The On-Page SEO Checklist for 2026
A no-jargon on-page SEO checklist covering titles, headings, internal links, schema, and the AI-extractability basics most checklists miss
On-page SEO is everything you control on the page itself to help it rank and get cited: the title, headings, content structure, internal links, schema, and page basics. Most checklists stop at title tags and keyword placement. In 2026 you also have to make the page easy for AI engines to extract, because a growing share of search traffic now flows through AI answers, where 83% of citations come from outside the organic top 10. This checklist covers both the classic on-page factors and the AI-extractability layer most lists still ignore.
Work through it for every important page. None of these items is hard on its own; the value is in doing them consistently.
Content and structure
- The primary keyword is in the title, the H1, the URL, and the first 100 words.
- The main question is answered in the first 150 words. Do not bury it under a long intro.
- Headings use a real hierarchy: one H1, H2s for sections, H3s for subsections.
- H2s are phrased the way people search ("how do I...", "what is...") where it fits naturally.
- Each section opens with a direct, self-contained answer of roughly 40 to 80 words, so it makes sense if an AI lifts it out of context.
- The page is scannable: short paragraphs, lists, and at least one table where a comparison or set of data fits.
- Claims are specific and sourced. Numbers have units and dates.
- The content covers the related sub-questions, not just the head term. This feeds Google's query fan-out and makes you eligible for more AI citations.
Title and meta
- The title tag is compelling and under ~60 characters so it does not truncate in results.
- The meta description is 120 to 155 characters, includes the keyword, and reads like ad copy, not a flat summary.
- The URL is short, readable, and keyword-relevant. No dates, no random IDs, no deep nesting.
Links
- Two to three internal links to related pages, using descriptive anchor text (not "click here").
- A link up to the relevant pillar or hub page, and a link from the hub down to this one.
- A few outbound links to authoritative primary sources. Linking out to good sources correlates with higher AI citation rates, not lower.
- No broken links anywhere on the page.
Media
- Images have descriptive alt text, which helps accessibility, SEO, and AI selection.
- Images are compressed so they do not slow the page down.
- Where relevant, an embedded video with a transcript. Multimodal pages get selected for AI answers at much higher rates than text-only ones.
Authority and freshness signals
- A real, named author with a bio and a link to a verifiable profile like LinkedIn.
- A visible publish date and a "last updated" date, kept genuinely current.
- The page reflects real experience or original information, not just a rehash of the current top results.
Technical and schema
- The page loads in under ~3 seconds and passes Core Web Vitals.
- It is mobile-friendly and served over HTTPS.
- It is indexed (confirm in Search Console) and not blocked by
noindexornosnippet. - AI search crawlers are allowed in robots.txt (
OAI-SearchBot,PerplexityBot,Google-Extended, and peers). - JSON-LD schema is in place:
Articlewith a linked author at minimum, plusFAQPageorHowTowhere they genuinely apply. - Schema matches the visible content and validates in a structured-data tester.
How to actually use this checklist
Do not try to perfect every page at once. Two practical ways to apply it:
- For new posts: keep the checklist open while you draft, and clear it before publishing. After a few posts the habits stick and you stop needing the list.
- For existing pages: start with the pages that already get impressions but rank on page two (Search Console shows these). They are closest to a breakthrough, so a checklist pass there has the highest payoff. Work down from there.
The highest-leverage items, if you only do a few, are: answer the question in the first 150 words, put a real author on the page, add Article schema, and keep the updated date current. Those four punch above their weight for both ranking and AI citation.
Where Okara fits
A checklist is only useful if someone runs it, and running it across a whole site, repeatedly, is exactly the kind of task that slips. Okara's SEO agent automates most of this: it audits every page against on-page and technical criteria, surfaces a prioritized fix list (buried answers, missing schema, stale dates, broken links, crawl issues) rather than a wall of jargon, and its coding agent can apply the technical fixes. Instead of manually checking 30 boxes per page, you get the specific gaps and the fixes, refreshed continuously. Point it at your site for an on-page audit.
Frequently asked questions
What's the most important on-page factor? Answering the searcher's question clearly and early, in a well-structured page. Everything else amplifies that. A perfectly optimized page that does not actually answer the query will not rank.
Does keyword density still matter? No. Use your keyword naturally in the key places (title, H1, opening, a heading or two) and then write for the reader. Keyword stuffing does nothing for SEO and actively hurts AI citation.
How often should I update old pages? Refresh important pages at least quarterly, and sooner if the information changes. Pages updated within 60 days are meaningfully more likely to appear in AI answers.
Do I really need schema? It is not strictly required, but it gives a measurable edge in AI answer selection and helps search engines parse your page. Start with Article and an author, and add FAQ or HowTo where they fit.
What's the difference between on-page and technical SEO? On-page covers what is on the page itself (content, headings, links, schema). Technical SEO covers site-wide infrastructure (crawlability, speed, indexing, sitemaps). This checklist spans both, since at a small site they blur together.