Category: Search Engine Optimisation

A client came to us having already built their entire site on AI-generated content. Every page, every post, churned out at a pace no human writer could match. Not one piece of it performed. Rankings never came. What did come, eventually, was the work of pulling it all down and starting again – properly this time, with a person behind every page.

That’s not a hypothetical. It’s the shape of a mistake I keep watching people make, and it’s the reason this article exists.

Search “SEO content strategy” and you’ll land on nine-step frameworks that all say the same thing in a different order: research your audience, do keyword research, build a content calendar, optimise on-page, measure results, repeat. None of it is wrong exactly. It’s just not a strategy. It’s a checklist wearing a strategy’s clothes, and following it to the letter still leaves you exactly where you started – producing content because the calendar says today’s the day, not because you’ve got something worth saying.

Here’s the part nobody selling you a framework wants to say out loud: quality beats quantity, and most teams are optimising for the wrong one. Not as a nice sentiment. As the actual difference between a site that compounds and a site that has to be rebuilt.

What a Content Strategy Actually Is (And Why a Calendar Isn’t One)

A content calendar tells you what publishes on what day. A content strategy tells you why any of it exists.

Keyword research is a tactic. It tells you what people are searching for, what the intent behind that search is, whether there’s volume worth chasing. That’s useful – it feeds the strategy. But it isn’t the strategy itself. The strategy is the layer above it: what are we solving for this reader, what’s the theme connecting this piece to the next one, how does this get in front of the person who needs it, and what happens to it after it’s published.

Without that layer, a content calendar is just a map with no destination. You can execute it perfectly – hit every date, tick every box – and still have no idea whether it’s working, because there was never anything to measure it against. No objective means no way to say “this cluster is underperforming, let’s pivot” or “this piece did better than expected, let’s build on it.” You’re just producing.

Ask yourself what each piece of content is actually for. Is it there to build links? Get someone to sign up for a newsletter? Answer a question that stops them cannibalising two other pages on your own site by asking a slightly different version of the same thing? If you can’t answer that before you write it, you don’t have a strategy yet. You have a production schedule.

The Budget Myth: What HubSpot’s Collapse Actually Proves

If money solved this, HubSpot wouldn’t have lost most of its organic traffic.

Between November and December 2024, HubSpot’s monthly organic visits fell from around 13.5 million to 8.6 million – the sharpest month of a slide that took them from roughly 24 million visits down to somewhere near 6 to 7 million. A drop in the region of 70 to 75%. This is a company with a marketing budget most of the sites reading this article will never touch, and it still happened to them.

The trigger was Google’s December 2024 update, which leaned harder into topical authority – rewarding sites that go deep in a specific area and penalising the ones that spread thin across dozens of unrelated ones. HubSpot had spent years publishing content well outside their actual expertise: inspirational quotes, resignation letter templates, generic business forms. Content that generated volume, not relevance. Their topic list grew and grew, and their topical authority – the thing that told Google “this site actually knows CRM and inbound marketing” – diluted along with it.

They had the budget to do everything right. They did everything, instead. That’s not a budget failure. That’s a strategy failure wearing a budget’s clothes.

HubSpot isn’t the only one. I’ve watched the same graph shape play out on a project management brand’s site – traffic climbing hard for a couple of years as the content library grew into every tangentially related topic they could find, then collapsing back down near where it started once the sites’ worth of content stopped being about the product and started being about everything else. Different company, same mistake: mistaking “more topics” for “more authority.”

It’s a useful thing to sit with, because HubSpot’s own research says the opposite approach works: high-quality, relevant content generates 9.5 times more leads than low-quality content aimed at nobody in particular. They had the data. They didn’t follow it. And 83% of marketers, when asked directly, say they’d rather post less often and keep the quality high than chase the calendar. Most people already know the answer. Most people still don’t act on it, because volume feels like progress and quality feels like waiting.

How Smaller Teams Actually Compete: The Pineapple Insurance Approach

Pineapple Insurance had almost no content when we started, and they were up against players with five times their marketing budget in one of the most competitive verticals there is.

We didn’t try to out-publish anyone. We built out one cluster at a time, deliberately slower than what a bigger team could manage, and set a cap on length for every piece – not to hit a word count, but to force the writing to earn every sentence it kept. Say what needs saying, cut what doesn’t, move to the next point. Concise wasn’t a constraint. It was the strategy.

The traffic grew month over month, then year over year, in a space where nobody hands you rankings. Organic traffic grew 435% – not from outspending the competition, from out-thinking them on what actually got published.

The other part of it was recognising that “content” didn’t mean “blog post.” Once the written pieces were live, we’d turn the article into a script, break it into short-form video, and get it in front of the audience where they were actually spending time – TikTok, YouTube Shorts. Pineapple’s audience skewed younger, newer drivers, and a lot of them weren’t reading a blog at all. They were scrolling. Meeting them there, in a format built for how they actually consume content, did more for the brand than another written piece would have.

None of that needed a bigger budget. It needed a strategy that knew what it was for.

Content Strategy vs. Content Marketing Strategy Isn’t a Real Split

People treat these as two separate disciplines. They shouldn’t be.

Content strategy is the angle – what you’re creating, the theme behind it, the platforms it belongs on. Content marketing strategy is how you get it in front of people – the distribution, the promotion, the repurposing. Split them and you get exactly what happens on most sites: brilliant articles nobody reads, sitting untouched three months after publish, because nobody planned past “hit publish.”

What’s the point of a well-researched, well-written piece if the only people who see it are the ones who happened to already be on your site that week? A strategy that stops at the content itself is half a plan. The other half – how it reaches someone, how it gets reused, how it keeps earning after day one – has to be baked in from the start, not bolted on afterwards as an afterthought nobody owns.

Repurposing: Turning One Piece Into Reach, Not Into More Content

Repurposing gets talked about like it means “post the same thing on five platforms.” That’s not what made it work for Pineapple.

The process was specific. Take the article, distil it into a script, turn that into one or two short-form videos – if the article makes two distinct points, that’s two videos, not one video trying to cram in both. If a single long-form video suits the topic better, build that instead. Then pull social posts out of the same source material, tying everything back to the original piece so someone can land on it however they found it – video first, post first, or the article itself.

That’s the whole point of repurposing done right: one piece of real thinking, redistributed into the formats where different parts of your audience already are, instead of writing five separate mediocre pieces that all say roughly the same thing. It’s not extra content. It’s the same content working harder.

Where AI Search and Query Fan-Out Actually Change the Plan

Query fan-out doesn’t make content strategy harder. It makes ignoring your audience more expensive.

For a long time, content calendars were built almost entirely off keyword volume – target this term, it gets this many searches, done. That’s where the thinking stopped. Now that people can get a synthesised, in-depth answer straight from ChatGPT or Claude without visiting five different sites, volume alone tells you less than it used to. What matters more is anticipating the actual question behind the question.

Someone searching “Nike running shoes” might mean trail running, might mean road running, might be comparing two specific models. Query fan-out is Google – and increasingly, AI search – trying to anticipate which of those they actually meant, and serving up an answer that covers the ground before the follow-up search happens. Bake those variations into one genuinely thorough piece and you’ve replaced three or four thin, near-duplicate pages with one that actually earns its ranking. It’s not about writing for an algorithm’s prediction. It’s about knowing your audience well enough that you’d have anticipated the same question anyway.

This is also where content becomes harder for competitors, and AI, to simply copy. Generic answers to “what is X” get replicated everywhere – AI has read every version already. What can’t be replicated is the version built on what you’ve actually seen happen, with a client, in a market, over real time. That’s not a defensive move. That’s the whole advantage.

The First 30 Days With No Budget and No Agency

Start with the audience, not the content.

Understand who you’re writing for and what’s actually keeping them up at night – the real pain points, not the ones that sound good in a strategy deck. Do the competitive research. See what’s already being said about the brands in your space, and more usefully, what isn’t being said – the questions nobody’s answering properly yet. Tools like Ahrefs and the “People Also Ask” data are enough to start seeing the shape of that gap without paying for anything more sophisticated.

Resist the pull toward the highest-volume, highest-difficulty terms because they look like where the real prize is. They’re not, not yet. Go after low-difficulty, lower-volume terms first – the ones you can actually win – and use them to prove the content is attracting people who convert, not just people who click. Build from there. Slowly, deliberately, one cluster at a time, the same way Pineapple did it.

That’s the whole starting point. Not a twelve-month roadmap. Just: know who you’re talking to, find what nobody’s answered well yet, and prove it works before you try to scale it.

The One Thing to Stop Worrying About

Volume.

