What makes content work in the age of average?

Cover art made with AI. Every detail, from the palette to the bite, was decided before the first prompt.

When content costs nothing to make, intention becomes the advantage.

Picture an apple. What color is it? Is it on a table, in your hand, on a tree, in a store?

Ask a designer to build a concept around that same apple, and they start where designers always start, with inspiration and context. They might look at what the apple means to them, its color and shape, perhaps its cultural relevance to New York or a teacher’s desk, and design toward its appeal. Ask an architect, and they see structure: an opaque envelope, a malleable interior that can be shaped for a variety of internal programs, and a central core for circulation through the space.

Trained in both, you see the two views merge into one system, where how it looks and how it’s built get decided together. Then someone takes a bite, cutting a section through the envelope to expose the interior. What was an object becomes something consumed, and it carries a mark of where they went.

Now we’re not talking about an apple at all, are we?

We’re talking about every piece of content your team puts in front of a buyer, and what happens once they’re inside it.

The cost of making content has fallen toward zero. Buyer attention costs as much as it ever did.

Ten years ago, I was handed an interactive content tool and asked to make magic, an assignment most designers know well. Which tool matters less than what it taught me: interactive content could tell me what people did inside it, something a flat document never could. The designers who do this work well learn that the principles are about behavior, and that understanding arrives long before the tool does.

Every bite into digital content leaves a record. The industry loosely calls it analytics, and most teams use it only to confirm what was opened, but it holds far more. When the content is designed to reveal it, you can see the other side of intention: what buyers chose to explore and how they found their way through it. You’ll find the section you were sure would matter and everyone skipped, along with the order people followed your calls to action, or didn’t. Architects read the same evidence in desire paths, the worn lines across a lawn where the sidewalk should have been. The good ones wait to see where people walk, then pave the path they’ve already chosen.

Build with that evidence in mind and read it closely, and the work compounds. Each piece starts from what the last one taught, and before long the analytics stop answering design questions and start answering business ones: what to expand, what to pull back, where the next conversation should start. Over time the lessons outgrow any single piece and become a way of seeing you bring to any tool.

The age of average

Now the tool is AI, and designers are getting the same assignment.

The cost of making content has fallen toward zero. Buyer attention costs as much as it ever did, and that gap is where average does its damage.

Every model builds from what already exists. Its training spans the most prolific work in human history, from Plato and Confucius to Bach and Nina Simone, and, squarely on the other side, the consensus summary of Wikipedia. Ask it for an interactive page and it returns the middle of that spectrum in the time it takes to describe one. The years of watching people use what you built don’t come with it.

Most generated pages look the same: dark planes, stacked cards, a button in the corner that says Explore, and nothing underneath that a printed slide couldn’t hold. They borrow the look of depth and apply one answer to every section.

What follows is what the work taught, condensed, starting with the lesson that came first.

Design changes what people do

The surface earns the first look, but the structure decides what a buyer does next.

The graphic above takes a static deck and gives each section what it needs. Some sections earn interaction, while others need reordering, or a different format entirely. Only one of the three lets the buyer choose their own path, but each change shifts what the buyer does, and the idea is older than any screen.

Early in World War II, experienced B-17 pilots kept reaching for the flaps after touching down and raising the landing gear instead, dropping their bombers onto their bellies on the runway. The crashes were blamed on pilot error until a psychologist named Alphonse Chapanis looked at the cockpit instead. He found two identical controls side by side and gave each a different shape, and the errors nearly disappeared. The B-17 was a beautiful airplane, and none of that beauty kept a pilot’s hand off the wrong lever. The fix met the pilot where the work happened: at the lever, in the few seconds that mattered.

Sales content has its own version of pilot error. When a buyer never reaches the next step, the usual diagnosis is a disengaged buyer or a slow seller. More often, the next step sits on the last slide in the same type and weight as every bullet around it, one line among many that look alike. Give it its own shape and its own place on the buyer’s path, and every click on it becomes a signal a seller can act on. Expedia Group saw what that signal is worth on a competitive RFP: its sellers walked into the meeting already knowing what the buyer cared about, and when the partner chose Expedia, it credited the format as a key reason why.

What the work kept showing

In the cockpit and in the deck, the fix came from looking at the system instead of the surface. Across 55 million sessions, the Microapps that outperform the rest share the same pattern. They have depth, and depth has three parts.

• Layering stacks detail on top of the presentation, anything from a single point of detail to a layer that responds to the viewer. The viewer opens what they want more of, closes it when they’re done, and never loses their place.
• Branching turns navigation into a path the viewer chooses. Content about your company is still written for the buyer, and when they choose the path, the conversation leads the content instead of following it.
• Persistence keeps the Microapp working after it’s sent. It lives at one link, so the experience can evolve without sending a new file, while reporting what people opened, skipped, and lingered on.

Content that works needs all three, and each section still earns its own decision. Interaction applied everywhere is the same mistake as interaction applied nowhere.

Knowing where to hold back is the harder skill, and the greatest creators have always treated restraint as the discipline. In the age of average, anyone can produce consumable volume, and quality depends as much on what you leave out as on what you put in.

The latest tool

Ten years ago, I was handed a tool, and in those same years, teams trusted Tiled with more than 83,000 pieces of static content to turn into something people explore, each one leaving a before and an after. Where those paths intersect, one lesson has held: the tools kept changing, from Sketch to Adobe XD to Figma and now to AI, while the way people move through content stayed the same. AI is the latest, and like every tool before it, its value depends on the person directing it. Today, that person could be almost anyone.

In the urgency to ship, average becomes the default, and it scales faster than any team can review it. Most of the conversation about AI and design leans heavily on taste and judgment as the answer; both matter in their own way, but both are symptoms of something earlier.

Intention comes first, and it only works when paired with context. No matter how capable AI becomes, it still doesn’t know who this buyer is or where this conversation stands. The context lives in the record, in what buyers came for and the paths they chose. Across every before and after, Tiled has learned what those paths tend to have in common, and it’s turning that understanding into the lens AI works through before it builds anything.

Deciding what a specific piece has to do, however, is still a leadership call. It belongs to the people who know how buyers move, and it has to be made before the first prompt.

Getting it right matters more now, because the window keeps shrinking. Gloria Mark at UC Irvine has tracked how long people hold attention on a single screen for two decades, and it has fallen from about two and a half minutes to 47 seconds. Content has less than a minute to earn its place. Intention sets what a piece has to do in that minute, and structure is how it does it.

Designers know better than anyone that a system is never finished. Most of us have a story about the better idea that arrived mid shower, days after the project shipped. Interactive content still has room for it, and the next bite someone takes tells you whether it was right.

Once you see content this way, you can’t unsee it.

Originally published on LinkedIn on October 1, 2026.