Why Keywords Are the Wrong Starting Point (And What to Focus on Instead)

If you’ve spent any time thinking about your content’s visibility online, you’ve probably been told to focus on keywords. Find the right ones, use them consistently, and the right people will find you. It sounds logical. And for a long time, it was roughly correct.

But search has fundamentally changed, and businesses that still optimize around keywords as a primary signal are building on a foundation that’s quietly eroding beneath them. Understanding why, and what to focus on instead, is one of the most useful shifts any growth leader can make in how they think about content and organic visibility.

Why Keywords Dominated for So Long

To understand where we are, it helps to understand where we started.

For most of the internet’s history, search engines classified content using a formula called TF-IDF: Term Frequency multiplied by Inverse Document Frequency. In plain terms, it measured how often a keyword appeared on a given page relative to how often it appeared across other documents. The logic was simple: if a page mentions a specific term more than other pages do, it’s probably more relevant to that term. So it should rank higher for it.

This approach worked reasonably well early on. But it had three fundamental limitations that became increasingly problematic as the web scaled.

It ignored meaning. TF-IDF looked at keywords in isolation, without considering variations, synonyms, or relationships between words. The same word can mean entirely different things depending on context, and the formula had no way to account for that. A search for “soap” could mean dozens of completely unrelated things, and frequency-based scoring couldn’t distinguish between them.

It ignored importance. Just because a keyword appears frequently on a page doesn’t mean the content is more valuable or more relevant to the user. A page with fewer keyword mentions but deeper, more nuanced treatment of a topic may be far more useful, but TF-IDF couldn’t recognize that.

It ignored purpose. Most critically, TF-IDF compared content across pages without considering what those pages were actually trying to do. It might weigh a blog post against a product page, a FAQ against a pricing page, content for beginners against content for advanced practitioners. The user’s reason for searching, and the page’s reason for existing, were both invisible to the formula.

The result was predictable: once website owners figured out how TF-IDF worked, they exploited it. Keyword-stuffed pages flooded search results. Content quality degraded. And search engines were forced to evolve.

The Shift From Keywords to Intent

Over the past decade, major search algorithm updates have progressively reduced reliance on keyword frequency and increased reliance on something more sophisticated: understanding what the searcher is actually trying to accomplish.

This is called search intent, and it has become the organizing principle of modern search.

When someone types a query into a search bar, they’re not just entering words. They’re expressing a need. Sometimes that need is obvious from the query itself. More often, it requires interpretation, context, and understanding of where the searcher is in their thinking.

Google now uses machine learning and natural language processing to make those interpretations at scale. The result is a search engine that increasingly thinks less like a keyword-matching system and more like a librarian who understands what you’re actually looking for.

For businesses creating content, this changes the fundamental question. It’s no longer “what keyword should I target?” It’s “what is my audience trying to accomplish, and does my content actually help them do it?”

The Four Types of Search Intent

Understanding intent starts with recognizing that not all searches are the same. There are four primary types worth knowing.

Informational intent. The searcher wants to learn. They’re researching a topic, exploring a problem, or trying to understand something they don’t yet know. They’re not yet in buying mode, but they’re building the knowledge that will eventually lead them there. Content for this intent should educate without immediately selling.

Navigational intent. The searcher is trying to find something specific: a website, a business, a person, or a location. They know where they want to go; they just need help getting there. Branded searches are almost always navigational. Content for this intent should make it as easy as possible to find what the searcher is looking for.

Transactional intent. The searcher is ready to act. They’ve made or are close to making a decision, and they’re looking for the mechanism to execute it: a booking page, a contact form, a product page, a download. Content for this intent should reduce friction and make the next step obvious.

Commercial or investigational intent. The searcher wants to buy or commit, but isn’t quite ready. They’re comparing options, reading reviews, looking for validation, or narrowing a shortlist. This intent sits between informational and transactional, and content for this stage should provide the reassurance and specificity that moves someone from “interested” to “decided.”

Most businesses create content that targets transactional and investigational intent almost exclusively, which means they’re invisible to the large majority of buyers who are still in the informational stage of their journey. By the time those buyers are ready to act, a competitor who was present earlier in their research has already built the relationship.

This maps directly to the awareness spectrum. Buyers at the Oblivious and Apathetic stages are searching with informational intent. Those at the Thinking stage are searching with investigational intent. And those at the Hurting stage are searching with transactional intent. Matching your content to the right intent type means meeting buyers where they actually are.

Intent Is a Signal. Behaviour Confirms It.

Here’s what makes intent so important from a visibility standpoint: search engines don’t just guess at it. They measure it through user behaviour.

When someone searches for a term, clicks on a result, and immediately bounces back to the search results, that’s a signal the content didn’t satisfy their intent. SEO practitioners call this “pogosticking,” and it tells the search engine something useful: this result wasn’t what the user was looking for.

The inverse is also measured. When someone clicks a result and stays, reading deeply before eventually leaving, that’s a “long click.” It signals the content was relevant, valuable, and aligned with what the searcher needed. Over time, content that consistently generates long clicks earns stronger search visibility. Content that generates short clicks loses ground, regardless of how well it was optimized for a target keyword.

This has a practical implication: you can optimize perfectly for a keyword and still underperform if your content doesn’t satisfy the actual intent behind the search. Rankings are a means to an end. The end is meeting the user’s need well enough that they stay.