You’ll look at a competitor with ten times your output and feel like you’re losing before you’ve started. Most of the time, you’re not. A lot of what they’re publishing has nothing to do with what their brand actually does – the same drift that cost HubSpot most of its traffic. Some of it is quietly cannibalising their own rankings, two or three pages fighting each other for the same search. You won’t see that from the outside. You’ll only see the volume, and volume looks like winning until you check what it’s actually doing.

You can say in one well-built piece what they’re saying across ten mediocre ones. That’s not a consolation prize for having less budget. That’s the actual strategy – find where their coverage is thin, say the thing properly once, and let it outlast the ten pieces built to hit a quota instead of a purpose.

Where to from here? Pick the one thing your competitors haven’t said properly yet, and go say it properly. That’s the whole game.

FAQ

What’s the difference between a content strategy and a content calendar? A content calendar is the schedule – what publishes, when. A content strategy is what sits above it: the objectives, the themes, and the reason each piece exists. A calendar without a strategy behind it is just a production schedule with no way to measure whether it’s working.

How often should a small business publish content to compete with bigger sites? Less than you think, and better than you’re currently managing. One or two genuinely strong pieces a month, properly researched and repurposed across formats, will usually outperform ten average ones. The goal is coverage of the gaps competitors have left, not matching their output.

Is AI-generated content bad for SEO in 2026? Unedited, mass-produced AI content gets caught and demoted – Google has shown it can detect low-effort, machine-written text at scale, and sites that lean on it hard tend to get hit when quality-focused updates roll through. AI as part of the process, with a person shaping the final piece, is a different thing entirely.

Should I focus on written content or video content for SEO? Go where your audience actually spends time, not where you’re most comfortable creating. If your audience is younger and scrolling TikTok or YouTube Shorts, written-only content misses them no matter how good it is. Test both at a small scale before committing either way.

How do I know if my content strategy is actually working? You need objectives set before you publish, not after. If you can’t say what a piece of content was trying to achieve, you can’t measure whether it did. Track whether traffic is converting, not just whether it’s arriving – quality traffic beats volume every time.

What is content repurposing and how do I start? Taking one well-built piece and reshaping it for the formats and platforms your audience actually uses – a script for short-form video, posts pulled from the same source material, all tied back to the original. Start with your strongest existing piece rather than your newest one.

My GEO cluster has ten pages in it right now. Two of them are trying to answer the same question.

/generative-engine-optimisation/ and /generative-engine-optimization-geo-the-next-layer-of-search-visibility/ sit on my own site, different URLs, different word counts, near-identical intent. I wrote them four months apart and didn’t notice the overlap until I went looking for it while writing this. There’s a second pair doing the same thing on LLM SEO: /llm-seo-a-search-practitioners-honest-guide/ and /how-to-optimize-for-llm-search-the-complete-2025-guide-to-geo-ai-visibility/. Two pillar pages, one topic, both competing for the same reader.

I’m telling you this before I tell you anything else because it’s the honest starting point. Every SEO content strategy guide you’ll find gives you the same nine steps: understand your audience, do the keyword research, map the content, optimise, measure, repeat. That framework isn’t wrong. Audience research matters. Keyword research matters. You need all of it.

What it doesn’t tell you is that you can follow every one of those steps and still end up with what I’ve got. A cluster with real depth and a real problem sitting inside it.

The checklist isn’t wrong, it’s incomplete

Go and read the top ten results for “SEO content strategy” right now. Siteimprove, HubSpot, seoClarity, half a dozen others. They all converge on the same shape: research your audience, find your keywords, build a content calendar, optimise on-page, track your metrics. Some stretch it to nine steps, some compress it to six, but it’s the same skeleton wearing different clothes.

I don’t disagree with any of it. If you skip audience research you’re writing for nobody. If you skip keyword research you’re guessing at demand. Those fundamentals aren’t going anywhere.

The problem is what the checklist assumes without saying it out loud. It assumes the reader’s journey starts on Google. It assumes one persona per keyword. It assumes a cluster is “done” once every page has a home in the sitemap. None of those assumptions hold the way they did five years ago, and a strategy built only on the checklist misses the parts of the job that actually decide whether the cluster works.

The journey doesn’t start on Google anymore

Someone researching your topic today might never touch a search results page before they’ve already formed an opinion about you. They ask ChatGPT. They ask Claude or Perplexity. The tool reads across the web, synthesises an answer, and hands them something that feels complete. By the time they do search on Google, if they search at all, they’re often searching your brand name, not the generic topic.

That’s what search everywhere optimisation actually means in practice. Not “optimise for five platforms instead of one.” It means your content has to hold up wherever it gets encountered, including inside an AI answer that fans your topic out into six sub-questions you never wrote a heading for. A cluster built purely for Google’s crawler misses that a chatbot doesn’t crawl your site the way Googlebot does. It pulls a chunk, synthesises it into an answer, and moves on. If your cluster only makes sense as a whole document read top to bottom, it doesn’t survive that process well.

I’ve written about how query fan-out actually reshapes this in more depth elsewhere on the site, but the short version for cluster planning is this: your pillar page and its supporting content need to work as standalone, extractable answers, not just as chapters in a book only Google reads start to finish.

Topics, not keywords, and not one persona either

Here’s a gap that shows up early in most cluster plans. Someone maps “project management software” to a handful of keyword variants: best project management software, project management software for small business, project management software comparison. Reasonable enough on paper.

But a senior manager evaluating that software and a junior team member trying to figure out how to use it after it’s already been bought are two completely different readers. One wants a comparison built around ROI, team size, and integration risk. The other wants a walkthrough. Same rough topic, same product, entirely different job to be done.

A cluster built keyword-first tends to flatten that difference because it’s optimising for search volume, not for the actual range of people typing something close to that phrase. Building topic-first instead means starting with “who is asking this and why” before you start with “what phrase do they type.” The keyword research still happens. It just stops being the first decision and becomes the second one.

Does the cluster actually cover the funnel?

This is where most cluster diagrams look better than the clusters actually are. The pillar page exists. A handful of spokes exist. Someone’s drawn the hub-and-spoke picture in a deck and called it a strategy.

What that picture usually doesn’t show is whether the cluster covers the reader at every stage of actually deciding something. Top of funnel: what is this, why does it matter. Middle: how do the options compare, what’s the real trade-off. Bottom: how does this specific thing solve my specific problem, and can I trust the answer.

The gap I see most often sits at the bottom. Cluster plans are generous with explainer content and comparison content, then thin out right where the reader is closest to acting. If you’ve built or bought a tool, that’s the moment to explain what it actually does and how it solves the reader’s problem, without turning it into a sales page. Not “book a demo.” What does it do, step by step, and why does that specific mechanic matter for the problem the reader showed up with. That page is harder to write than a TOFU explainer because you can’t just summarise what’s already been said everywhere else. It has to come from having actually built or used the thing.

That’s also, not coincidentally, the content that’s hardest to replicate. Generic “what is X” content is well covered territory. An AI model has seen a hundred versions of it and can produce a hundred and first without much trouble. A page built on your own hands-on experience with a specific product, backed by what you actually saw happen when you used it, is a different kind of asset. It can be cited. It can’t be duplicated.

Intent cannibalisation, not keyword cannibalisation

This is where I’ll go back to my own cluster, because it’s the clearest way to explain the distinction that most cannibalisation advice misses.

The usual warning is: don’t have two pages targeting the same keyword. Fair, as far as it goes. But two pages can share a keyword and coexist fine if they’re answering different questions. A “what is generative engine optimization” page and a “generative engine optimization services” page can both rank for variations of “generative engine optimization” without stepping on each other, because one is explaining a concept and the other is selling an engagement. Different intent, different reader need, same rough phrase.

The real problem is intent cannibalisation: two pages trying to answer the exact same underlying question, for the exact same reader, regardless of what keywords each one happens to target. That’s harder to spot because it doesn’t show up as an obvious keyword clash. It shows up as two pages quietly splitting the same traffic and the same authority that one stronger page would have carried on its own.

I let my own cluster get away from me, here’s the fix

So back to my ten pages. /generative-engine-optimisation/ and /generative-engine-optimization-geo-the-next-layer-of-search-visibility/. Run the test: is a reader landing on either page trying to answer a genuinely different question? No. Both are explaining what GEO is and why it matters. Different titles, different word counts, same job. That’s intent cannibalisation, not a coexistence case.

Same story with /llm-seo-a-search-practitioners-honest-guide/ and /how-to-optimize-for-llm-search-the-complete-2025-guide-to-geo-ai-visibility/. Both are trying to be the definitive LLM SEO explainer. Neither one is doing anything the other isn’t already doing.

What I’d actually do here, and what I’m doing after this goes live: pick the stronger page in each pair, fold whatever’s genuinely additive from the weaker one into it, and 301 redirect the loser. Not because having two pages is inherently wrong. Because these four specifically aren’t serving different readers, they’re just splitting one audience across two URLs and asking Google to guess which one matters more. I’d rather make that decision myself than leave it to chance.

