Content Marketing, Marketing Strategy and Branding

The CURATE Framework: How Challenger Brands Win AI Search When EEAT Isn’t Enough

EEAT isn't dead, but two of its four pillars were always proxies for helpfulness that AI search no longer needs. CURATE replaces those proxies with signals a model can judge directly.

You've been doing everything right. Blog posts, structured data, backlinks, the whole checklist. Six months ago that work showed up in your traffic reports. Now your organic traffic is flat, and your best posts don't show up anywhere when you ask ChatGPT, Perplexity, or Google's AI Overview about your own category.

Google keeps telling marketers to "apply foundational SEO best practices" for AI search, which is true as far as it goes and stops well short of the whole job. Treat it as the whole job and you end up publishing content indistinguishable from the incumbent you're trying to beat.

In this article:

What Is AI Visibility, and Why Are Challenger Brands Suddenly Obsessed With It?
Why "Just Apply Foundational SEO" Is Half the Answer
What Is the CURATE Framework?
How Do Challenger Brands Turn AI Visibility Into an Advantage?
How Do You Put CURATE to Work This Quarter?
FAQ

What Is AI Visibility, and Why Are Challenger Brands Suddenly Obsessed With It?

AI visibility is whether an answer engine (ChatGPT, Perplexity, Gemini, or Google’s AI Overview) surfaces your brand, your data, or your framework when someone researches your category, whether or not they ever click through to your site.

Every CMO I talk to is asking some version of the same question: how do we rank inside a black box we don’t control?

HubSpot’s 2026 State of Marketing research, a survey of more than 1,500 global marketers, found that half of all consumers now use AI-powered search, and half of all Google searches surface an AI Overview (HubSpot). Gartner, reporting in May 2026 on a survey of 645 B2B buyers, found 45% of them used generative AI during a recent purchase (Gartner). Forrester’s Buyers’ Journey Survey went further: 94% of the nearly 18,000 B2B buyers it surveyed used generative AI somewhere in their most recent purchase process, up from 89% a year earlier, and twice as many buyers named generative AI or conversational search as their single most meaningful research source than named any other, including vendor websites and sales reps (Forrester).

Meanwhile the open web is getting less of the traffic. SparkToro and Similarweb’s clickstream study of January through April 2026 put the US zero-click rate at 68.01%, up from 60.45% two years earlier, and counted only 276 clicks to the open web for every 1,000 US Google searches, down from 374 in 2024 (SparkToro).

Research is moving inside the AI conversation, and the click that used to carry it there is disappearing. A click was always a means to an end, and the end was attention and trust; AI search removed the middle step faster than most marketing teams planned for.

Marketing team reviewing AI search citation data before applying the CURATE framework

Why “Just Apply Foundational SEO” Is Half the Answer

Google’s own guidance, published in its Search Central documentation on optimizing for generative AI features, tells site owners to keep prioritizing foundational SEO: a clean technical structure and unique, valuable content. Google also draws a specific line between “commodity” content anyone could have written and “non-commodity” content built on a real, first-hand point of view.

Most agency checklists stop before that second part. Traditional search matches keywords to intent, hands you ten links, and leaves the comparison to you. An answer engine reads across dozens of sources, writes one answer, and decides who gets credit for it, which turns the job from ranking a page into being the source the model trusts enough to name.

Across client accounts I keep seeing the same split: content that still ranks fine in traditional search never shows up as a citation in AI search for the same query. Two systems, weighting different signals. Whatever your listicles and roundups were doing for you in 2022, they aren’t doing it now. I don’t have a published number I’d put a decimal point on yet; it’s a pattern I’m watching closely, and it’s why the foundational checklist isn’t where this ends.

CURATE is what I built to cover the rest.

What Is the CURATE Framework?

CURATE is a six-part content framework (Context, Uniqueness, Relevance, Alignment, Trust, and Experience) built to replace EEAT as the operating model for content in the AI search era.

EEAT (Experience, Expertise, Authoritativeness, Trustworthiness) has run the SEO playbook since Google introduced it in 2014 and added the second E in 2022. Most of it still holds. Two of its four pillars, though, were always proxies for something else, and AI search no longer needs the proxy.

