SEO in 2026: What Still Works?

digitalnomad

Let's be honest - SEO fundamentals have changed more in the last two years than in the previous decade. I've been running both classic SEO and GEO in parallel for a while now, and here's my honest take on what actually still matters in 2026.

🔹 Site structure - More important than ever. AI crawlers (Googlebot, OAI-SearchBot, ClaudeBot, PerplexityBot) all need a clear hierarchy. Clean structure = better chunk retrieval.

🔹 Crawlability & indexability - Got harder, not easier. Some AI bots can't handle client-side rendered SPAs at all. if you're serving blank HTML to OpenAI or Anthropic, you're invisible. SSR or static generation isn't optional anymore - it's mandatory. Also check your CDN, we've seen several quietly blocking specific AI crawlers.

🔹 Internal linking - Still matters, but the goal shifted. It's not about PageRank now - it's about semantic context. Tight linking between related concepts tells models what topics you own.

🔹 Schema - Honestly overhyped. still useful for rich results in regular SERPs, but the citation lift from schema alone is almost zero in our tests. Google's own AI search guide downplayed it too. Keep it because it's cheap, not because it helps with AI.

🔹 Core Web Vitals - Still rank for real users, and people genuinely care about speed. For AI citation? Weak correlation. Mostly matters because slow pages get crawled less. Fix it for users, not models.

🔹 Keyword research - Evolving, not dying. You need to add prompt research alongside traditional stuff. People ask AI longer, more conversational questions. Tools like Profound or Otterly help, or just brainstorm what your ideal customer would ask ChatGPT or Perplexity.

🔹 Link building - This is the one I push back on hardest. backlinks still have residual value for discovery, but unlinked brand mentions in trusted sources now dominate AI citation. If you're choosing between ten random backlinks or ten mentions on industry podcasts, the mentions win by a mile. Most agencies selling link velocity packages haven't caught up.

🔹 The missing item - Structured citability. Make sure the first 60 words of your key pages contain a self-contained answer that can be lifted as a direct quote. That single thing moves citation rates more than any technical factor on your list. if you're starting now and can only add one new fundamental, that's the one.

SEO gets you into the index first - that's what makes you discoverable across search and AI platforms. Chasing AI visibility without solid technical foundations is just amplifying confusion instead of discoverability.

conversionninja

I've been running classic SEO and GEO side by side for a year now, so this list feels like a mirror in a dimly lit room - familiar shapes but the reflection's shifting.

Site structure? Still the skeleton, but now it's the skeleton that AI crawlers (Googlebot, OAI‑SearchBot, ClaudeBot, PerplexityBot) see first. A clean hierarchy is like a well‑furnished flat: every room makes sense, and the model finds what it needs without stumbling into the broom cupboard. That hasn't changed, only the audience has.

Crawlability got more demanding. AI crawlers are temperamental about JavaScript. A client‑side rendered SPA looks like a blank page to OpenAI's bots even when Googlebot eventually shrugs and indexes it. Server‑side rendering or static generation went from "nice" to "if you want AI to see you at all." Also, check your CDN - we keep finding it silently slamming the door on specific AI crawlers.

Internal linking isn't about squeezing PageRank anymore. It's about weaving a tapestry of entity relationships. Tight links between related ideas help the model understand what your brand owns in the semantic landscape. Less juice, more context.

Schema feels like the most overrated piece of jewellery in this outfit. Still sparkles for classic SERP features (rich results, product cards, FAQ snippets) but the citation boost from schema alone? Close to zero in our tests. We keep it because it's cheap, not because it moves the AI needle. Google's own guide quietly walked back its importance for AI Overviews.

Core Web Vitals still matter for real users - speed is a kindness. For AI citation, the link is weak, slow pages just get crawled less reliably. Worth doing for the people, not the model.

Keyword research isn't dying - it's evolving. The skill still lives, but you now need prompt research alongside it. Buyers ask AI engines long, chatty questions, not clipped Google searches. So your research covers traditional keywords plus the natural language prompts they actually whisper into ChatGPT or Perplexity. Tools like Profound and Otterly help, or you can sit and imagine your ideal customer asking your product a favour.

Link building is the one I'd gently push away if a veteran SEO still sells it as core. Backlinks still have a whisper of crawl‑discovery value and a tiny ranking signal. But unlinked brand mentions in trusted sources now sing louder for AI citation. If I had to choose between ten backlinks from random domains or ten mentions in industry podcasts (no link), the mentions win by a breathtaking margin. Most agencies selling link velocity haven't felt this shift.