If you’re auditing your own cluster, the question to ask about every pair of pages that feels a little too close is simple: if I put these in front of the same reader on the same day, are they solving different problems, or just repeating each other in a different voice. If it’s the second one, you don’t have two pages. You have one page that isn’t sure what it wants to be yet.

Content isn’t just articles anymore

Google made a change in July 2026 that’s worth paying attention to here. Search Console added platform properties covering Instagram, TikTok, X, and YouTube, letting site owners see clicks, impressions, and top-performing posts for content on those platforms the same way you’d view a normal site property. That’s Google formally telling you it’s tracking multi-format visibility as part of the same measurement layer as your written pages, not as a separate world.

Attention spans are shorter than they were even two years ago. Trust in written content took a hit as unedited, ungrounded AI output flooded feeds, and readers got sharper at spotting it. Video, in both short and long form, has become a real part of how people search and how Google surfaces answers. So has structured content: tables people can scan, infographics that compress a comparison into one image, formats that don’t ask for the same sustained attention a 2,000-word article does.

None of that means every cluster spoke needs to become a video. It means format has to be chosen the same way audience is chosen, deliberately, based on what the specific reader at that specific stage actually wants and how they’re actually looking for it. A comparison page might work better as a table than a paragraph. A “how it works” explanation might land harder as a two-minute video than a wall of text. There’s no single format that wins across every persona and every funnel stage, and treating video or tables or infographics as a mandatory checkbox misses the point as badly as ignoring them entirely.

Building the cluster: a working structure

Put the pieces together and the shape looks like this. A pillar page anchors the topic and routes into spokes built around real funnel stages, not just related keywords. Each spoke is checked against the intent test before it gets built, so you’re not writing a fourth version of the same explainer under a new title. Internal links move the reader (and the crawler, and whatever’s fanning your content out into an AI answer) from broad to specific, TOFU to BOFU, in a path that actually makes sense to follow. Format gets chosen per page based on the reader at that stage, not applied uniformly across the whole cluster.

And once it’s built, it gets checked again. Not just for keyword rankings, but for the two questions that actually matter: does this cluster cover the full range of people searching this topic, from the senior manager to the junior researcher, and is every page in it still earning its place, or has something started quietly competing with something else.

Mine didn’t pass that check when I ran it for this piece. Worth running yours too.

FAQ

How many pages should a content cluster have?

There’s no fixed number. What matters is coverage, not count. A cluster with four pages that map cleanly to different stages of the reader’s journey beats a cluster with fifteen pages where half of them are answering the same question in different words.

What’s the difference between a pillar page and a cluster page?

A pillar page covers the topic broadly and links out to more specific supporting pages. Cluster pages (sometimes called spoke pages) go deep on one angle, one funnel stage, or one sub-question the pillar only touches on. The pillar is the map, the cluster pages are the destinations.

How do I know if I have keyword cannibalisation or intent cannibalisation?

Check the keyword first. If two pages rank for the same phrase but answer different questions for different readers, that’s coexistence, not a problem. If two pages answer the same underlying question for the same reader, regardless of which exact phrases each one targets, that’s intent cannibalisation, and it’s the more common issue in mature clusters.

Does a content cluster actually help with AI search visibility?

Structure helps because AI models fan a topic out into sub-questions and synthesise answers from whichever pages best answer each one. A cluster that’s mapped to real sub-questions, with each page written to stand on its own, gives an AI system more clean, extractable answers to pull from than one long page trying to cover everything at once.

Should every cluster include video content?

Only where it fits how that specific audience searches and consumes information. Google’s July 2026 Search Console update, which added performance reporting for Instagram, TikTok, X, and YouTube content, shows video and social formats are part of the same visibility picture as written pages now. That’s a reason to consider it, not a reason to force it into every spoke.

How often should a content cluster be reviewed?

At minimum, whenever you add a new page to it. That’s the moment to check whether the new page duplicates an existing one in intent, not just in keyword. Beyond that, a full review every few months catches drift before it turns into the kind of pair I found in my own GEO cluster.

Most on-page SEO checklists get you 80% of the way there, then quietly skip the part that actually sinks people. You can tick every box – title tag, meta description, alt text, keyword in the first paragraph – and still watch an article fail, because the checklist never asks the one question that matters most: can a person actually navigate what you built?

I had a client whose blog section was, functionally, an HTML sitemap. Articles and glossary terms jammed into one list, no hierarchy, nothing telling a visitor or a crawler what belonged where. Every individual page could have passed a checklist audit. The section as a whole made no sense. That’s the gap this guide is built to close – the on-page work that lives on the page, and the structure around it that decides whether anyone ever reads that far.

Key Takeaways

• On-page SEO covers everything on the page itself – structure, keywords, headings, links, content formats – and it fails the moment the section around it has no navigable structure.
• The basics (title tags, meta descriptions, alt text, page speed) are still the fastest, most valuable wins available, even in 2026.
• E-E-A-T isn’t a background trust score. It’s something you build into the page itself – bylines, cited sources, reviewer credit, original data.
• Content formats like tables and chunked sections help human readers and AI engines the same way, for the same reason. Neither should be built “for the algorithm” alone.
• URLs, schema, and page speed each deserve their own attention – they’re not interchangeable checklist items, and each has its own rules.
• Real click-through data shows a massive spread at every single ranking position – proof that on-page work, not position alone, decides whether people actually click.
• If rankings are climbing but engagement is falling, on-page SEO was never your problem. Navigation was.

What On-Page SEO Actually Means

On-page SEO is everything that happens on the page itself: layout, keyword usage, headings, internal links, and whether the person reading it actually gets an answer that makes sense. That last part is the one every checklist assumes rather than checks.

It’s tempting to treat on-page and AI-engine optimization as separate jobs, one for Google and one for ChatGPT. They’re not. A page structured clearly enough for a human to skim in ten seconds is also structured clearly enough for an AI system to extract an answer from – the same argument I make at length in why answer engine optimization builds on SEO rather than around it. This guide doesn’t split those into two tracks. Every section below is written for both, because the work is the same work.

Where On-Page SEO Actually Breaks

A client of mine ran their glossary and blog content as one undifferentiated list. No separation, no hierarchy – open the blog homepage and you got a wall of links that read more like a sitemap than a section a person would want to browse. Every individual glossary entry and article, taken on its own, might have looked fine.

We rebuilt it properly: a real blog section, a real glossary section, each following a conventional structure a visitor would recognize instantly. Alongside that, we developed the topic clusters underneath both sections and did the on-page work page by page. The result was rankings for terms they hadn’t ranked for before – not from one fix, but because the architecture and the on-page work compounded together. People could finally navigate what was already there, and so could the crawlers indexing it.

That’s the argument underneath this whole guide: on-page SEO isn’t just what’s on one page. It’s whether the section that page lives in makes sense.

Content Architecture and Internal Linking

Before you touch a single title tag, the section your content lives in needs a shape a visitor recognizes: clear categories, a logical hierarchy, and internal links that connect related pieces instead of leaving them stranded.

It’s the least glamorous part of on-page SEO and, in my experience, one of the most consequential. Internal links are entirely within your control – unlike backlinks, which depend on someone else’s decision – and a well-linked cluster helps both a reader who wants to go deeper and a crawler trying to understand how your content relates to itself. Group related articles under a clear parent structure, link between them naturally in the body text, and make sure your navigation reflects the same hierarchy your content actually has.

Title Tags: Structure and Examples

Few fixes are faster wins than a properly written title tag, and I mean that literally – a rewritten title tag can move click-through rate within days, not months.

The formula I test with most often: primary keyword and modifier first, brand name last, if there’s room. Title tags have a limited character budget – keep them under roughly 60 characters so they don’t get cut off in the results – and when space is tight, the brand name is the first thing to drop. What matters more is that the title genuinely describes what the page delivers.

For this exact article, that formula plays out a few different ways:

• On-Page SEO: The Complete Guide to Optimizing Every Page
• On-Page SEO: The Ultimate Guide | Gus van der Walt
• What Is On-Page SEO? A Practical Guide

Notice the pattern: the primary keyword (“on-page SEO”) is present in every version, but it doesn’t have to sit at the very start. A modifier word – “complete,” “ultimate,” “practical” – helps signal depth and can lift click-through rate on its own, and a year or date can do the same when freshness matters to the searcher. The brand name only appears when there’s character budget left over to spend on it.

If the page is answering a direct question, a question-format title works too, as long as the keyword is in there naturally: What Is On-Page SEO, and Why Does It Matter? The rule I keep coming back to: does this read naturally to a person scanning results, and does it answer what they were actually searching for? If yes, the exact placement of the keyword inside the title matters far less than people assume.

