Your SEO skills from a year ago need a refresh

Hiba Fathima · 18th Feb 2026 · 5 min read

If you learned SEO 12–18 months ago and haven't updated since, you're still right about the fundamentals. You're missing a lot of the execution.

This isn't a "SEO is dead" take. It's the opposite. SEO has expanded. The skills required now are wider than they were, and some techniques that worked a year ago are either table stakes or actively less effective.

Here's what changed and what to update.

What hasn't changed (anchor here)

Before the refresh: authority, relevance, and trust are still the core. If your content doesn't demonstrate genuine expertise and earn links and mentions from credible sources, no amount of tactical optimization fixes that.

Technical hygiene still matters too. Crawlability, indexability, page speed, Core Web Vitals—these aren't exciting, but they're still the floor.

What has shifted

1. The click is no longer the primary unit of value

A year ago, the goal of a page ranking in position one was to earn the click. That's still true, but it's no longer the whole picture. AI Overviews, featured snippets, and Perplexity responses increasingly answer the query without a click.

This changes how you measure. Traffic is not the same as visibility. You need to track:

  • Impressions vs. clicks (the gap is widening)
  • Brand mentions in AI-generated answers
  • "Zero-click" brand awareness that shows up later in direct or branded search

If you're only measuring organic traffic, you're missing a growing share of the value your content is generating.

2. Entity optimization matters more than keyword density

Search engines—and LLMs—understand entities. Your page shouldn't just contain the keyword; it should clearly establish relationships between entities. Who wrote it, what brand it belongs to, what topics it covers, what it cites.

Structured data (Schema.org markup) has been around for years, but it's now doing more work than before. Person, Organization, Article, FAQPage, and HowTo schemas feed directly into how AI systems understand and attribute your content.

The shift: optimize for entity clarity, not keyword repetition.

3. Content structure is load-bearing

In 2023, a well-written long-form article could rank on the strength of its substance alone. Today, structure is how models extract answers. An article that buries its key point in paragraph six loses to a structurally similar but shallower piece that leads with the answer.

Practical changes this requires:

  • Lead with your conclusion or answer, not a warm-up
  • Use H2s that are answerable questions, not generic topic labels
  • Write in paragraphs of 2–4 sentences, not walls of text
  • Include a FAQ section for high-intent query variants

4. Topical authority requires depth and breadth

It used to be enough to have one great post on a topic. Now search engines and AI systems evaluate your whole content footprint. Do you have a cluster of related, interlinked content? Do multiple pages on your site address the same topic from different angles?

This matters for both traditional rankings and AI citations. Models favor sources that consistently cover a domain, not one-off pieces.

5. The tools have changed

Some tools you used a year ago now have features you haven't touched. Ahrefs added Brand Radar for AI citation tracking. Google Search Console data patterns have shifted with AI Overviews launching. Perplexity and ChatGPT are new surfaces to audit manually.

The skill isn't just knowing the tools—it's knowing what questions to ask them now. What's my click-through rate on branded queries? Where am I being cited in AI answers? What topics do I rank for but get no clicks on?

The refresh list

If I were auditing my own skillset, here's what I'd update:

  1. Re-learn structured data. Specifically FAQPage, Article, Person, and HowTo. Implement and validate with Google's Rich Results Test.
  2. Set up AI citation monitoring. Brand Radar in Ahrefs, or manual spot-checks in Perplexity and ChatGPT for your core queries.
  3. Audit your content structure. For your top 10 pages: does each one lead with the answer? Does it have a clear H1, scannable H2s, and a FAQ?
  4. Map your topical coverage. Where are the gaps in your content cluster? What related queries do you not have a page for?
  5. Separate visibility from traffic in your reporting. Impressions, brand mentions, and zero-click exposure need their own line.

The uncomfortable truth

A lot of the SEO industry has been slow to update. Advice from established practitioners is sometimes 18 months behind what's actually working. That's not cynicism—it's just how any industry operates when the underlying platform shifts faster than the writing about it.

The people doing well right now are treating search as a multi-surface problem. They're optimizing for the blue link and the AI citation and the brand mention in the same strategy. That's a different way of working than most teams were built for.

It's worth figuring out sooner than later.


Questions or pushback? Find me on X or LinkedIn.


FAQs

Has SEO fundamentally changed in 2024–2025?

The core principles—authority, relevance, trust—haven't changed. But the surfaces and tactics have shifted significantly with AI Overviews, LLM-powered search, and changing click behavior.

What SEO skills are most important to update?

Understanding how AI Overviews select content, optimizing for entity recognition, structuring content for extraction, and tracking brand mentions in AI-generated answers are the highest-priority updates.

Is keyword research still relevant in 2025?

Yes, but the goal has shifted. Keyword research now needs to map to query intent at a topic cluster level, not just individual page optimization. Volume matters less; answer intent matters more.

What is GEO and how does it relate to SEO?

GEO (Generative Engine Optimization) is the practice of optimizing content to appear in AI-generated answers. It overlaps significantly with SEO but adds requirements around structure, specificity, and citability.


More essays