From Keywords to Entities and Topics

Alongside the shift to intent-based search, another fundamental change is underway in how search engines understand language.

Natural language processing, the technology that enables machines to understand human language, has moved search engines away from treating keywords as isolated signals and toward treating them as “entities.” An entity is a keyword understood in context: who or what it refers to, how it relates to other concepts, and what it means in a given situation.

This matters because the same word can carry entirely different meanings. “Apple” means something different in a discussion about nutrition than it does in a discussion about technology stocks. “Lead” means something different to a sales team than it does to an environmental chemist. A search engine that understands entities can distinguish between these meanings. One that relies on keyword frequency cannot.

The practical implication for content creators is significant. Trying to rank for a specific keyword by optimizing frequency is increasingly futile. What builds visibility now is covering a topic with genuine depth and breadth, using the full range of related terms, concepts, and contexts naturally, in a way that signals to the search engine that the content understands the subject rather than just mentions it.

Topics, not keywords, are the organizing unit of modern search. A topic is an idea with a full context: related concepts, relevant entities, user needs, and awareness stages. When your content reflects that kind of depth, relevant keywords appear naturally throughout, without any forcing.

Why Long-Tail Queries Are More Valuable Than They Look

One of the clearest windows into intent is the specificity of a search query.

Generic short-head keywords, the one or two-word searches that appear to have massive volume, are almost always ambiguous. A search for “consulting” could mean almost anything. A search for “how to price consulting services for the first time” tells you exactly who is searching, what they need, and where they are in their thinking.

This specificity is not a limitation. It’s a signal. Longer, more specific queries carry clearer intent, which means content that matches them is more likely to satisfy the searcher, generate long clicks, and convert. And because they’re less contested than generic terms, they’re often easier to rank for as well.

More importantly, the pattern of long-tail queries across a topic area reveals what your actual audience is actually thinking about. It’s the closest thing available to listening in on the internal monologue of your buyers as they research, compare, and eventually decide. That intelligence is more valuable than any keyword volume report. I cover a practical method for mining those patterns in my piece on competitive intelligence.

What This Means in the Age of AI Search

The shift from keywords to intent isn’t just a search engine story. It’s now an AI story, and the implications are even more significant.

AI-powered tools like ChatGPT, Claude, Perplexity, and Google’s AI Overviews are changing how people find and consume information. Instead of returning a list of links for the user to evaluate, these systems synthesize answers directly, drawing on content they’ve been trained on or can retrieve. The user may never click through to your website at all.

This changes the visibility question in a fundamental way. The old goal was to rank on page one of Google. The emerging goal is to be the source that AI systems reference, cite, and draw from when answering questions in your domain.

And what determines whether your content earns that position? Topical authority and genuine depth. LLMs are trained to recognize and surface content that demonstrates real expertise on a subject: content that covers topics comprehensively, uses the full range of relevant language naturally, addresses the questions buyers actually ask, and does so with clarity and specificity. Keyword-stuffed content that was written to game a frequency algorithm has no place in this model. Content that genuinely serves a reader’s intent does.

The actionable shift is this. Instead of asking “what keyword should I target,” ask “what question does my buyer have at this stage of their awareness, and can I answer it more clearly and completely than anyone else?” That answer, written well and structured properly, is what earns visibility in both traditional search and AI-generated responses.

There’s a second dimension worth noting. AI systems give disproportionate weight to content from sources they recognize as authoritative on a given subject. A business with a deep, coherent library of content on a specific topic, where pieces interlink and reinforce each other, signals that kind of authority far more effectively than a collection of unrelated articles optimized around individual keywords.

Topical depth is no longer just good content strategy. It’s increasingly the mechanism of AI visibility.

The underlying principle hasn’t changed. Understand your audience deeply, address their real needs, and produce content that demonstrates genuine expertise. What’s changed is where that content now needs to show up, and how it gets discovered.

The Practical Framework

Putting this together, here’s how to approach content strategy through the lens of intent rather than keywords.

Start by identifying the real problems your audience is trying to solve at each stage of their awareness journey. What does someone think about when they first start recognizing they have a problem? What questions do they ask when they’re actively researching solutions? What objections do they have when they’re close to committing?

Map each of those problems and questions to the intent type it represents. Some are informational. Some are investigational. Some are transactional. Each requires a different kind of content and a different measure of success.

Then research how your audience talks about those problems using their own language, in their own words, at their own level of sophistication. SERP analysis, reading forums and communities where your buyers spend time, and studying the questions they ask in sales conversations are all useful inputs here. The keywords and phrases that emerge from this research are more valuable than any keyword tool, because they come directly from observable buyer behaviour.

Finally, create content that genuinely serves the intent behind each query. Not content that mentions the right words. Content that answers the real question, in the right format, for the right stage of awareness.

If you do this consistently, the SEO signals follow naturally. You’ll use the right language because you understand your audience. You’ll generate long clicks because your content genuinely helps. You’ll build topical authority because your library reflects real depth across a subject area. And you’ll attract the right buyers, at the right stages, rather than generating traffic that goes nowhere.

That’s what modern search visibility actually is. Not a keyword strategy. A buyer understanding strategy.

---

Frequently Asked Questions