Meta Descriptions: Structure and Examples

A meta description has one job: give someone a reason to click, in roughly 150 characters or less, without giving away the entire answer.

The structure I use: state the question the page answers, then give a compressed version of the answer, in two sentences. Something like:

“On-page SEO isn’t just tags and keywords – it’s whether your content is structured well enough for readers and AI engines to actually use it. Here’s the full breakdown.”

Or, for a more direct variant:

“A complete, example-led guide to on-page SEO: title tags, headings, schema, page speed, and the structure that ties it all together.”

Both versions name the topic, hint at what’s inside, and stop short of answering the question completely – because if the snippet already tells someone everything they need, there’s no reason left to click through. Strike that balance deliberately: enough substance to prove the page is worth the click, not so much that the click becomes optional.

Headings and Content Structure

Here’s the structure I use for every long-form piece, including this one – not as a rigid template, but because it consistently works for readers and for extraction. This article is itself the worked example: notice the key takeaways section near the top, the table of contents directly underneath it mirroring every heading below, and each H2 covering one distinct idea with its own H3s where the topic needs further breakdown.

If you stripped this pattern down to a generic outline, it looks like this:

H1: [Primary Keyword]: [What the Page Delivers]

Key Takeaways

 – 3-5 bullet points summarizing the whole article

Table of Contents

 – Mirrors every H2 below, in order

H2: [First major subtopic]

 H3: [Supporting detail]

 H3: [Supporting detail]

H2: [Second major subtopic]

 H3: [Supporting detail]

H2: [Third major subtopic]

Conclusion / Next Steps

 – What to do now, and where to read next

Use your H1 once. Follow it with an opening that either states the core point directly or opens with a short story, then gives the reader the key takeaways, because plenty of readers won’t make it past the first screen. From there, break the body into H2 sections, each covering one distinct part of the topic, with H3s and deeper subheadings nested underneath where a section needs it. Use bullet points and short paragraphs inside each section – readers skim before they commit, and so does an AI system pulling an answer out of your page. Close with a conclusion or next-steps section: what the reader should do now, and where to go for more.

This isn’t decoration. It’s the difference between an article someone reads to the end and one they bounce off after the first paragraph.

Keywords and Intent Inside the Page

Keyword placement still matters – in your H1, your opening paragraph, a subheading or two, your URL slug – but the framing has shifted from “hit this term five times” to “does this page clearly serve the intent behind the search.” That shift didn’t happen overnight; it tracks the same thirty-year arc I walk through in the history of SEO, from keyword stuffing to something closer to topic modeling. Write naturally, place your target term where it reads correctly, and don’t force it into every paragraph. If a page covers a topic properly, the related terms show up without you engineering them in.

The Data: How On-Page Work Actually Moves CTR

Everything above is argued from experience. Here’s the same argument from data – an aggregate analysis across a working keyword set spanning multiple industries, several hundred keywords deep, tracked for search volume, ranking position, and real click-through rate over time.

Volume and intent don’t move together the way people assume. Across the tracked set, informational keywords carry the most search volume by a wide margin – an average of roughly 14,700 searches a month, against about 5,550 for commercial-intent terms and 4,550 for transactional ones. But transactional keywords are the hardest to rank for, averaging a keyword difficulty score nearly seven points higher than informational or commercial terms carrying similar or greater volume. In plain terms: the biggest search demand sits around questions and research, not purchases, and the terms closest to a sale are also the most competitive relative to what they’re worth in volume. That’s the numeric version of the point made earlier in this guide – intent has to drive your keyword targeting, because volume alone will point you at the wrong terms.

Intent	Avg. monthly volume	Avg. keyword difficulty
Informational	~14,671	30.8
Commercial	~5,552	30.7
Transactional	~4,552	37.7

Ranking position explains far less of your click-through rate than people assume – and that gap is where on-page work lives. Looking at real click-through data for informational content at each ranking position, the spread is enormous. At position 1, average CTR sits around 42%, but individual pages ranged all the way from under 0.1% to over 92%. At position 3, the average drops to roughly 17%, with a range stretching from near-zero to 88%. Even at position 5, where average CTR falls to under 8%, some individual pages were still pulling in CTR above 75%.

Position	Average CTR	Observed range
1	~42.4%	0.05% – 92.6%
2	~19.7%	0.14% – 88.4%
3	~16.9%	0.05% – 88.1%
5	~7.8%	0.01% – 76.7%
10	~2.1%	0.00% – 42.7%

Two pages can sit at the exact same position and see wildly different click-through rates. Position gets you in front of someone. What happens next – whether your title tag and meta description actually earn the click – is entirely on-page work. The food blogger case study earlier in this guide, where basic on-page fixes moved click-through rate from roughly 1% to 3%, is a real result, and this data says there was still considerably more room in that number. A threefold CTR lift from adding the basics is good. The ceiling, based on what’s actually achievable at the same ranking positions elsewhere, is considerably higher.

E-E-A-T as an On-Page Discipline

Most guides treat E-E-A-T as something that happens in the background – a trust score Google calculates from signals you don’t directly control. I think that’s a mistake. E-E-A-T is something you build into the page itself, and it belongs in this checklist, not off to the side of it.

Concretely, that means: an author byline naming who wrote the piece and what their experience actually is, a note on who reviewed it, quotes or citations from people with real expertise in the topic, and – where you have it – original data or research your own work produced rather than borrowed. When you cite another source, credit it properly and link to it; when you reference an industry publication or an expert’s take, name them. Every one of those signals tells a reader, and increasingly an AI system summarizing your page, that this wasn’t produced from nothing – the same groundwork that determines whether you ever show up in a ChatGPT answer at all, covered in more technical depth in my guide to getting cited in ChatGPT responses.

For a landing page rather than an article, the same principle shows up as reviews, ratings, third-party trust marks (a Google rating, a Trustpilot score, a G2 listing if it’s relevant to your industry), case studies, and testimonials. None of it should be an afterthought bolted onto a finished page. Build the piece with the question already in mind: what evidence can I bring in here that proves this wasn’t just thumbsucked?

Content Formats for Humans and AI

Tables, images, video, and modular chunking all help a page get parsed by an AI system and read by a human, for exactly the same reason: both are trying to extract a clear answer as quickly as possible. Research on AI citation patterns backs this up directly – content that’s structured with clear, self-contained sections and entity-dense language gets cited more often, because the system doing the summarizing can find the answer without having to infer it. I go deeper on chunk design and retrieval specifically in my honest guide to LLM SEO, if you want the mechanics behind why this works.

Where this goes wrong is when formats get added for their own sake. A table only earns its place if it’s genuinely comparing something – specs, prices, before-and-after figures. An article about running shoes stuffed with tables nobody asked for won’t get read by a human, and forcing structure that doesn’t fit the content won’t fool an AI system either, because the underlying information still has to make sense.

My rule: look at what’s already ranking well for a topic, see what structure they’re using, and test variations against your own audience – an added table here, a video there – before assuming a format works. If your readers engage well with a format, it tends to translate to AI visibility too, because both audiences are responding to the same underlying clarity.

Internal and External Links, Done Right

Internal links should connect a reader to something genuinely useful, not just pad out a word count. Use descriptive anchor text – “content architecture,” not “click here” – and link from your strongest, most-visited pages down to newer content that needs the visibility.

External links work the same way in reverse: cite the sources you’re actually drawing on. If you’re writing about SEO, that might mean referencing Search Engine Journal, Search Engine Land, or a specific practitioner’s research on LinkedIn – and crediting them properly when you do. This isn’t just good practice. It’s part of the E-E-A-T argument above: showing your work is itself a signal.

Image Alt Text and Accessibility

Alt text describes an image for anyone using a screen reader and for any system that can’t render the image directly – which, increasingly, includes AI crawlers building an understanding of your page. Write it plainly: describe what’s actually in the image, skip “image of” or “picture of,” and keep it under roughly 125 characters.

This isn’t a minor accessibility checkbox tacked onto SEO. Treat it as its own discipline that on-page SEO happens to benefit from. A site that’s genuinely accessible – proper alt text, sensible heading order, readable contrast – serves every visitor, and search engines reward exactly the same signals they’re built to detect. It’s also part of what determines whether a page gets pulled into an AI Overview in the first place, which I break down separately in how to actually show up in AI Overviews.

URL Slugs: Structure and Examples

Keep URL slugs short, readable, and free of anything that doesn’t earn its place. Drop stop words – “a,” “the,” “and” – and stick to three to five words that describe the page.

