Most on-page SEO checklists still circulating were written before AI-generated answers started absorbing a meaningful share of search traffic, and haven't been updated to account for it. This checklist covers the fundamentals plus the additions that actually matter now; for the structural layer around individual pages, see my internal linking strategy and generative engine optimization checklist.
The fundamentals that haven't changed
- Title tag — unique per page, includes the primary target term, under roughly 60 characters to avoid truncation in results.
- Meta description — every page should have one; a missing meta description is a surprisingly common miss even on otherwise well-built sites, and it directly affects click-through rate on the pages that do rank.
- Single, descriptive H1 — one per page, matching the actual content and search intent, not a generic company tagline reused everywhere.
- Logical heading hierarchy — H2s and H3s that genuinely reflect the content's structure, not just formatting for visual variety.
- Descriptive URL slugs — readable, keyword-relevant, and stable once published, since changing them later requires redirects to preserve existing rankings.
- Image alt text — describes the image content accurately, both for accessibility and for image search visibility.
What's newly essential in the AI-answer era
- Direct-answer paragraphs early in the content — a clear, quotable answer to the page's core question near the top, since AI overviews and answer engines tend to pull from concise, well-structured direct answers rather than content buried after several paragraphs of preamble.
- FAQPage schema for genuine Q&A content — structured markup makes it easier for both traditional rich results and AI answer systems to parse and cite specific answers.
- Clear source attribution and specificity — content citing where numbers and claims come from tends to be favored by systems trying to verify what they're synthesizing from, over vague, uncredited claims.
| Checklist Category | Key Items |
|---|---|
| Metadata | Unique title tag, meta description on every page |
| Structure | Single H1, logical H2/H3 hierarchy, descriptive URLs |
| Content quality | Direct-answer paragraphs early, specific sourcing |
| Schema | FAQPage, BreadcrumbList, and content-type-appropriate structured data |
| Technical | Alt text, internal links, mobile page experience |
The check most checklists skip
Before optimizing a new page, check whether an existing page on the site already targets the same core term — publishing without this check is the most common cause of keyword cannibalization, covered in more depth in my keyword cannibalization guide. A checklist run against a page that shouldn't have been created in the first place doesn't fix the underlying problem.
How to use this without over-optimizing
Every item here should read naturally on the page once implemented — a title tag stuffed with keywords or a direct-answer paragraph that reads like it was written for a machine rather than a person tends to underperform a more natural version, even on the exact metrics the checklist is meant to improve. The checklist is a floor to meet, not a formula to maximize past the point of natural readability.
FAQ
Does on-page SEO still matter if AI answers are absorbing more search traffic?
Yes, arguably more than before — AI answer systems still need to find, parse, and trust a source to cite, and well-structured on-page content with clear direct answers, proper schema, and specific sourcing is exactly what makes a page easier for those systems to pull from and credit.
- Being citable by AI answer systems depends on the same structural clarity that traditional on-page SEO already rewards.
- Direct-answer content early in a page tends to be favored by both traditional rankings and AI citation.
What's the most commonly missed item on this checklist?
A missing or generic, non-unique meta description is one of the most common misses even on otherwise well-optimized sites, and it directly affects click-through rate on pages that do rank, making it a disproportionately high-impact fix relative to how often it gets overlooked.
- Meta descriptions are frequently skipped despite being low-effort to fix.
- Click-through rate impact makes this a higher-leverage fix than its simplicity suggests.