In a nutshell: Onpage SEO covers all optimizations directly on your page, from keyword mapping through structure and content to structured data. In 2026 it decides your visibility in Google and in AI answers at once, because the moment Google shows an AI Overview, the click rate on position 1 drops from 27 to 11 percent (Sistrix, 2025).
Onpage SEO covers all optimizations directly on your page: structure, content, meta data, internal links, images and structured data. Why this matters more than ever: the moment Google shows an AI Overview, the click rate on the first organic result drops from 27 to 11 percent (Sistrix, 2025). Sloppy onpage work makes you lose twice.
To draw the line: technical SEO handles crawlability and indexing, such as load times, mobile optimization, duplicate content and XML sitemaps. Offpage SEO is about external signals like backlinks from high-quality websites. These three areas form the foundation of search engine optimization. Beyond them are specialist fields like local SEO or video SEO.
Speed, by the way, is not a purely technical topic but an onpage lever: how to get your load times under control is something we show in our article on Webflow page speed optimization.
Classic onpage SEO is no longer enough in 2026: only around 360 of every 1,000 Google searches in the EU lead to a click on the open web (SparkToro/Datos, 2024), and the trend is falling. The reason is AI Overviews that deliver answers directly in the search results. In parallel, a second channel is emerging: answer engines like ChatGPT, Perplexity and Google AI sent 527 percent more referral traffic between January and May 2025 (SE Ranking, 2025).
The key consequence for your onpage work: from now on you optimize for two readers at once. For the Google algorithm, which still evaluates structure, relevance and authority, and for AI systems, which extract individual passages and cite them as a source. Both reward the same fundamentals: clear answers, sourced facts and a clean structure. How to serve this second channel deliberately is something we explain in our guide to AI search optimization for B2B companies.
The good news: the following 10 steps pay into both channels. We summarize the AI-specific fine-tuning in a dedicated section afterwards.
It all starts with keyword research, because it decides whether your content is in demand at all. The goal is to identify the terms and questions your target audience actually types in. Only then do you integrate these into your content in a targeted way.
It works without paid software too, but it costs time. We use Sistrix or Semrush for a clean overview of search volume, search intent and ranking difficulty. Think in topics rather than single terms: clusters of primary and secondary keywords rank more stably. In B2B especially, niche terms are worth a look, as we show in our article on B2B SEO with low search volume.
Define exactly one focus keyword per page that this page should rank for. This keyword mapping sends clear signals to search engines about which page is relevant for which topic, and improves both user guidance and ranking.
The second reason matters just as much: you avoid keyword cannibalism. If several pages compete for the same keyword, Google cannot decide which is the most relevant and weakens both in case of doubt. A well-thought-out mapping prevents this from the start.
Google rewards a logical structure because it directly influences the user experience. The rules are simple: an outline by topic groups, exactly one H1 per page as the main topic, and below it as many H2 and H3 as needed for sub-structure.
A pro tip for the AI age: answer the question of a heading right in the first paragraph below it, in 40 to 60 words. Both readers who scan quickly and AI systems that cite exactly such passages love this answer-first structure.
And: write in paragraphs. Nobody likes a wall of text, neither people nor search engines. Keep paragraphs under 80 words and write for your audience, simple rather than complicated.
Comprehensive, well-founded content ranks better, because search engines check how completely you cover a topic. The deciding factor here is semantically relevant terms. Prepare a list of such terms before writing so you do not forget any important detail.
A quick, free method: ask an AI tool for semantically relevant terms on your topic. For an article on „vegan dog food“, ChatGPT returns, for example: plant-based proteins, grain-free, omega-3 fatty acids, vitamin B12, hypoallergenic, sustainability. That shows you at a glance which sub-topics you should cover.
But beware of generic AI text. From our own projects we know: the content that ranks and gets cited by AI is exactly the content that offers real experience, own data and concrete examples. No language model can take this individual touch off your hands, it is your most important differentiator. More on this in our article on AI content marketing.
Your focus keyword should appear concisely at the decisive spots, without tipping into keyword stuffing. If the density gets unnaturally high, Google detects the over-optimization and ranks quality lower. Rule of thumb: if the keyword does not fit fluently into a sentence, drop it or use a synonym.
These four spots matter most:
Your title is SEO indicator number one. It makes clear what the page is about, so the focus keyword belongs here.
Sometimes a title with a keyword simply does not sound good. On our own page about „Website Relaunch Agency“ we therefore used a design trick: people interpret relevance visually, search engines only through the H tags in the source code. So we marked a small „over-heading“ as the H1 and the large, promotional sentence as an H2.