A clean, simple structure looks like this:

gusvanderwalt.com/blog/on-page-seo-complete-guide

Not this:

gusvanderwalt.com/2026/07/27/the-ultimate-and-complete-guide-to-on-page-seo-for-beginners-and-experts

If you want your URLs to reflect where content sits in your site’s hierarchy, a category-based pattern works well too:

gusvanderwalt.com/blog/technical-seo/on-page-seo-complete-guide

That’s a genuinely useful signal for search engines, but it isn’t mandatory – if you implement breadcrumbs on the page itself (home > category > article), you get the same hierarchy signal through navigation without needing it baked into every URL. Either approach works. What doesn’t work is a slug so long it stops being readable at a glance, or one stuffed with dates and parameters that add nothing for a reader trying to guess what the page is about before they click.

Schema Markup: Structure and Examples

Schema is use-case specific – there’s no single schema type that covers every page, and adding schema that doesn’t match your content does more harm than good.

For a standard blog article, the baseline is Article or BlogPosting schema plus BreadcrumbList, so search engines understand both what the page is and where it sits in your site:

{

  “@context”: “https://schema.org”,

  “@type”: “BreadcrumbList”,

  “itemListElement”: [

    { “@type”: “ListItem”, “position”: 1, “name”: “Home”, “item”: “https://gusvanderwalt.com/” },

    { “@type”: “ListItem”, “position”: 2, “name”: “Blog”, “item”: “https://gusvanderwalt.com/blog/” },

    { “@type”: “ListItem”, “position”: 3, “name”: “On-Page SEO: The Complete Guide” }

  ]

}

If the page includes a genuine FAQ section, FAQPage schema is still worth adding – with one caveat worth knowing. Google deprecated FAQ rich results from search on May 7, 2026, meaning the expandable Q&A dropdown no longer shows up in Google’s results the way it used to. The schema type itself hasn’t been removed, though, and other crawlers – including the retrieval systems behind AI search tools – can still parse it. Keep the schema if your FAQ content is real and answers questions visible on the page. Don’t add it purely chasing a rich result that no longer exists.

For a landing page, swap Article for Service or Product schema depending on what the page is selling, and layer in Organization schema so search engines and AI systems can tie the page back to your brand as an entity – the same entity-relationship groundwork covered in more depth in my complete guide to generative engine optimization:

{

  “@context”: “https://schema.org”,

  “@type”: “Organization”,

  “name”: “Gus van der Walt”,

  “url”: “https://gusvanderwalt.com”,

  “sameAs”: [

    “https://www.linkedin.com/in/gusvanderwalt/”,

    “https://www.instagram.com/gusvanderwalt/”

  ]

}

The pattern to remember: Article/BlogPosting and Breadcrumb as your baseline for content, FAQPage when the content genuinely earns it, Service/Product for commercial pages, and Organization to tie everything back to a recognizable entity.

Page Speed: What Actually Moves the Needle

Page speed is on-page SEO in the sense that it’s entirely within your control on a page-by-page basis, even though it touches technical SEO too. A handful of things move the needle more than the rest:

• Lazy load what isn’t immediately visible. Fonts, images further down the page, and embedded video shouldn’t load before the content someone actually landed to read.
• Compress and format images properly. The food blogger case study above is proof of this directly – some of those images were well over a megabyte each, and compressing them improved both load time and, downstream, click-through rate. Use modern formats like WebP or AVIF over older JPEGs where your CMS supports it; they hold visual quality at a fraction of the file size.
• Keep related-post and thumbnail images genuinely small. A featured image sized for a hero banner has no business being reused, full-size, as a thumbnail three rows down the page.
• Watch video and embed weight. A single heavy video embed can undo every other speed optimization on the page – keep the number of resources a page has to load in check, especially anything auto-playing or auto-loading above the fold.

None of this is exotic. It’s the unglamorous, compounding work that shows up in both your Core Web Vitals and in how long someone’s willing to wait before bouncing.

Using AI for On-Page SEO Without Losing the Human Touch

AI can genuinely help with on-page work, but it has to be a collaborative process, not a replacement for judgment. This article is an example of what that looks like in practice: I answered a structured set of interview questions, drove the direction of the piece, and the research and drafting came together around what I actually said – rather than the other way around.

If you’re using AI tools for on-page work, connect them to real data rather than letting them guess – keyword and SERP data from a tool like Ahrefs or Semrush, and your own Search Console data, so what gets produced is grounded rather than generic. Set guardrails, test the output, and don’t take what an AI tool produces at face value – I’ve written up the fuller version of what that setup actually looks like in AI for SEO: how to use it without losing your rankings.

Here’s a test I use: if you swapped out your brand name and your specific details, would the article still say anything real, or would it read exactly the same for any competitor in your space? If it’s the second one, you haven’t produced anything a reader has a reason to trust, no matter how well-structured it looks on paper.

How to Tell When On-Page SEO Was Never the Problem

If a page is ranking reasonably well but bounce rate is high and engagement time is low in GA4, your on-page SEO didn’t fail – your content’s ability to hold a reader did, and that has a way of catching up with your rankings anyway. A high bounce-back tells search engines the page didn’t solve what the searcher actually needed, and pages that get sent back quickly tend to slide down the results over time, even if they climbed there initially.

Tools like Microsoft Clarity make this easier to diagnose directly: session recordings, rage clicks, and dead clicks show you exactly where someone gave up, rather than leaving you to guess from the rankings graph alone. If the pattern shows people arriving, scrolling briefly, and leaving, the fix isn’t another keyword or another paragraph. It’s the structure – the thing this whole guide has been building toward.

The On-Page SEO Checklist

• H1 used once, containing your primary keyword
• Title tag under 60 characters, keyword and modifier first, brand last if there’s room
• Meta description under 150 characters, question plus a partial answer
• Key takeaways section near the top, for readers who won’t make it to the end
• Table of contents mirroring your heading structure
• Clear H2/H3 hierarchy, one topic shift per H2
• Bullet points and short paragraphs used deliberately for scannability
• Author byline naming who wrote the piece and their relevant experience
• Cited sources and expert quotes, credited properly
• Tables, images, or video used only where they genuinely add clarity
• Descriptive internal links connecting to related content
• External links to sources that actually back up your claims
• Alt text on every meaningful image, under roughly 125 characters
• Short, readable URL slug, three to five words, no stop words
• Article/BlogPosting and Breadcrumb schema as the baseline; FAQPage, Service, Product, or Organization added by use case
• Images compressed and served in a modern format (WebP or AVIF)
• Lazy loading on fonts, below-the-fold images, and embedded video
• A conclusion or next-steps section pointing the reader somewhere useful

FAQ

Is on-page SEO still worth doing if AI search is taking over? Yes. AI systems still draw heavily on the same structural and content signals that traditional on-page SEO targets – clear headings, well-organized sections, and genuine expertise on the page. Strong on-page work benefits both.

What’s the single most valuable on-page fix for a site that’s never done any SEO? Title tags and meta descriptions, in most cases. They’re fast to implement and directly affect click-through rate, which compounds with whatever ranking position you already have.

How is on-page SEO different from technical SEO? Technical SEO is about whether a site can be crawled and indexed at all – architecture, speed, and site-wide structure. On-page SEO is about the content and structure of individual pages once they’re accessible. They overlap, particularly around page speed and site architecture, but they’re answering different questions.

Does FAQ schema still matter if Google removed FAQ rich results? The rich result (the expandable dropdown in Google’s results) is gone as of May 2026, but the schema markup itself is still valid and can still be parsed by other crawlers, including AI search systems. Keep it if your FAQ content is genuine; don’t add it purely chasing a rich result that no longer exists.

Do tables and structured content actually help with AI search visibility? Generally, yes, when they’re used for genuine comparisons or breakdowns rather than added for their own sake. The same clarity that helps a table make sense to a reader tends to help an AI system extract information from it.

How do I know if my on-page SEO is actually working? Track rankings, but also track engagement – bounce rate, time on page, and scroll depth in an analytics tool. Rankings without engagement usually mean the content isn’t holding the people it’s attracting, which tends to show up in your rankings eventually anyway.

Can I use AI tools to do my on-page SEO for me? You can use them to speed up parts of the process, but treat it as a collaboration, not a replacement. Connect any AI tool to real keyword and performance data, and test whether the output would still make sense with your brand name removed – if it wouldn’t, it hasn’t said anything real.

Google “what is SEO” today and you’ll get three different answers depending on which corner of the internet you land in. One site tells you it’s technical, on-page, and off-page – full stop, textbook definition, moving on. Another tells you that’s outdated, that you now need GEO for AI search and AEO for answer engines, three separate disciplines with three separate playbooks. A third tries to split the difference and gives you a glossary entry with links to six other glossary entries.

None of them are lying to you. They’re just each holding one piece of something that stopped being simple a while ago.

Here’s my answer, the one I’d give you at a coffee shop instead of in a pitch deck: SEO is optimizing your page, your content, your app – whatever you’ve built – so the systems people use to find things can find it, and so the people on the other end of that search actually get what they came for. Google, Bing, ChatGPT, Perplexity, the app store search bar, YouTube’s algorithm. All of it. I call it search everywhere optimization, because that’s closer to how people actually search now than any of the three-letter acronyms floating around.

