Do Keywords Matter for SEO? Here’s the Truth

TL;DR

• Keywords still matter, but they’ve gone from headliner to supporting act. The real question isn’t whether to use them. It’s when keywords should lead your strategy and when they shouldn’t.
• Zero-click searches now hit 64.82% of Google queries in 2026 [1]. Nearly two-thirds of searches end without anyone clicking a result. Keyword-driven traffic is shrinking even when rankings hold steady.
• High topical authority sites gain organic traffic 57% faster than low-authority sites [2]. Topic depth is eating keyword breadth’s lunch, and the gap keeps widening.
• AI Overviews reduce organic CTR for the #1 position by 58% [3]. For every 100 clicks you used to get, Google keeps 58. And that’s just the position at the very top.
• Use the framework below to decide, page by page, whether to target a keyword or build around a topic cluster.

I spent the first half of my career obsessing over keywords. Exact match. Long-tail. LSI. I had spreadsheets with color-coded difficulty scores and monthly volume projections that would make a data analyst weep. Some of that effort paid off. A lot of it was wasted motion.

Here’s what made me rethink everything: in late 2025, I watched a client’s page hold its #3 ranking for a 12,000-volume keyword while organic traffic to that same page dropped 40% in six months. The ranking never moved. The clicks just vanished. That’s the world we’re in now, and most keyword advice hasn’t caught up.

This article won’t give you the “keywords matter, but differently” speech that twelve other articles already cover. Instead, I’ll give you a framework for deciding when keywords should drive your strategy and when they shouldn’t, plus a look at why SEO has split into two games that most people still treat as one.

Why “Do Keywords Still Matter?” Is the Wrong Question

It’s the wrong question because it demands a yes-or-no answer to something that isn’t binary. Of course keywords matter. They’re how search engines initially understand what a page is about. But asking if keywords “matter” is like asking if flour matters when you’re baking. Obviously yes. The better question is: what’s the recipe, and has it changed?

It has. Google’s ranking systems now process queries through layers of semantic understanding that didn’t exist five years ago. BERT (2019) taught the engine to read sentences bidirectionally. MUM (2021) cross-references information across languages and formats. And the constant iterative improvements to Google’s AI stack mean the engine no longer matches strings of text. It interprets meaning. A page about “budget project management tools for small teams” can rank for “cheap PM software for startups” without that phrase appearing anywhere on the page.

But here’s where most articles on this topic stop. They say “intent matters more than exact match,” toss in a few examples, and move on. That’s true but incomplete. What they skip is this: the value of a keyword depends entirely on where the searcher’s click ends up. And increasingly, it ends up nowhere.

The Two-Game Split: Ranking vs. Getting Cited

Here’s something I rarely see framed clearly. Since AI Overviews, ChatGPT search, Perplexity, and Google AI Mode became mainstream, SEO has fractured into two distinct games. Most marketers are still playing just the first one.

Game 1: Ranking for clicks. Traditional SEO. You target a keyword, optimize your page, earn links, and collect traffic when someone clicks through from search results. Keywords are the entry ticket here.

Game 2: Getting cited by AI. Your content gets referenced in AI Overviews, ChatGPT responses, or Perplexity answers. The user may never visit your site, but your brand and ideas get distributed at scale. Keywords matter less here. Topic authority, original data, and clear, extractable writing are what get you cited.

These two games require different strategies. And different pages on your site might be playing different games.

An Ahrefs study analyzing 300,000 keywords with December 2025 data found that AI Overviews reduce the organic click-through rate for position one content by 58% [3]. For every 100 clicks you used to collect, Google now keeps 58 of them. And that’s just for queries where AI Overviews appear - which, depending on methodology, now trigger on roughly 21% to 25% of Google searches [4].

Meanwhile, zero-click searches - queries where the user never clicks any result - climbed to 64.82% of all Google searches in 2026 [1]. On mobile, which accounts for about 63% of all searches, that number reaches 77.2%. More than three-quarters of mobile searches end right there on the results page.

So if you’re only playing Game 1, you’re fighting for a shrinking share of a shrinking pie. And if you’re ignoring Game 2, you’re invisible in the fastest-growing segment of search - AI platforms like ChatGPT Search (900 million weekly active users as of February 2026 [5]), Perplexity, and Google AI Mode.

Pro Tip: Audit your existing content by asking one question per page: “Is this page primarily trying to earn clicks, or earn citations?” Then optimize accordingly. Click-focused pages need precise keyword targeting and compelling meta descriptions. Citation-focused pages need unique data, named expert perspectives, and clear, self-contained answers that AI can extract and quote.