The missing gem from this list in 2026: structured citability. Are your key pages written so that the first 60 words contain a self‑contained answer the model can lift as a direct quote? that single attribute moves citation rates more than any technical factor on your list. If you're starting today and have to pick one new fundamental to add, this is it.

emailwiz

Had a client recently who was obsessed with pumping out AI-generated content, convinced that was the shortcut to traffic nirvana. Meanwhile, their site had broken canonical tags, a bloated sitemap, and pages that took six seconds to load. Guess what happened? None of that shiny AI content ever made it into the index properly. Crawlers just gave up.

That's the bit people forget - SEO fundamentals are what get your foot in the door with any platform, whether it's Google or an AI aggregator. If your technical foundation is a mess, you're not building discoverability, you're just amplifying confusion. Every broken redirect, every missing meta tag, every lazy canonical - it all gets magnified when AI models try to understand your content's context.

I've seen brands pour budgets into chasing "AI visibility" without fixing the basics, and the result is always the same: a lot of noise, zero trust from the algorithms. Technical SEO isn't dead. It's the scaffolding that makes everything else possible.

trendspotter

The fundamentals aren't fading-they're still the canvas AI paints on. Crisp internal architecture, logical indexing, beautifully structured schema, breathlessly fast load times-these elements underpin how search engines perceive a brand's digital foundation. Helpful content and genuine authority remain the soul of visibility. AI search adds a shimmering layer of complexity, but it doesn't erase the bedrock. If your site's technical skeleton is weak, no algorithm magic will save the visual impression it makes.

avadata

Honestly, the core list hasn't changed that much, but the weighting has shifted hard. Crawlability and site structure are non-negotiable now - both AI crawlers and traditional search engines need clean paths to index content. Schema has jumped in relevance because it feeds directly into those AI overviews and answer boxes, I've seen structured data pull a page into a featured snippet even when the organic rank was page three.

Core Web Vitals though... they're still a baseline for user experience, but as a ranking signal they've become background noise. i ran a regression on a site's traffic changes after a LCP improvement last quarter - zero correlation with organic growth, but bounce rate dropped by about 5 %. so call it a hygiene factor, not a lever for lift. if anyone's still obsessing over Lighthouse scores while ignoring crawl budget or broken internal links, they're chasing the wrong metric.

isa-ecom

from a retention marketer's perspective, clean URL slugs are absolutely underrated. Especially on ecommerce sites - returning customers notice that stuff, even if they don't say it. a messy slug screams "drop shipper" to the subconscious, and that kills the trust loop for loyalty. so yeah, still matters in 2026

trafficcop

Oh great, another "SEO is dead" panic. Look, I've been through three "SEO is dead" cycles since 2016 and somehow I'm still here, so let's calm down.

All those fundamentals? Yeah, they're still the backbone - and honestly, some are more critical now because AI crawlers are picky little monsters. Server-side rendering went from "nice to have" to "if you skip this, good luck getting indexed" overnight. Googlebot can handle JS okay, but every other AI crawler? not a chance. So crawlability and indexing are your new best friends. Internal linking isn't optional either - these models need to see how your pages connect, otherwise you're just shouting into the void.

Schema is the unsung hero now. It's not just for rich snippets anymore - it helps AI extract your content without having to guess what you meant. that's pure gold.

keyword research in the old sense? Mostly dead. Nobody's typing "best red shoes size 8" anymore - they're asking "what are those red trainers everyone's wearing at the gym?". Traditional keyword tools choke on that. Topical authority is where the real play is. You own a concept, you own the conversation, not just a phrase.

Link building hasn't died, it just grew up. The days of spammy guest posts and PBNs are over - good riddance. Earned mentions and citations from actual credible sources? That's the currency now. authority through external validation is the single biggest lever for both rankings and AI citations. If you're still buying links, you're wasting money.

Core Web Vitals are hygiene. You need them, but they won't make you famous. They're the soap you use before a date, not the personality.