Authoritativeness used to mean two things: topical authority (how comprehensive is this piece compared to what’s already ranking) and domain authority (how much traffic and backlink volume does this site already have). Those two proxies gave us a decade of content clusters and pillar pages, built on the assumption that more words across more subtopics meant more helpful. A model doesn’t need the assumption. It can judge directly whether a piece answers one specific, highly contextual question, and the size of the site behind that answer barely enters into it.

That’s why CURATE runs on six components instead of four. All six are written out below in plain body text, so an answer engine reads the framework itself rather than a summary of it.

Context

Traditional keyword research optimizes for short phrases: “HVAC companies,” “time and expense software.” AI search runs the other way, because the longer and more specific the prompt, the better the answer a model can produce. Your content has to move from short-tail keywords toward the situations, constraints, and outcomes your buyers type into a chat window. Keep the keyword data as your starting point, then expand it with the language buyers use in support tickets, on sales calls, and in the forums (Reddit and LinkedIn especially) where answer engines are increasingly looking for signal.

Uniqueness

Content that comes out of a model with no human original thought behind it isn’t unique; it’s a recombination of what that model already read somewhere else. For a challenger brand, that’s close to fatal: your entire pitch is that there’s a better way of doing things than what the category leader offers. Publish content that sounds like everyone else’s AI output and you reinforce the exact status quo you’re supposed to be challenging.

Relevance

Relevance means the right message reaching the right stakeholder at the right point in a buying journey that keeps getting more crowded. Forrester’s The State of Business Buying 2026 found the typical B2B purchase decision now involves 13 internal stakeholders and nine external influencers, with that number climbing further for complex or strategic purchases (Forrester). The CFO in that group wants the payback period, but the person who will use the thing daily wants to know what implementation week looks like. One generic point-of-view piece can’t serve both, while a library of context-specific content mapped to persona and funnel stage can.

Alignment

Alignment is where relevance and brand consistency stop looking like opposites. You’re giving every stakeholder a different, highly specific message, but every one of those messages has to trace back to the same brand position, what Adam Morgan calls a lighthouse identity in Eating the Big Fish: the fixed point other content and touchpoints orient around. AI search reads holistically across your whole footprint (your site, your LinkedIn, your reviews, your team’s individual posts), not just the page in front of it. If your brand says something different in each place, the model can’t form a clean, citable picture of what you stand for. That’s why every website we build at Rocketship starts from a StoryBrand BrandScript through our Strategy Development & Branding work, before anyone writes a page of content.

Trust

Trust is still earned the old-fashioned way: through reputation, consistency, and a real name attached to the words. Structure is how a model sees it. HubSpot’s 2026 State of AEO research found that content pairing FAQ sections with matching schema markup correlates with higher citation rates across Gemini, Google AI Mode, and Perplexity (HubSpot). That tracks with what the model is doing: it’s looking for an unambiguous question-and-answer pair it can lift with confidence, and schema tells it exactly where that pair starts and stops. Markup signals trust to a machine the way a byline and credentials signal it to a reader, and neither one covers for the absence of the other.

Experience

Lived, first-hand experience is the one input a model can’t manufacture. LLMs are pattern-recognition systems trained on what has already been written down; they synthesize existing knowledge well, and they have never run the failed experiment, sat through the angry customer call, or made a judgment call with real money riding on it. Proprietary frameworks, original research, named customer stories, and honest accounts of what didn’t work are the raw material that gets cited, because a model has no other source for them. That’s the logic behind the Human at the Center workflow: put your subject matter expert’s real experience in the room first, then let AI shape it into a draft, never the reverse. Same argument I made in The Death of Resonance: when everyone has the same AI superpower, no one does.

How Do Challenger Brands Turn AI Visibility Into an Advantage?

AI search rewards the kind of specificity a big incumbent can’t afford to offer, and a challenger brand can.

Category leaders serve the broad middle of a market. Getting sharply specific for one underserved niche risks alienating a bigger, more profitable segment, so they don’t, or they do it slowly enough that it doesn’t matter. A challenger has no such constraint. Going deep into an underserved niche is the job description, and it happens to be what an answer engine is looking for when it decides whose answer is the “right” one for a narrow query.