What SEO Actually Is

Fifteen years in, here’s what I’ve stopped believing needs a subcategory: getting found and getting understood are the same job, whichever engine is doing the reading.

Search Engine Land calls this “search everywhere optimization” too, in their own guide to the topic – it’s not just my framing, it’s a name the industry is groping toward because the old one stopped covering the territory. A person shopping for running shoes might start on TikTok, get a brand recommendation from ChatGPT, then type that brand into Google to check reviews before buying. That’s one search journey across three platforms. Treating it as three separate optimization problems – one for TikTok, one for AI answers, one for Google – misses that every one of those touchpoints is being shaped by the same underlying work: is your information structured so a machine can parse it, and does it actually solve the problem the person showed up with?

That’s the sentence I want you to leave with if you read nothing else here: SEO isn’t dying, and it isn’t being replaced by GEO or AEO. It’s the same discipline, doing more jobs than it used to, because the places people search multiplied and the work compounds across all of them.

How Search Engines (and AI Engines) Actually Find and Rank Things?

Strip away the jargon and every search system – Google, Bing, ChatGPT’s retrieval layer – runs the same three-step process underneath.

Crawling. A bot visits your site and follows the links it finds, the same way you’d follow a trail of breadcrumbs. If your site’s structure doesn’t give it a clear trail, whole sections of your content might as well not exist.

Indexing. Whatever the crawler found gets logged in a database, alongside what the content is actually about.

Ranking. When someone searches, an algorithm sorts through everything in the index and decides what to show, in what order, based on relevance and a few hundred other signals nobody outside Google has the full list of.

The part people miss with AI search specifically: Ahrefs’ own research on ChatGPT citations found that 88% of the URLs ChatGPT cites come from its underlying general search index – meaning the content still has to earn a place in a traditional search index before an AI assistant will ever mention it. Answer engines aren’t a parallel universe. They’re mostly reading off the same map, then summarizing what they find.

The Three-Part Map: Technical, On-Page, Off-Page

Ten years in, this part of the map hasn’t moved. What’s changed is what falls under each heading.

Technical SEO

Can the machines – search engines and AI crawlers both – actually get into your site and make sense of it? Site architecture, crawlability, page speed, and now: whether you’ve got an llms.txt file and markdown-friendly formats sitting alongside your regular HTML, because AI crawlers increasingly prefer to ingest a clean, structured version of your content rather than parsing a page full of navigation and ads.

I had a SaaS client whose product was genuinely good, but the site had no real structure behind it. Content wasn’t organized into anything resembling clusters, and the internal linking was close to non-existent – pages just floated, disconnected from each other. We went cluster by cluster: full technical audit first, then systematically optimized every page’s on-page elements, then built out internal linking properly – navigation, in-body links, the works. That systematic approach – fix the foundation, then link it together – got them a 190% increase in site traffic. Not from one clever trick. From doing the boring structural work properly and in sequence.

On-Page SEO

The page itself: layout, keyword usage, internal links, and whether someone reading it actually gets an answer that makes sense. I worked with an HR company once whose site was performing well, but their blog section was a mess – a glossary and articles all jumbled together with no real hierarchy, so the blog homepage read like a dumped list of links rather than something a visitor could navigate. We restructured it properly, and it had a real, measurable positive impact on their SEO. Nobody wants to read a glossary that looks like a junk drawer, and neither does a crawler trying to understand what the page is for.

Off-Page SEO

What happens away from your site: backlinks, brand mentions, whether your content gets shared or cited elsewhere. If ten years ago the industry was obsessed with backlinks almost to the exclusion of everything else, the correction I’d make for anyone starting today is: internal linking is the boring cousin nobody wants to work on, and it’s one of the most consequential things you can actually control yourself. Backlinks depend on someone else deciding to link to you. Internal linking depends on you deciding to do the work.

Where GEO and AEO Actually Fit

Here’s where I think the industry overcomplicates things, and where I want to be specific rather than vague about it.

GEO and AEO aren’t a fourth pillar bolted onto the three above. They’re the same three pillars, doing an additional job. Technical SEO already covered “can the machine read this” – now that includes llms.txt files and markdown formats for LLMs, alongside the traditional crawl signals. On-page SEO already covered “does this page make sense to a reader” – now that includes modular structure, tables, and shorter, more scannable sections, because AI engines are looking for concise, extractable answers just as much as human readers are. Off-page already covered “does the wider internet vouch for this” – now that includes whether your brand shows up consistently across the third-party content an AI assistant might draw from when it’s deciding who to cite.

Splitting these into separate disciplines with separate tactics misses that the work compounds. A well-structured, table-heavy, clearly-headed article helps a human skim it and helps an AI system extract an answer from it. It’s not two jobs. It’s one job that happens to serve two audiences at once.

Where I’d push back on some of the current GEO advice: there’s a real push toward heavy entity density and rigidly declarative language, and Kevin Indig’s research (cited via Ahrefs) found that heavily-cited AI content runs at roughly 20% entity density against 5-8% in standard prose, and uses definitive phrasing like “is defined as” nearly twice as often as content that doesn’t get cited. That’s real signal. But I’ve also seen a counter-example float through the industry – a small test where stacking in more stats, quotes, and confident tone actually dropped one page’s citation rate on a specific model. I don’t think that cancels out the entity-density finding. I think it’s a warning against treating entities as a checkbox instead of a structure.

Take Nike. Nike is an entity. Running shoes is a related entity. Usain Bolt, as a brand ambassador, is another entity, and the relationship between them – Bolt endorses Nike, Nike makes a specific shoe he’s associated with – is what an AI system is trying to map when it decides what to cite for a “best running shoes” query. You don’t need to force that relationship into a single sentence like a court deposition. You need it to exist naturally across your site and in how other people talk about your brand. Force it, and you’ve built something that reads like keyword stuffing wearing a new outfit. AI engines might still parse it. Google won’t reward it, and your actual readers will feel the seams.

How We Got Here

None of this appeared overnight. SEO has gone through PageRank, the Panda and Penguin quality crackdowns, mobile-first indexing, and now the shift toward AI-mediated search – each one forcing a genuine rebuild of what “doing SEO well” actually meant at the time. I’ve written the fuller version of that story, three decades condensed into the moments that actually mattered, in The History of SEO – worth reading if you want the “why” behind why the three-part map keeps holding even as the tactics inside it change.

The Specialisms

Once you’ve got the three-part map down, SEO branches into specialisms depending on what you’re optimizing for:

• Local SEO – visibility in map packs and location-based results, built around your Google Business Profile, reviews, and local citations.
• Ecommerce SEO – category and product page structure, faceted navigation, and the internal linking patterns specific to a storefront.
• Enterprise SEO – the same fundamentals at a scale where a single change might touch a million pages and three teams need to sign off before it ships.
• Technical SEO audits – a dedicated, periodic deep-dive into crawlability, indexation, and site health, distinct from the ongoing technical work described above.

And then there’s the layer this whole article has been circling: Generative Engine Optimization, Answer Engine Optimization, and what I’d call practical LLM SEO – each one a deeper look at a specific slice of the AI-search extension to the three-part map. If you want the technical mechanics of getting cited directly inside a ChatGPT response, that’s covered in my guide to ChatGPT citations. If you’re trying to understand where AI Overviews specifically fit into this, I’ve written that up separately too: how to actually show up in AI Overviews.

What SEO Actually Costs

SEO doesn’t cost you money to rank – nobody’s paying Google per click for an organic result. But it costs time, and time has a value whether you’re the one spending it or you’re paying someone else to.

If you’re doing it yourself, work out your own hourly rate and be honest about how many hours a proper technical audit, a content rebuild, and an ongoing cadence of internal linking and iteration actually take. If you’re hiring an agency or a freelancer, you’re paying for hours – ours typically run in blocks of ten, twenty, thirty, forty hours depending on the scope, plus the strategy and oversight wrapped around that time. Compared to paid search, where you’re buying clicks directly and the traffic stops the day you stop paying, SEO is slower to compound but tends to keep paying out well after the active work slows down. Calling it free is a bit of a misnomer. Calling it a long-term investment with a real, calculable cost is closer to accurate.

How I Actually Do SEO

My process, stripped down: I start by understanding what problem the business or the piece of content is actually trying to solve, and who it’s trying to solve it for. Then I look at the real search results – type the target term into Google myself, see who’s actually ranking, and work out what they’re doing well.

