exists; 2-3 links and about 20 minutes are operating examples, not quotas or guarantees.
• Metadata and image-alt corrections are useful when evidence shows they are inaccurate, missing, or inaccessible, but they should not outrank a critical indexing defect merely because they are easy.

High Impact, Higher Effort

• Content-gap remediation may require research, subject review, writing, and editing. Schedule it when the page is strategically important and the missing information is evidence-backed.
• Performance work may require engineering changes. Diagnose the measured bottleneck rather than treating Core Web Vitals as a universal explanation for search performance.
• Structured data should be corrected when inaccurate or invalid, but do not implement markup solely to chase an unsupported search feature.

Lower Impact or Batchable Work

• Image compression, canonical monitoring, and breadcrumb checks can often be automated at template or pipeline level once the intended rules are documented.
• Batch work still needs validation. Automation reduces collection effort; it does not remove the need to confirm that the resulting page state is correct.

The existing on-page SEO tool benchmarks page can provide supporting context, but any unsourced benchmark should be reconciled before it is presented as verified evidence.

Historical workflow estimates suggest that a manual pass can take 45-60 minutes on a single page, that reviewing 10 pages can consume a working day, and that 100 pages can become a multi-day project. Treat those values as historical workflow examples rather than guaranteed effort estimates.

Around 15-20 pages is an operating reference point where automation may become more useful. Page complexity, template reuse, data quality, and team experience matter more than a fixed threshold. Use the related on-page SEO audit guide when the problem is diagnosis and prioritization rather than checklist execution.

Good tooling can collect repeatable evidence such as title and description duplication, missing H1 values, internal-link counts, image-alt coverage, canonical conflicts, index directives, status codes, and performance observations. It should surface evidence, not declare causality.

Human review is still required for search intent, factual accuracy, editorial quality, claim support, useful internal-link context, severity, business priority, and acceptance of residual risk.

The best operating model uses on-page SEO tools to collect and recheck repeatable signals while owners reserve manual review for judgment-heavy decisions. The goal is not to automate every decision; it is to make evidence collection consistent enough that reviewers can focus on what the page should actually do.