Optimizing for Meaning, Not Matches: Why the Old Search Playbook Is Broken

Media Minute is WPP Media's series, specifically crafted to empower our clients and marketers for their intelligence era.

There's a certain kind of search marketer who knows what a SKAG is. If you were deep in the paid search trenches in the 2010s, you know exactly what we're talking about: Single Keyword Ad Groups. The practice of isolating your highest-value exact match keywords, writing custom ad copy for each one, setting individual budgets, and squeezing every last drop of performance out of hyper-granular keyword data. 

It was meticulous. It was data-driven. And for a long time, it worked. 

The difference between "Cottage Rental Ontario" and "Cottage Rentals Ontario" actually mattered. Clickstream data told you which form of the query was worth more. You'd build content and campaigns around both, optimized to the letter — literally. 

That playbook is now obsolete. And if your team is still running it, you're not just leaving performance on the table. You're actively optimizing for a version of search that no longer exists.  

What Changed (And Why It Matters) 

Two things happened simultaneously, and together they've fundamentally rewritten the rules of search. 

First, the technology evolved. Match types in paid search have shifted dramatically. True exact match, the kind where only that precise query triggers your ad, is largely a relic. Close variants, semantic matching, and misspellings are now folded in automatically, because search platforms aren't asking "did the user type these exact words?" They're asking "what is this person actually trying to find?" 

That shift from lexical to semantic matching isn't a bug. It's a feature and it reflects a deeper truth about how people are searching today. 

Second, query behavior changed. The volume of searches consisting of four words or fewer is declining in many verticals. Searches for luxury vehicles, for example, are seeing shorter queries shrink while longer, more descriptive queries grow. Meanwhile, searches that are five-plus words — full questions, detailed descriptions, conversational prompts — are growing significantly faster. 

People have stopped typing keywords. They've started having conversations.  

The Cottage Rental Isn't Dead. The Query Just Grew Up. 

Here's what that looks like in practice. Ten years ago, someone planning a summer trip might have searched: 

"Cottage rentals Ontario" 

Today, that same person is typing (or speaking) something closer to: 

"I'm looking for a pet-friendly cottage in the Muskoka area that's available during summer break and has direct sandy-bottom access to a lake." 

That's not a keyword. That's a request. And it's being answered, not with a list of blue links to wade through, but with a direct, curated response.  

This shift happened because the technology finally caught up to what users always wanted to do. In the old world, you'd type a short query, get imperfect results, tack on a modifier, search again, narrow it down, iterating your way toward what you were looking for. People learned to speak the language of the search engine. 

Now, the search engine speaks theirs. 

Multimodal Search Made It Even More Complex 

And it's not just about longer text queries. Search is now multimodal meaning people are searching with images, not just words. 

Think about Circle to Search on Android, Google Lens, or Pinterest's visual search. Someone scrolling Instagram sees a celebrity carrying a beautiful handbag. They don't type anything. They screen-grab the image and search with it. In the background, the search engine converts that image into a rich, detailed text description: identifying the bag, its color, the leather type, the brand. Then serves up results that answer questions the user hasn't even fully formed yet: Where can I buy this? What are similar options under $750?  

The input is visual. The underlying mechanics are still semantic. And the intent is what the system is actually optimizing for. 

What This Means for Your Strategy 

If you're still building paid search campaigns around tight, exact-match keyword lists, or writing SEO content to rank for hyper-specific two-word terms, here's what's likely happening: your results look like search is declining. But search isn't declining. You're just limiting yourself to the slice of search that is shrinking. 

The shift required is both strategic and structural: 

• On the paid side, it means embracing broad match and AI-powered tools like Google's AI Max, which leverage your site content, landing pages, and thematic signals to match against relevant intent — not just specific strings of text. 

• On the organic side, it means moving from keyword-by-keyword optimization to thinking in topics, categories, and intent clusters. What questions does your audience ask? What themes does your brand genuinely have authority on? Build for those. 

In short: stop asking "what keywords do I want to rank for?" and start asking "what questions do I want to answer, and for whom?" 

The Bottom Line 

The brands that will win in search aren't the ones who can best engineer an exact-match keyword list. They're the ones who most clearly understand the meaning behind what their audience is asking across every format, every platform, and every stage of discovery. 

Granular keyword optimization was never really about the keyword. It was always about the intent behind it. The tools have just finally evolved to let us optimize for that directly. 

It's time to catch up.