The Keyword Value Spectrum: A Decision Framework

I got tired of advice that boils down to “it depends,” so I built a framework. I call it the Keyword Value Spectrum. It helps you decide, for any given page, how much keyword optimization effort is worth your time.

Signal	Keywords Lead (High Keyword Value)	Topics Lead (Low Keyword Value)
Query type	Transactional, commercial, navigational	Informational, educational
AI Overview present?	No AI Overview in the SERP	AI Overview dominates the SERP
Click-through potential	High (product pages, comparisons, tools)	Low (definitions, quick-answer questions)
Competitive gap	You can realistically reach page 1	Top 10 dominated by massive domains
Content uniqueness	Your page offers a distinct product or service	Your content covers the same ground as 10 others
Business value	Directly tied to revenue (buy, sign up, demo)	Brand awareness or thought leadership

When the signals cluster on the left, go hard on keyword targeting. Optimize your title tag, URL, H1, and on-page copy around a specific primary keyword. This is the playbook that still works for product pages, landing pages, comparison content, and bottom-of-funnel queries. Transactional queries show only a 39.4% zero-click rate [1] - people searching to buy still need to reach a destination.

When the signals cluster on the right, shift your energy to topic authority. Build a cluster of interlinked content around a broad subject. Write with depth and specificity. Include original data, named sources, and perspectives other articles don’t have. This is how you win citations from AI engines and build durable authority that compounds over time.

Most pages fall somewhere in the middle. That’s fine. The point isn’t to pick a side forever. It’s to stop applying the same keyword-optimization checklist to every page you publish.

What Actually Gets Cited by AI (Spoiler: Not Keyword Density)

Google published official guidance in May 2025 about succeeding in AI search experiences [6]. Their core recommendation: “Focus on making unique, non-commodity content that visitors from Search and your own readers will find helpful and satisfying.” That word “non-commodity” is doing heavy lifting. It means if five other articles say the same thing, yours won’t get cited.

The Animalz content team explored this through Google’s information gain patent [7]. Their conclusion matches what I’ve seen firsthand: “If your content repeats what 10 other articles already say, AI makes it redundant before you hit publish.”

I’ve tracked what earns citations across client campaigns and spotted consistent patterns:

1. Original data or statistics. If you ran a survey, analyzed your own product data, or compiled primary research, AI engines reference it. You become the primary source. A Stratabeat analysis of 300 B2B SaaS websites confirms this: sites offering original research grew organic traffic by 29.7% versus 9.3% for those that didn’t [8].

2. Named expert perspectives. A direct quote from a real person with credentials carries more weight than generic advice. AI tools frequently pull blockquotes with attribution.

3. Specific, tactical advice. Not “use long-tail keywords.” More like “we shifted from targeting ‘email marketing software’ to building a 15-article cluster around email deliverability, and organic traffic to that section grew 34% in four months.”

4. Clear, self-contained answers. Every sentence AI cites needs to make sense without surrounding paragraphs. Write like each paragraph could be extracted and shown in isolation.

5. Structured, scannable formatting. Clear headings, short paragraphs, and lists make content easier for AI to parse and cite. An Otterly analysis of 1+ million AI citations across ChatGPT, Perplexity, and Google AI Overviews found that content with proper schema markup and chunked formatting received 3-5x more citations [9].

Notice what’s not on that list? Keyword density. Exact-match keyword placement. Meta keyword tags - Google publicly abandoned those as a ranking signal in 2009 [10]. The things that once formed the backbone of keyword strategy don’t influence AI citation at all.

Where Keywords Genuinely Still Earn Their Keep

I don’t want to overcorrect here. Keywords haven’t become irrelevant. They’ve become contextual. There are specific situations where smart keyword targeting still makes or breaks performance.

Title tags and H1s. These remain the strongest on-page signals for telling Google what a page is about. Ahrefs research confirms the average #1 ranking page also ranks in the top 10 for nearly 1,000 other keywords, but the primary keyword in the title tag anchors initial relevance [11].

Transactional and commercial pages. When someone searches “buy standing desk under $500” or “best CRM for real estate agents,” they have specific purchase intent. These queries reward precise keyword targeting. The person isn’t looking for a think piece. They want a match. Commercial investigation queries show a zero-click rate of just 51.2%, compared to 74.3% for informational queries [1].

New sites building initial authority. If your domain is young and hasn’t established topical authority yet, keywords are your entry point. You need to signal clearly what your content covers before Google trusts you enough to rank for semantically related terms. Think of it like introducing yourself at a party. You state your name before people recognize your face.