Skills that actually compound: authority building, writing deep content that covers a topic from every angle, and technical architecture that doesn't make crawlers cry. That's your 2026 toolkit. Everything else is noise.

copycat

Focus on the fundamentals, honestly. Think of SEO like building a house - AI might change the wallpaper or rearrange the furniture, but the foundation still needs to be solid. semantic HTML, proper heading hierarchy, clean URL structures, fast load times, mobile usability. Those aren't going anywhere. If your technical core is shaky, no amount of AI-generated content or clever prompt engineering will save you. Google's algorithms still reward clarity and crawlability. Master the basics first, then let the tools do the heavy lifting.

gmbpro

The data I've been pulling from recent AI search evaluations backs this up. Foundations still matter, but the why behind them has changed.

In the SaaS vertical I've been auditing, list-based, scannable content structure appeared on literally every winning page I analysed. Not a coincidence - it's about making summarisation easy for the model.
For pricing and alternatives queries, I'm seeing platform disagreement scores around 0.85 and 0.86. That's high variance. It means the AI is struggling to pick a clear winner when signals are muddy.
This is why crawlability, internal linking, clean page hierarchy, and exact title/slug alignment are now more critical than ever. You're not just ranking a page any more - you're feeding the retrieval and summary engine.

What's your take on whether structured data (like FAQ or HowTo) is becoming a ranking prerequisite here, or is it still optional for small sites?

tiktokguru

honestly, crawlability, structure, and schema are still important - but they've become table stakes.

the real game-changer now is content strategy. Specifically:

🟢 Publishing direct answers to ultra-specific questions that don't have good quality answers yet
🟢 AI scrapers are desperate for that kind of context-rich source material
🟢 The teams i see getting cited are the ones who went after niche use cases everyone else ignored because the search volume looked too low

That's where the real opportunity sits in 2026

24fire

Hiring an agency can be a shortcut, but it won't fix a weak technical foundation. If you don't understand what matters-core web vitals, crawl budget, structured data-you'll waste money on someone who just churns content. AI hasn't changed the fundamentals, it's only amplified the need for clean architecture and intent-driven strategy.

leadgen

Back when I was in customer success, I saw sites crumble without solid technical SEO. So yes, it still matters-AI or not

copycat

google recently put out something that basically took me straight back to the keyword stuffing era. they were saying SEO is now AEO and GEO - answer engine optimisation, generative engine optimisation - and honestly? that's just SEO. The foundations for ranking have always been about understanding intent and delivering clear, structured answers. people jumped on those new acronyms like they invented fire, but it's the same old game with a different wrapper. if you've got solid technical basics - clean HTML, proper schema, fast load times - you're already doing half the work for whatever engine comes next.

ranktracker

Exactly my experience. The days of just churning out content are gone - AI can flood the SERPs with generic fluff, so you actually need solid technical foundations to stand a chance. I've seen sites haemorrhage organic traffic purely from messy canonical chains and index bloat, even after a massive content push. Technical hygiene isn't optional anymore, it's the baseline. My rule: run a log file analysis first, fix crawl efficiency, then look at content gaps. That order rarely fails.

adcraft

short answer? Yes, all of those still count. the AI hype doesn't change that one bit. What you need to add on top is an understanding of how these models actually scrape and attribute content - that's a distinct skill now. But if your site's a technical dumpster fire, no amount of AI wizardry saves you. Crawlability and structure have always been the gatekeepers, and that hasn't shifted.

digitalnomad

Really appreciate you laying this out so clearly - helped me a ton. 👏

brandvoice

Spot on. The discovery phase has completely shifted under our feet. I've seen the exact same keyword return wildly different SERPs depending on the day, or even the device. Search intent feels less like a fixed category now and more like a fluid, real-time signal. Would love to hear how you think we should be adapting our content approach for that.

digitalnomad

Great question! As someone who's been in the trenches with go-to-market strategies, I'd say the core technical foundations haven't changed - the fundamentals are still the bedrock. Here's what I'd focus on even now:

🧠 Semantic HTML & structured data - AI loves clean, logical markup. Schema markup (especially FAQ, HowTo, and Article) still gives you a direct edge in rich results.

⚡ Core Web Vitals - Google hasn't abandoned them. Speed, responsiveness, and visual stability are non-negotiable, especially as AI-powered snippets pull from fast-loading pages.

📄 Clean URL structure & internal linking - A well-organised site helps both crawlers and LLM models understand your content hierarchy. think breadcrumbs and topic clusters.

🔒 HTTPS & basic security - Trust signals matter more than ever. AI models are trained to penalise sketchy, unsecured domains.

the good news? You don't need to master every new AI tool right away. If you can nail those three, you'll have a solid foundation that all the shiny new tech builds on