So „Website Relaunch Agency“ carries the H1 tag, while „Boost your business with an effective website relaunch“ dominates visually. Mission completed.
Meta title and meta description appear in the search results (SERPs) and are your storefront. The meta title should be 50 to 60 characters, the description 150 to 160. Place the focus keyword as far to the front as possible so it is not cut off, and write in a way that makes users want to click.
Place the keyword organically in body text and subheadings, which signals relevance. Use synonyms in between so the text stays natural and you avoid keyword stuffing.
The URL is a topic indicator too. Use „speaking URLs“ that describe the content in a human-readable way, and integrate the focus keyword. Avoid cryptic strings, auto-generated URLs and special characters.
Internal links are one of the most underrated onpage levers: they connect related content, help Google understand your site structure, promote deeper indexing and strengthen the authority of important pages. For readers, they create clear paths through your topic.
Use meaningful link texts (anchor texts) that describe what the user can expect. Vague texts like „click here“ waste context that search engines and AI systems use for classification.
Images are a traffic channel of their own: Google Lens now processes nearly 20 billion visual searches per month (Google, 2024). Since the crawler does not see images but reads the source code, you have to describe the image content in text.
Two levers matter. First, the file name: „cat-in-tree.jpg“ tells a search engine more than „539xZb9899-RB.jpg“.
Second, the alt attribute, which you set in the CMS or source code. It improves visibility in image search, serves as alternative text when the image fails to load, and is central to accessibility, because it makes the content available to people with visual impairments. Describe as precisely as possible and as generally as necessary, and add a relevant keyword where it fits.
You optimize videos primarily where they are hosted, usually on YouTube. Through title, description, tags and transcript you describe the content for search. Here is how you transfer the effect to your website:
Structured data are labels you use to explain the content of individual elements to search engines, such as reviews, prices or video details. The measurable effect: pages with FAQ schema increased their traffic by around 25 percent after the markup was applied (seoClarity, 2024). You usually use JSON-LD for this.
Structured data are doubly valuable in 2026: they not only create more eye-catching rich snippets in Google, they also help AI systems assign the entities and facts of your page unambiguously. You can find a detailed guide in Google's introduction to structured data.
Freshness is a ranking and trust factor: search engines want to give users relevant, up-to-date information at all times. Revise existing content regularly, add new developments and update the publication date.
For AI search this applies even more: answer engines favor current, sourced content. Content that is substantially updated every year has clearly better chances of being cited. A practical template for this is our website relaunch checklist.
Anyone optimizing onpage in 2026 should also pay attention to Answer Engine Optimization (AEO), because ChatGPT alone accounts for around 78 percent of AI referral traffic (SE Ranking, 2025). And in Germany a good one in three companies already uses AI actively (36 percent, Bitkom, 2025), so your target audience is searching there too.
For AI systems to name you as a source, you need citable content. In concrete terms: short, self-contained answer passages of 40 to 60 words, one sourced number per statement, clear entities (who, what, where) and an FAQ section with real value. A complete answer always sits directly under its heading.
The most important AEO levers onpage:
How classic SEO and AI optimization differ is something we explain in detail in AEO vs. GEO for B2B and in the practical article How to optimize your website for AI search results.
To help you keep track of all these tips, here is a checklist that sums up every essential step:
Onpage SEO covers all optimizations directly on your website, such as structure, content, meta data, internal links and structured data. Offpage SEO refers to external signals like backlinks from other websites. Both complement each other, but onpage is fully in your own hands.
One focus keyword per page is the rule of thumb. It prevents keyword cannibalism, where several pages compete for the same term. You add semantically related secondary keywords so you cover the topic fully, without pitting your pages against each other.
When Google shows an AI Overview, the click rate on position 1 drops from 27 to 11 percent (Sistrix, 2025). Onpage SEO therefore has to optimize for Answer Engine Optimization too: citable answer passages, sourced facts and FAQ schema, so AI systems name your content as a source.
At least once a year you should substantially revise important pages, update statistics and add new developments. Freshness is a ranking factor and increases the chance of being cited by AI systems, which favor current, well-sourced content.
Not necessarily, but it pays off wherever it fits. FAQ, article, product or video schema improve your rich snippets in Google and help AI systems classify your content correctly. In one test, traffic rose by around 25 percent after FAQ schema was applied (seoClarity, 2024).
