Authenticity as a measurable marketing asset in 2026

In December 2025, the Association of National Advertisers named its Marketing Word of the Year and, for the first time in eleven years, picked two: "agentic AI" and "authenticity." One was the problem; the other was, supposedly, the cure. 

As of May 2026, worldwide search volume for "brand authenticity" is up roughly 400% year over year, but you don’t have to go on Google Trends to see this. Open any marketing publication since and the prescription is identical: counter the flood of AI-generated content with a more "authentic" brand. Yet almost no one can tell you what authenticity actually is, let alone how you'd know if you had it.

That is the problem: the industry named the antidote to “AI brands” but left it as a vibe. We'd argue authenticity in branding is both definable and measurable, and that if you're not measuring it, you don't actually know whether your brand is doing the thing the entire industry is now telling you to do.

First: what we mean when a brand "feels AI"

When people say a brand "feels AI" today, they'd have said "feels generic" a few years ago. They're pointing at a constellation of signals that fail to separate one brand from its category: a homepage built on category-wide vocabulary, stock-feeling imagery, claims so slippery they commit to nothing.

AI didn't invent the generic brand, but it made it a lot more widespread. We’ve been hearing “this looks like every other brand in our category" as the most common unfavourable diagnosis for as long as we've been building brands. AI took the existing trend and turned it into a visible crisis, producing artefacts that are structurally sound and substantively empty across the board: content, visuals, even product features.

So what the market is filing under "authenticity" is something more specific: the visible presence of a human, or humans, with a particular point of view behind the brand. Those are the signals that point to the thinking layer underneath the output. When it's there, the work is harder to replicate, and it reads as distinctly itself rather than as a category.

Two attributes consistently sit underneath. Both can be observed, and both can be measured.

Specificity: could this only be your brand?

Specificity is the degree to which a piece of brand expression could only have come from you. Its opposite is interchangeability, the expression lifts cleanly onto a competitor's site and no one notices the seam.

Strip the logo off a homepage hero and drop it onto three other companies in the category. If it survives the transplant, it isn't specific. The same test can be run on taglines, sales decks, social posts, and case studies, which means specificity isn't one number, it's a rate you can observe across a sample. Pull twenty pieces of expression from across a brand’s channels, ask of each whether a competitor could credibly have published it. The share you can confidently say no to is the brand's specificity, and it's a figure that moves: it can be tracked over time and compared channel against channel.

The score tells you how often expressions are interchangeable, but it won’t tell you why. To understand it, you will need to look a little deeper. High specificity traces to one of two sources: a sharp, ownable point of view, or an operational reality a competitor can't claim. It may be useful to ask which one it is in each instance, but the most useful metric is when the answer is “neither.” This usually points to an arbitrary strive for distinctiveness that isn’t rooted in what your brand actually stands for.

A brand can look bold and still be nothing in particular. This is apparent visual or verbal choices that no competitor would make but that are disconnected from a strategic intent. So the specificity worth measuring isn't "how unusual is this," but "how much of this could only be true of us.”

Provenance: can it be traced to a real source?

Provenance is the degree to which a point of view can be traced back to a real human or a real operational fact. Its opposite is the anonymous and the unsourceable, the content now flooding the internet with no clear author or origin behind it.

Where specificity is a rate, provenance is a ladder, and the rungs are what you measure against:

• Named human. Traceable to a specific person who genuinely holds the view and would defend it in a room.
• Operational fact. Traceable to internal data or a working reality the company actually possesses.
• Generic "we." Attributed to a "team" or a house voice with no locatable source behind it.
• Unsourceable. A claim about a trend, written as if from nowhere, that no one inside the company could stand behind.
  ‍

Every owned expression sits on a rung, and the distribution across rungs is the measurement. A channel where most pieces sit on the top two has provenance; one where they cluster at the bottom doesn't, regardless of how polished it reads.

On AI disclosure

Provenance is also where the AI-disclosure conversation tends to land, and where it tends to land wrong. Within the brand authenticity discourse, the conversation usually leans towards arguing that the use of AI in itself is the thing that compromises a brand.

But "did AI touch this" is the wrong axis. Provenance asks whether a recognizable human point of view shaped the work and traces to a real source, and that's independent of how it got made. An AI-drafted post sits near the top of the ladder if a named person holds the view; a hand-built sales deck sits at the bottom if it restates what every competitor pitches. A product decision sourced from real customer behavior has provenance; one copied off a competitor's roadmap does not. Provenance decouples authenticity from the AI tools because the substance is more important than how it got made.

Specificity is the opposite: it's what AI erodes by default. AI output regresses toward the category mean, which is the actual mechanism behind "feels AI": not that a machine was involved, but that the work drifted to the center of the category, where specificity goes to die.

Where the two axes meet, a point of view appears

Specificity without provenance reads as clever positioning that may be backed by nothing real. Provenance without specificity reads as a sincere person saying exactly what everyone around them is saying. Neither is what the market is reaching for when it asks for authenticity.

Both together is what we'd call a brand's point of view: a consistent perspective, traceable to a real source, that no competitor could credibly adopt. It shows up across surfaces and compounds over time, because each piece of expression reinforces the others. It also can't be retrofitted: it's a reading you take off the work, not a coat you apply to it.

Most playbooks today will treat authenticity as a veneer over whatever is already being produced: friendlier copy, more behind-the-scenes content, a founder's LinkedIn presence. None of it is wrong, it’s just downstream. A low specificity rate isn't a copy problem, it's a brand that hasn't decided what it believes that makes it stand out in its category. A provenance ladder stacked at the bottom isn't a tone problem, it's a brand with no real human or fact underneath the work.

You can warm up the copy all you like. Authenticity won’t emerge until you do the hard work of locating what’s underneath.