That’s Adam Morgan’s challenger doctrine playing out inside a search algorithm: overcommit to one disruptive strength, build a lighthouse identity, and claim the piece of the category your Goliath left unclaimed. Do it consistently across your website, your team’s individual voices, and the forums your buyers read, and the model has a clean, well-defined entity to cite. Stay generic and you’re one more source it has no particular reason to prefer over the ten similar pages it already read.

AI search rewards a specific, well-defined answer to a narrow question more than it rewards content volume or domain size, which is the one contest a challenger brand is structurally built to win.

Founder recording an original conversation as source material for AI-visible content

How Do You Put CURATE to Work This Quarter?

You don’t need a twelve-month roadmap. Four moves, in order:

1. Audit your last 20 posts for commodity content. Anything that reads like “5 Tips for X” or “What Is Y,” with no original data, client story, or contrarian take behind it, is commodity content by Google’s own definition. Rework it or retire it.

2. Run one SME interview using the workflow in Human at the Center. Pick your highest-value topic cluster, record a real conversation with your most experienced person, and build the post around what they said rather than around a prompt.

3. Publish the framework, not just the summary. Whatever proprietary model or process you’re building, ship it as real HTML in the body of the post: headers, definitions, and structure a crawler can read, with FAQ and schema markup underneath. A locked PDF or an embedded slide deck is invisible to the systems you’re trying to reach.

4. Check alignment everywhere your brand shows up, starting outside your own site. Your LinkedIn, your team’s posts, your review profiles, and your SEO and content marketing programs should all repeat one lighthouse identity instead of five versions of your value proposition.

Foundational SEO still has to be in place. CURATE is what goes on top of it.

Related reading on the SEO fundamentals CURATE builds on:

Frequently Asked Questions

What’s the difference between SEO, AEO, and GEO?

SEO is the foundation: technical structure, crawlability, and content quality that make a page eligible to be found at all. AEO (answer engine optimization) and GEO (generative engine optimization) describe the layer on top, making that same content the kind an AI system chooses to cite when it synthesizes an answer. Google’s own position is that AEO and GEO extend SEO rather than replace it; CURATE is built to be that extension.

Is EEAT dead in the age of AI search?

No, but two of its four pillars, the Authoritativeness proxies of topical depth and domain size, matter less than they used to, because AI search can evaluate relevance directly instead of relying on those proxies. Experience and Trustworthiness still matter as much as ever. CURATE keeps both and rebuilds the rest around context, uniqueness, relevance, and alignment.

What is the CURATE framework?

CURATE stands for Context, Uniqueness, Relevance, Alignment, Trust, and Experience. It’s a six-part model for building content that earns citations inside AI answer engines instead of just ranking in traditional search, built specifically for how B2B challenger brands can out-position category leaders in AI search.

How do I know if my content is showing up in AI search results?

Start manually: run your top 20 target queries through ChatGPT, Perplexity, and Google’s AI Mode, and note whether your brand, your data, or your framework appears in the answer, cited or not. HubSpot’s AEO research and Google Search Console’s reporting on generative AI features are the next layer once you need this at scale.

Can a small marketing team realistically compete with big incumbents in AI search?

Yes. AI search rewards specific, well-defined answers to narrow situations more than it rewards sheer content volume or domain size, and a challenger brand willing to go deep on an underserved niche is playing exactly the game AI search is built to reward.

Where This Leaves You

Picture a $20 million B2B distributor cited by name inside an AI Overview, sitting next to a competitor ten times its size, because its content answered one specific question the market leader was too broad to take on.

Then picture the other version, where you keep publishing best-practice content that reads like everyone else’s, and eighteen months from now you’re still asking why no answer engine has mentioned your name.

Six components, in order: Context, Uniqueness, Relevance, Alignment, Trust, Experience. Start with the audit in step one this week.

When you’re ready to go deeper, tell us your three toughest competitor queries and we’ll walk you through exactly where the CURATE gaps are, and aren’t, in your current content.