Every day, your customers write you thousands of creative briefs. They’re short. They’re brutally honest. And they tell you exactly what they want, in their own words.
They’re called search terms — and most advertisers throw them away.
Someone who types “waterproof hiking boots for wide feet” into Google isn’t browsing. They’ve told you the product, the use case, and the thing they’re worried about, all in seven words. That’s not a keyword. That’s a brief. And what do most accounts serve back? The category-level ad: “Shop Quality Footwear. Great Prices. Free Shipping.”
The gap between what someone searched and what your ad says is the most expensive blank space in paid media. Let’s talk about how to close it.
Your account is full of briefs nobody reads
Here’s an exercise we run with almost every brand we work with: pull your search term report, pick twenty high-value queries, and put them side by side with the ads that actually served against them.
The pattern is nearly universal. The queries are specific — a product line, a problem, a comparison, a location, a “for” clause (“…for sensitive skin,” “…for small businesses,” “…for beginners”). The ads are general. One set of headlines, written months ago, doing its best to be all things to all searchers.
This isn’t because anyone is lazy. It’s math. A brand with 40 product lines, running campaigns across hundreds of ad groups, would need thousands of pieces of tailored copy to genuinely answer every intent in the account — and it would need to refresh them as search behavior shifts. No human team can write at that scale. So we compress: one message per campaign, maybe per ad group, and we let “relevance” mean “technically related.”
What this means for you: if your brand sells more than one thing — multiple product lines, multiple services, multiple programs — your account isn’t one business. It’s dozens of micro-businesses sharing one voice. Every intent that gets the generic ad is a brief you received and didn’t answer.
The multi-product brand feels this the most
For a single-product company, one great ad can carry a lot of weight. But the more you sell, the more your search demand fragments into distinct intents — and the more expensive generic copy becomes.
Think about a home services company that does roofing, gutters, siding, and windows. Or a university with fifty degree programs. Or a retailer with a catalog that runs from cookware to camping gear. Behind each of those lines sits a completely different customer, with a different problem, a different vocabulary, and a different objection to overcome:
- The searcher typing “emergency roof leak repair” needs speed and reassurance.
- The one typing “how much does siding replacement cost” needs transparency and a reason to trust your estimate.
- The one comparing “online MBA vs in person” needs a point of view, not a slogan.
Serve all three the same “Trusted Local Experts — Get a Free Quote” headline and you haven’t advertised to any of them. You’ve advertised near them.
What this means for you: the richer your catalog, the bigger the payoff for copy that meets each intent on its own terms — and the harder it is to get there manually. This is precisely the problem shape where AI-generated, feed-grounded creative earns its keep.
What “answering the brief” actually looks like
So what does it take to treat the search term as the brief? Three ingredients, and all of them matter:
Grounding in your actual product data. Copy that answers “waterproof hiking boots for wide feet” can only be written by something that knows you stock wide-fit waterproof boots, what they’re called, and what makes them good. That knowledge lives in your product feed, and your site — which is why intent-matched creative has to be generated from your data, not from a template with the keyword bolted in. Old-school keyword insertion put the searcher’s words in your headline; feed-grounded generation puts your answer to their words in the headline. Those are very different things.
Fluency in how people actually buy. A great brief response doesn’t just repeat the request back — it persuades. The craft that good copywriters apply instinctively (address the hesitation, make the value concrete, give a reason to act now) can be applied systematically, at scale, across thousands of ad variants. The searcher worried about “wide feet” gets reassurance about fit. The one searching “cost” gets clarity. The comparison shopper gets a difference worth remembering.
A human hand on the wheel. Scale without control is how brands end up with ads they’ve never read. Every piece of generated copy should be reviewable — on-brand by rule, approved by a person, and visible to the advertiser at all times. If AI is writing in your brand’s voice, you should be able to open the book: see what’s being said, where, and why. Glass box, not black box.
This is the philosophy behind ad-machina, our AI creative engine: it turns your product data, brand rules, and real search behavior into thousands of on-brand, intent-matched ads — written by a team of specialist AI agents, reviewed by humans, and improved month after month as the search data comes back. Not one campaign message stretched across every query. An answer for each brief.
“But isn’t Google’s AI doing this for me now?”
Fair question — and probably the one your CMO is asking. Google’s AI Max expands your reach into queries you’d never have covered and generates assets automatically. Google reports that AI Max delivers meaningfully more conversions than manual campaigns for advertisers who adopt it, and we’ve said it before on this blog: you should absolutely be testing it.
But here’s the thing about generative systems: they generate from something. AI Max writes and adapts based on the substrate you give it — your site, your feed, your existing ads. Feed it generic, category-level copy, and it will faithfully iterate on generic, category-level copy, just across more queries. Feed it specific, intent-matched, feed-grounded creative, and it has dramatically better raw material to learn from and build on.
In other words: AI Max isn’t your competitor. Weak inputs are.
We think about it as two engines doing two different jobs. Platform AI is the distribution engine — exploration, reach, the long tail of queries you can’t predict. Your creative layer is the generation engine — precision on the searches that matter most, with copy strong enough that the platform’s AI gets smarter by referencing it. The classic pattern still holds: a small share of search terms drives the large majority of revenue. Those head terms deserve deliberate, human-validated, continuously improving creative — and everything the platform builds on top inherits that quality.
What this means for you: the question isn’t “AI Max or crafted copy?” It’s “what is AI Max working from?” The brands getting the most out of platform AI are the ones investing in the creative substrate underneath it — with the control and transparency to know exactly what’s being said in their name.
Our honest take
Search didn’t just change once — it changed twice. People now search in full sentences and expect answers as personal as the question. And the platforms answered with automation that expands reach but writes generically by default. The advertisers winning in this environment aren’t choosing between human craft and AI scale. They’re putting craft into the scale: treating every search term as the brief it is, answering it with copy grounded in what they actually sell, and keeping a human hand on the approval button.
Your customers are already telling you exactly what they want. The only question is whether your ads are listening.
Want to see what your search terms have been trying to tell you? Get in touch — we’d love to take a look at your account with you.
Cookie configuration