Long-tail queries with low competition. A 6-word query with 200 monthly searches and no AI Overview? That’s still free traffic if you nail the intent. Long-tail keywords convert at roughly 2.5x the rate of head terms [12] because they capture specific intent. And since AI Overviews heavily favor informational queries, long-tail transactional and commercial searches remain a reliable source of click-through traffic.

“Go where your audience goes. Rather than chasing general trends, focus on understanding your specific audience’s behavior on the internet.”

• Rand Fishkin, CEO of SparkToro, during the Datos Q4 2025 State of Search webinar

Fishkin’s point is relevant because it reframes keyword research from “find the biggest volume” to “find where my specific audience is actually clicking.” That distinction matters more every quarter.

Topic Authority: The Thing Quietly Eating Keyword Strategy’s Lunch

Topical authority is the degree to which Google (and AI engines) trust your site as a reliable source on a specific subject, built by consistently publishing deep, interlinked content around that subject.

A Graphite study published in April 2026 found that sites with high topical authority scores gain organic traffic 57% faster than sites with low scores [2]. High-authority pages also reached 100% visibility within three weeks of publishing, compared to lower-authority counterparts that took significantly longer. Google itself confirmed in 2023 that topical authority exists and impacts search results.

Why does this matter for the keyword debate? Because topical authority is what lets you rank for keywords you never explicitly targeted. When Google trusts you on a subject, it fills in the gaps. You don’t need separate pages for “email open rate benchmarks,” “average email open rate by industry,” and “what’s a good email open rate.” One thorough, authoritative page supported by related content in a cluster can rank for all three and dozens more.

This frustrates people who learned SEO through keyword research tools. The tools make it look like every keyword deserves its own page. The reality in 2026 is that building depth on a topic is almost always more effective than scattering thin pages across dozens of keyword variations. The Otterly AI citation study found that community-driven content (Reddit, Quora, Wikipedia) captures a staggering share of AI citations - Reddit alone accounts for 9.7% of all LLM citations across platforms [9] - because these sources demonstrate deep, authentic topical expertise that brands often lack.

The Practical Shift: How I Changed My Own Workflow

I used to start every content brief with a target keyword and a list of secondary keywords. Now I start differently.

Step one: I pick a topic where we have or can build genuine expertise. Step two: I look at what currently ranks and what AI engines are saying about the topic. Step three: I identify the gap - the thing nobody else is covering, the angle nobody has taken. Step four: only then do I pull up keyword data to validate that real people are searching for this, figure out what specific phrases to use in titles and headers, and check whether AI Overviews dominate the SERP.

The order matters. Starting with the keyword made sense when Google was a matching engine. It makes no sense when Google is an understanding engine.

Real example. A B2B client in the fintech space wanted to rank for “payment processing for SaaS.” Volume: 880/month. Keyword difficulty: high. AI Overview present. The old approach would’ve been to write a keyword-optimized guide targeting that exact phrase.

Instead, we built a 12-piece content cluster around SaaS payment infrastructure: revenue recognition rules, dunning management, multi-currency billing, PCI compliance for subscription platforms. The “payment processing for SaaS” page became the pillar, but it earned its ranking because of the cluster supporting it, not because we stuffed the keyword into 14 places on a single page.

Traffic to that cluster grew 41% over five months. Three of the supporting articles started appearing in ChatGPT and Perplexity answers about SaaS billing - something that never would’ve happened with a keyword-first approach.

Frequently Asked Questions About Keywords and SEO

Do you still need to put keywords in your title tag and H1?

Yes. Title tags and H1s remain the clearest on-page signals for Google to understand a page’s primary topic. The average #1 ranking page also ranks for nearly 1,000 other keywords [11], but the primary keyword in the title tag establishes initial relevance.

Is keyword density still a ranking factor?

No. Google has not used keyword density as a ranking signal for years. A 2026 Rankability study analyzing 1,536 Google search results found “no consistent correlation between keyword density and ranking” [13]. Write naturally, include your primary keyword in key positions (title, H1, first 100 words, a subheading or two), and focus the rest of your energy on covering the topic well.

Should I still do keyword research in 2026?

Yes, but its role has shifted from driving content creation to validating content ideas. Use keyword data to confirm there’s actual search demand for your topic, identify the specific phrasing your audience uses, assess competition, and decide whether a query is worth targeting for clicks or for AI citations. Start with topic ideation, then use keyword research to sharpen your approach.