From there it’s gap analysis. What’s missing that I can add – a stronger internal linking structure, a section the top results gloss over, more backlinks pointing at a page that deserves them. I lean on Search Console to find pages losing ground or sitting just outside the top results, because a page that moves from position 11 to position 6 is often an easier, faster win than trying to build something from zero. Pull the keywords driving that page’s traffic, check the SERP for those terms in Ahrefs, do the same gap analysis, optimize, and push it back out. Then you iterate. SEO isn’t a project with an end date – it’s a loop you keep refining as the data tells you what’s working.

One thing that’s shifted in how I think about keywords specifically: it used to be about a target term and its long-tail variations – running shoes, buy running shoes, best running shoes for winter. Now I think in topics first, and the intent behind each variation matters more than the search volume attached to it. A lower-volume, high-intent term that gets someone to actually book, buy, or act is worth more to most businesses than a high-volume term that just gets someone browsing.

How to Spot Bad SEO Advice

If you’re a founder trying to hire someone, or you’ve already got someone and you’re not sure they’re any good, here’s what I’d tell you to check for.

Anyone promising a guaranteed ranking – a specific position, on a specific date – is telling you something no honest SEO can promise, because none of us control the algorithm. We can commit to the work and to trying. We can’t guarantee the outcome, because it isn’t entirely in our hands.

Watch the language in your reporting. If every update is “it’s coming, it’s coming, it’s coming” with no real explanation of what’s actually being done or why results haven’t shown yet, that’s a red flag. Good SEO reporting shows you the work, the reasoning, and an honest account of what’s working and what isn’t – not just reassurance. There’s a habit in this industry of leaning on heavy jargon to make simple things sound complicated, mostly so nobody asks a follow-up question. If an explanation makes you feel less informed than before you asked, that’s usually not because the topic is genuinely that complex.

Where This Goes Next

The three-part map isn’t going anywhere. What’s still shifting is what falls inside it – more markdown, more llms.txt files, more modular content built to be extracted by a machine as easily as it’s read by a person, more weight on genuine brand mentions across the web rather than just backlinks pointing at your own domain. None of that is a new discipline. It’s the same job, adjusting to more places search now happens.

FAQ

Is SEO dead now that people use ChatGPT and AI Overviews to search? No. Most AI citations still draw from the same underlying search index that traditional SEO targets – Ahrefs found 88% of ChatGPT’s cited URLs come from its general search index. Ranking well in traditional search remains the foundation AI visibility is built on top of.

Do I need a separate strategy for GEO or AEO, on top of my SEO strategy? Not a separate one – an extension of the same one. The structural and content-quality work that helps a page rank well in Google also tends to make it easier for an AI system to extract and cite. Treat GEO and AEO as additional considerations layered onto solid SEO, not a parallel workstream.

Can I do SEO myself, or do I need to hire someone? You can, particularly for a small site, if you’re willing to invest the time to learn keyword research, on-page structure, and basic technical health. Larger or more competitive sites usually benefit from someone who does this daily, simply because the gap analysis and iteration cycle gets more complex at scale.

How long does SEO take to show results? Most SEO work takes a few months before you see meaningful movement, and results tend to compound rather than spike. If someone’s promising fast, dramatic results, ask exactly what they’re doing to get there – some fast wins are legitimate optimization, and some are the kind of shortcut Google eventually penalizes.

What’s the difference between SEO and SEM? SEM (search engine marketing) is the umbrella term covering both SEO and paid search (PPC). SEO is the organic side – work that earns visibility without paying per click. PPC is the paid side, where you bid for placement and pay when someone clicks.

What are SEO keywords, and are they still relevant? Keywords are the terms and phrases people actually type or speak when searching for something. They’re still relevant, but the framing has shifted from isolated keyword lists toward topics and the intent behind each variation – what the searcher actually wants to do next matters more than the raw search volume of the term they used to get there.

How much does SEO actually cost? There’s no fee to rank organically, but SEO costs time – yours if you do it yourself, or an agency’s or freelancer’s hours if you hire out. Agency work is typically billed in blocks of hours; freelancers and consultants bill similarly. Weigh the hourly cost against the compounding, longer-term traffic SEO tends to generate versus the traffic that stops the moment you stop paying for ads.

I joined Synergize in 2011 – Saatchi & Saatchi wouldn’t acquire the agency until years later. About a year into the job, Google released Penguin. It didn’t just knock a few positions off client rankings. Sites got pulled from the index outright, and for the businesses behind them, that wasn’t a ranking dip. It was the channel that brought them customers disappearing overnight.

That’s not a dramatic origin story. It’s just the year I happened to walk in. But it means I didn’t read about the last fifteen years of SEO history in a blog post. I watched clients lose rankings to Panda, rebuild through Hummingbird, panic through Mobilegeddon, and now retool everything again for AI Overviews. Most of what’s been written about “the history of SEO” treats it as a museum tour – dates, names, a tidy timeline. It rarely explains why each shift happened at the infrastructure level, or what it actually did to the tactics people were running the week before.

So here’s the fuller version. Thirty years, condensed into the moments that mattered, with the technology change underneath each one – because the tactics only make sense once you understand what Google was actually building.

Before Google, there was no “SEO” – just guesswork (1990-1998)

The web’s first search tool wasn’t built to rank pages. Archie, launched in 1990 by a McGill University student, indexed file names on FTP servers. It had no concept of relevance, just a list of what existed.

By the mid-1990s, a handful of tools tried to make sense of the growing web: AltaVista, Excite, Infoseek, Lycos, and human-curated directories like Yahoo. Getting found meant getting a human editor to add your site to a category, or stuffing your meta tags with every word you wanted to rank for. There was no ranking algorithm to speak of – Yahoo’s directory was staffed by actual people deciding what belonged where.

Nobody called this “SEO” yet. Practitioners called it “web positioning” or “site promotion.” The industry veteran Ammon Johns has described this period simply: it didn’t have a name because it wasn’t yet a discipline, just a set of tricks people were quietly comparing notes on.

This matters more than it sounds like it should. Every era of SEO since has followed the same shape: a technique works, it spreads, and eventually the search engine has to build a system to stop it working. The 1990s are where that cycle started, before anyone had named the cycle itself.

PageRank rewrites the rules: links as votes (1996-2003)

In 1996, two Stanford PhD students, Larry Page and Sergey Brin, built a search tool called BackRub. It ranked pages by counting and weighing the links pointing at them – treating each link as a vote of confidence rather than a keyword to match.

That idea became PageRank, patented through Stanford, and it’s the single biggest technology shift in the history of search. Before it, ranking was a text-matching problem: does this page contain the word I searched for? After it, ranking became a trust-and-authority problem: does the web, collectively, vouch for this page?

Google launched in 1998. It took two more years to prove the model actually worked at scale, and the proof came from an unlikely source. In 2000, Yahoo – then the biggest portal on the internet – made the decision to power its own search results with Google’s index. Every Yahoo search result carried a small line: “Powered by Google.” Within a couple of years, Google wasn’t the underdog anymore.

For SEO, this meant one thing: backlinks became the currency. If Google was going to count links as votes, then getting more links – by any means available – was the fastest way to rank. That incentive didn’t age well.

The gold rush and Google’s first crackdown: Florida (2003-2005)

Between 2000 and 2003, link building turned into an arms race. Reciprocal link exchanges, directory submissions by the hundreds, exact-match domains, footer link stuffing – all of it worked, because PageRank had no real way yet to tell a genuine citation from a manufactured one.

In November 2003, Google released what practitioners still refer to by name: the Florida update. It was the first algorithm change that visibly, deliberately penalised keyword stuffing and manipulative linking at scale. Sites that had been ranking comfortably for competitive commercial terms vanished from the results within days. It was also the first time the SEO industry understood that Google could – and would – punish a tactic it had previously tolerated.

The same year, Google launched AdSense and acquired Blogger, which quietly created the economic engine behind an entirely different problem: content built purely to host ads, not to inform anyone. That tension – content built for search engines versus content built for people – has never gone away. It’s just changed shape every few years since.

2005 brought three infrastructure moves that mattered more than they seemed to at the time. Google, Yahoo, and MSN jointly introduced the nofollow attribute in January, giving webmasters a way to tell search engines not to pass ranking value through a link – a direct response to comment spam. In June, Google rolled out personalised search, tailoring results to a user’s history rather than treating every searcher as identical. In November, Google Analytics launched, and for the first time, SEOs had a free, detailed window into what was actually happening on their sites rather than guessing from rankings alone.

Spam scales, so does the countermeasure: Caffeine (2005-2010)

If the early 2000s were opportunistic, the back half of the decade was industrial. Article spinning software rewrote the same content thousands of times to avoid duplicate content filters. Link farms and automated forum-and-comment spam tools like Xrumer pumped out backlinks by the tens of thousands. The tactics that had worked in small doses in 2003 were now running at factory scale.

