Narrative Architecture Is Simple. Building It Is Not.

Why is a simple idea so rarely built?

A marketing lead can describe the goal in a sentence: when a customer asks an AI about our category, we want it to name us. Almost everyone who hears it nods. Almost no one has done it.

That gap is the interesting thing — not the goal, which is obvious, but the distance between grasping it and having it. The reason the distance matters more now than it did three years ago is that the place where buyers decide has moved. A 2025 Pew study found people click a result just 8% of the time when an AI summary sits on top of it, down from 15% without one, and by early 2026 roughly two-thirds of Google searches ended with no click at all. The answer has become the destination. So the simple sentence — be the one it names — is now most of the strategy. Which makes the gap between saying it and doing it the whole problem.

What does the napkin actually leave out?

Sketches are honest about shape and silent about labor. They show the idea and hide the hours.

An architect’s first sketch shows a building anyone can read at a glance — this wing here, light from the south, an entrance there. What it omits is everything that makes the building stand and worth standing in: the load paths, the tolerances, the thousand decisions that turn a shape into a structure. The sketch is true; it is just not the work. The same is true of the answer-first idea. “Be the brand AI cites” is the shape. The work is naming, for every page you publish, the exact question it answers and the part of the answer it is meant to move — and then doing that consistently enough that an engine can trust the whole. The sketch makes that sound like a detail. It is the job. Consider one ordinary page — a product explainer. On the napkin it is a single box. In the building it is a decision about which buyer question it owns, a first sentence that answers that question plainly enough for a machine to lift, a claim stated so it can be confirmed elsewhere, and a tone a person will stay for. Multiply that by every page a brand publishes, and the gap between the drawing and the structure stops being a metaphor.

Where do most teams actually fall in?

Not at the idea. They understand the idea. They fall in at the part where understanding has to become a hundred small, consistent decisions.

The common failure is to mistake the sketch for the work and reach for volume — more posts, more pages, more press releases — as if mass were the same as structure. It is not. A pile of materials on a site is a hazard, not a building, and a heap of content is not a story an engine will repeat. What the evidence rewards instead is substance a machine can verify. The Princeton team that first named generative engine optimization found that adding citable statistics and sources raised a page’s AI citations by about 41%, while keyword stuffing — the reflex inherited from search — did nothing or slightly hurt. The lesson is unglamorous: not more content, but content where every piece is doing a job. That is harder than volume, which is exactly why volume is the path most teams take.

Why can’t you skip the survey?

No architect designs blind. She measures the ground first — the soil, the sun, what is already next door — and the design answers what she finds.

The equivalent here is the least exciting and most skipped step: looking at the answer landscape before building into it. What does AI already say about your category? Who does it name? Which sources does it trust? That ground has a known shape. A 2025 Semrush analysis of more than 150,000 citations found engines lean heavily on third-party sources — forums, encyclopedias, the press — over a brand’s own pages, and Muck Rack found earned media drives roughly 84% of AI citations. So the ground you are surveying is mostly other people’s coverage of you, which tends to be sobering the first time anyone looks. Skipping the survey is the tempting shortcut, because the survey is the part that produces no visible output. It only tells you the truth before you spend money against it.

Isn’t simplicity supposed to make things easier?

It makes the idea easier to carry. It does not make the idea easier to build — and confusing the two is the costly mistake.

A framework that cannot be explained in a minute is one no one will use, so the simplicity is a feature; it is how the idea travels through an organization. But a simple idea and a simple execution are not the same sentence. Resolving the conflict on every page — structured and citable for the machine, genuinely worth reading for the human, in the same asset — is its own quiet discipline; we call it the Dual Legibility Tension, and naming it does not make it less work. This is the ordinary shape of any craft. The principles of cooking fit on a card. Restaurants are still hard. The difficulty was never in understanding; it was always in the doing, repeated until it holds together.

So what does the doing look like?

It looks like four steps anyone can list and few will sustain — which is, again, the entire point.

Survey the ground, so you design to terrain that exists rather than the one you imagine. Design each piece with a purpose, so nothing you publish is idle weight. Build toward one structure, so the parts corroborate each other instead of competing. Then keep measuring — how often, of the questions a buyer actually asks, the engines name you — and feed what you learn back into the next piece. That is the whole method, and it is short enough to fit beside the napkin. We do this for brands through White Wood, but the steps are the same whether anyone is hired or not. The advantage was never secret knowledge. It is that the work is patient and unglamorous, and most people, having understood the idea, assume they are nearly finished. They are at the beginning. In practice the first survey is usually the moment a team realizes how much work the napkin hid: they ask the engines the questions their own customers ask, and find a competitor — or no one — named where they assumed they would be. That is not a failure of the idea. It is the idea finally meeting the ground, which is where every build actually starts.

Frequently asked questions