How do AI Overviews affect keyword strategy?

AI Overviews reduce organic CTR by 58% for the #1 position [3] and trigger on roughly 21-25% of Google searches [4]. For informational queries where AI Overviews appear, optimizing for a specific keyword may earn you a ranking that delivers far fewer clicks than expected. In these cases, focus on creating content with unique data and clear, citable statements so AI Overviews reference your page as a source.

What’s more important in 2026, keywords or topics?

Both matter, but they serve different purposes. Keywords help you target specific queries. Topics help you build the authority that makes everything else easier. The Graphite study found that high topical authority sites gain traffic 57% faster [2]. For most sites, the best approach is to build topic clusters and use keywords strategically within that structure, rather than treating individual keywords as standalone targets.

Does content written for AI engines also rank on Google?

Generally, yes - when done right. Gartner projects traditional search volume will drop 25% by the end of 2026 [14], but the content that earns AI citations (unique data, clear structure, deep topical coverage) also performs well in traditional search. The fundamentals overlap more than most marketers realize. Whether a page primarily plays Game 1 or Game 2, the underlying quality signals are similar.

The Bottom Line

Keywords aren’t dead. They’re not dying. But they’ve been demoted, and pretending otherwise will cost you time and traffic.

The shift is straightforward: move from keyword-first to topic-first. Use keywords as a tool for validation and precision, not as the foundation of your entire strategy. Build depth on subjects you genuinely know. Create content that earns citations, not just rankings. Use the Keyword Value Spectrum to decide, page by page, where to put your energy.

The pages that will win over the next few years are the ones that would be worth reading even if Google didn’t exist. Start there.

You might also find these resources helpful:

• How to Do Keyword Research in 2026 - Our full guide to modern keyword research workflow.
• Content Clusters and Pillar Pages - How to structure topic clusters that build topical authority.
• LLM Visibility Guide - Optimizing for ChatGPT, Perplexity, and AI Overview citations.
• Zero-Click SEO Strategy - How to maintain visibility when clicks disappear.

Sources

1. Digital Applied, “Zero-Click Search Statistics 2026: Complete Data Guide,” April 2026. https://www.digitalapplied.com/blog/zero-click-search-statistics-2026-complete-data
2. Graphite, “Study Shows High Topical Authority Leads to Faster Organic Search Visibility,” April 2026. https://graphite.io/five-percent/topical-authority-white-paper
3. Ahrefs, “Update: AI Overviews Reduce Clicks by 58%,” February 2026. https://ahrefs.com/blog/ai-overviews-reduce-clicks-update/
4. QuickSEO, “Google AI Overviews Statistics 2026: 60+ Data Points Every SEO Should Know,” May 2026. https://quickseo.ai/blog/google-ai-overviews-statistics-2026-60-data-points-every-seo-should-know
5. Search Engine Land, “OpenAI: ChatGPT Now Has 900 Million Weekly Active Users,” February 2026. https://searchengineland.com/chatgpt-900-million-weekly-active-users-470492
6. Google Search Central, “Top Ways to Ensure Your Content Performs Well in Google’s AI Experiences,” May 2025. https://developers.google.com/search/blog/2025/05/succeeding-in-ai-search
7. Animalz, “Information Gain: The SEO Theory That AI Made Mandatory,” November 2025. https://www.animalz.co/blog/information-gain
8. Stratabeat, “2025 B2B SaaS SEO Performance Report,” 2025. https://stratabeat.com/research/b2b-saas-seo-study/
9. Otterly, “The AI Citation Economy: What 1+ Million Data Points Reveal About Visibility in 2026,” February 2026. https://otterly.ai/blog/the-ai-citations-report-2026/
10. Google Search Central Blog, “Google Does Not Use the Keywords Meta Tag in Web Ranking,” September 2009. https://developers.google.com/search/blog/2009/09/google-does-not-use-keywords-meta-tag
11. Ahrefs, “What Are Secondary Keywords? (And How to Use Them),” March 2026. https://ahrefs.com/blog/secondary-keywords/
12. Yotpo, “Long-Tail Keywords: The Ultimate Guide for 2026,” March 2026. https://www.yotpo.com/blog/long-tail-keywords-guide/
13. Rankability, “Is Keyword Density a Google Ranking Factor? Research Study 2026,” 2026. https://www.rankability.com/ranking-factors/google/keyword-density/
14. Gartner, cited via LinkedIn/Knotch, “Google Q1 Earnings: LLMs and AI Overviews Impact Brand Marketers,” May 2026.