Bing launched in 2009, positioned by Microsoft as a genuine Google alternative. It didn’t reshape the market – Search Engine Journal’s own comparisons found little meaningful difference in result quality, beyond Bing weighting URL keywords and capitalisation slightly differently. What did reshape things, in 2010, was Google Caffeine.

Caffeine wasn’t a ranking algorithm update in the way Florida was. It was an infrastructure rebuild – a new indexing system that let Google crawl and index content in near real time instead of in periodic batches. Before Caffeine, a new page might take days or weeks to show up in search. After it, that dropped to minutes. It’s easy to skim past as a technical footnote, but Caffeine is the reason breaking news, forum threads, and freshly published pages could start ranking within hours – and it laid the groundwork for the freshness signals Google now leans on constantly.

Google’s quality reckoning: Panda, Penguin, Hummingbird (2011-2013)

This is where I stop reciting history and start remembering it.

Panda hit in February 2011, and it went after content farms directly – sites publishing enormous volumes of thin, ad-heavy content purely to capture search traffic. It didn’t touch backlinks at all. It scored the content itself: was this page actually useful, or was it filler built to rank?

Penguin followed in April 2012, about a year into my time at Synergize. Where Panda judged content, Penguin judged links – specifically, the manipulative kind. Sites that had built their entire visibility on exchanged links, paid link networks, and over-optimised anchor text didn’t just slide down a few positions. Some got removed from the index entirely. I watched agencies that had built client strategies around link volume scramble to explain, in the same week, why a client’s site had effectively vanished from Google – not a ranking problem, a business problem, because that traffic was where their leads came from. The lesson landed hard and it landed fast: quantity of links was never the asset. Quality was, and Google had just built the system to tell the difference.

Hummingbird, in 2013, is the quieter update of the three but arguably the most important technically. It was a full rewrite of Google’s core search algorithm, built to interpret the meaning behind a query rather than just matching the words in it. This is the moment “search” started to mean something closer to “understand” – the first real step away from keyword matching and toward semantic search. It didn’t cause the same visible carnage as Panda or Penguin, but it set up everything that came after it.

Search starts to think: RankBrain, mobile-first, BERT (2015-2020)

By 2015, two shifts were running in parallel, and both were about the same underlying idea: Google needed to understand context, not just count signals.

RankBrain, introduced in 2015, was Google’s first machine-learning system built directly into ranking. It helped interpret ambiguous or novel queries – the roughly 15% of searches Google had never seen before – by learning patterns rather than following fixed rules. The same year, “Mobilegeddon” gave a ranking boost to mobile-friendly pages, responding to a milestone that had just landed: 2015 was the first year mobile searches overtook desktop searches on Google. Search had physically moved into people’s pockets, and the algorithm had to follow.

Then came BERT in 2019 – a transformer-based language model that let Google parse the nuance in prepositions, word order, and phrasing that previous systems missed entirely. “Can you get medicine for someone at a pharmacy” means something very different from “can you get medicine at the pharmacy for someone,” and BERT was the first Google system that could reliably tell the two apart. By 2020, it was running on nearly every English-language query.

Alongside the algorithmic shift, Google was rebuilding its infrastructure to match how people actually searched. Mobile-first indexing – crawling and ranking sites based on their mobile version rather than desktop – began rolling out from 2016, became the default for new sites in 2019, and Google confirmed the transition complete across the entire web in October 2023, with the last stragglers folded in by mid-2024. Nearly eight years, start to finish, to move the entire index onto a mobile-first foundation.

This era also introduced E-A-T – expertise, authoritativeness, trust – as a named framework in Google’s quality rater guidelines, sharpened further by the Medic update in 2018. For the first time, Google was explicit that who was saying something mattered, not just what was said.

The AI content flood and Google’s response (2022-2024)

ChatGPT launched publicly in November 2022. Within months, AI-generated content wasn’t a novelty – it was a production method, and a huge share of the web started using it to scale content output far beyond what any human team could write.

Google’s response came in stages. The Helpful Content Update in August 2022 was built specifically to demote content written primarily to rank rather than to help a reader – a direct shot at the exact problem generative AI was about to supercharge. It wasn’t enough. By March 2024, Google shipped one of the largest core updates in its history: a combined core and spam update that reduced “unhelpful, unoriginal content” in search results by an estimated 40%, according to Google’s own figures, and deindexed hundreds of sites outright. It took 45 days to fully roll out. Two months later, Google introduced a Site Reputation Abuse policy, targeting “parasite SEO” – the practice of publishing third-party content on an established, trusted domain purely to borrow its authority.

The pattern from Florida in 2003 repeated itself, just with a faster production tool behind it. A technique scales. Google notices. Google builds a system to demote it. What changed this time is how quickly the cycle turned – a matter of months, not years – because generative AI had made the exploit so much easier to run at volume.

Search stops sending clicks: AI Overviews to AI Mode (2024-2026)

This is the shift most SEOs are still catching up to, and it’s worth being precise about the timeline because it moved fast.

Google introduced AI Overviews – AI-generated summaries sitting above the traditional results – to US search in May 2024, built on a search-specific Gemini model. By December 2025, AI Overviews were appearing on roughly 34.5% of all queries. By March 2026, that had jumped to 48% – a 58% increase in three months. Where an AI Overview appears, click-through to the position-one organic result drops by around 18%, because the answer is often already on the page.

Then, at Google I/O in May 2026, Google went further: AI Mode, running on Gemini 3.5 Flash, became the default global search experience rather than an opt-in tab – what Google itself called the biggest change to the search box in twenty-five years. AI Mode’s zero-click rate sits around 93%. Most people asking a question inside it never visit a website at all.

Here’s the part that should reframe how you think about that number. Sites that do get cited inside AI Mode responses see roughly 35% more organic clicks than sites that only appear in traditional results. The competitive objective has quietly moved from “rank first” to “get cited” – and those are not the same skill. Ranking first has always rewarded relevance and authority signals a machine could measure. Getting cited rewards content a generative system chooses to quote, which means content specific enough, original enough, and evidenced enough that paraphrasing it produces something worse than linking to it – the mechanics of that are in the ChatGPT citation guide. Generative engine optimization exists as a discipline because that skill needed a name, the same way “SEO” needed one back in 1997.

What three decades of updates actually teach you

Strip away the dates and the pattern is almost boring in its consistency. A technique works. It gets exploited past the point Google can tolerate. Google rebuilds part of its system – sometimes the ranking logic, sometimes the infrastructure underneath it – to close the gap. The practitioners who survive each cycle aren’t the ones who find the next trick fastest. They’re the ones who never stopped optimising for the actual person doing the searching, so each new system has less to punish them for.

That’s not a nostalgic conclusion. It’s the same argument I’d make about AI Overviews and AI Mode today. Rankings were always half the story – intent and context mattered more than position long before an AI summary made that obvious. The platforms keep changing. Directories gave way to PageRank, PageRank gave way to semantic search, semantic search gave way to generative answers. The one thing that hasn’t moved in thirty years is who all of it is supposed to serve.

Where to from here? Understand how LLM search actually retrieves and cites your content, because that’s the version of Florida you’re living through right now.

FAQ

When did SEO actually start? The practice predates the name. Webmasters were tweaking meta tags and submitting sites to directories as early as 1994-1995, but the term “search engine optimization” only started appearing in 1997, popularised by early practitioners and publications like Search Engine Watch.

Who coined the term “SEO”? There’s no single inventor. Early digital marketers including Bob Heyman, Leland Harden, and Bruce Clay are credited with shaping the practice in the mid-to-late 1990s, while journalist Danny Sullivan helped popularise the term through Search Engine Watch.

What was the biggest Google algorithm update in history? It depends on the measure. Panda (2011) and Penguin (2012) caused the most visible, sudden ranking losses for individual sites. The March 2024 core and spam update may be the largest by scale, reportedly cutting unhelpful content in search results by around 40%. AI Mode’s rollout in 2026 may end up the most consequential of all, since it changes what “ranking” even means.

Is SEO dead because of AI Overviews and AI Mode? No, but the goal has shifted. With AI Mode carrying a zero-click rate near 93%, ranking first matters less than being the source an AI system chooses to cite – and cited sites see meaningfully more traffic than sites that only rank traditionally. The skill required to earn that citation is closer to old-fashioned E-E-A-T than to any new trick.

How is GEO different from the SEO history described here? GEO (generative engine optimization) isn’t a break from SEO history, it’s the next chapter in the same cycle – search engines rewarding content that generative systems can trust enough to quote. See GEO vs SEO for where the two overlap and where they diverge.

How often does Google update its search algorithm? Google makes thousands of changes a year, most too small to notice. The named, industry-defining updates covered in this article – Florida, Panda, Penguin, Hummingbird, RankBrain, BERT, Helpful Content, and the core updates since – are the small fraction that visibly reshaped how sites need to be built.