There was a brief, giddy moment when generative AI felt like free money. The marginal cost of a blog post, a banner, a product description, an email fell to roughly nothing. For anyone who has ever stared down a content calendar at six in the evening, this looked like deliverance. Publish more. Publish faster. Fill every channel, every week, forever.
The trap was hiding inside the word “more.”
New data from DoubleVerify, the media-measurement firm, suggests the audience has already adjusted to the flood. Its July 2026 study, Global Insights: Media Quality in the Age of AI, surveyed 22,000 consumers and more than 2,000 marketers across roughly twenty markets. The finding is not that people hate AI. Most of them rather like it. Sixty-three percent say AI-powered tools improve their online experience. The finding is what happens when the output is bad. Forty-two percent of consumers say a brand’s low-quality or “uncanny” AI advertising would actively worsen their opinion of that brand. Forty percent say polished, professional AI work improves it.
Read that gap twice. Same technology. Opposite outcomes. The variable is not the machine. It is the care taken with it.
The cost of cheap moved
For twenty years, the unspoken deal in Digital marketing was that volume was virtuous. More pages, more posts, more touchpoints. Reach was a numbers game, and the numbers rewarded whoever showed up most often.
Generative AI broke that deal by making volume trivial, and in doing so it quietly striped volume of its value. When everyone can produce everything, producing a lot is no longer evidence of anything. It is just noise wearing a logo.
DoubleVerify’s chief executive, Mark Zagorski, put it plainly: “AI itself is not what determines engagement. The quality of the output does.”
The uncomfortable part sits on the buyer’s side of the table too. Fifty-six percent of consumers admit they cannot reliably tell what is AI-generated and what is not. That might sound like permission to cut corners. It is the opposite. When people can no longer verify authorship, they stop judging on origin and start judging on feel. Does this seem considered, or does it seem churned out? The tell is not a watermark. It is whether the thing was worth making.
Slop has a price, and marketers are flinching
The same study found 53% of marketers are concerned about their ads appearing next to low-quality AI content, against 45% for high-quality content. Nearly half, 48%, are uneasy about using AI to build the creative itself.
That hesitation is not technophobia. It is Digital Maturity showing up as judgment. The mature response to a tool that makes bad work cheap is not to make more bad work cheaply. It is to raise the bar the tool has just made easier to clear.
Because here is the second-order effect that never made it into the sales deck. The same AI that lets you flood the zone lets your competitors flood it too. The floor rises for everyone. What passed for “good enough” content two years ago is now the wallpaper the whole market scrolls past without a flicker of recognition. In a world of infinite production, standing still is indistinguishable from falling behind.
Volume never mattered
Here is the reframe worth sitting with. Speed of production was never the thing holding good marketing back. Judgment was. Knowing what to say, to whom, and why it should matter to them. Knowing what to cut. Knowing when a piece is finished and when it is merely done.
AI collapses the cost of production and leaves the cost of judgment exactly where it was. Which means the scarce, valuable thing was never the writing or the designing. It was the deciding. The part a model cannot do for you, because it does not know your customer, your market, or the difference between filling a slot and earning attention.
This is why Digital Engagement has quietly stopped being a volume metric and become a trust metric. You do not build a relationship by being the loudest voice in a feed of synthetic sameness. You build it by being the one piece that was obviously made by someone who cared, for someone specific.
The one-question maturity test
There is a simple diagnostic buried in the DoubleVerify numbers. Ask it of any piece of work about to go out the door: if a customer could not tell whether a human or a machine made this, would they be glad it exists, or would it cost you a little of their regard?
If the honest answer is “cost,” no amount of volume fixes it.
That question, asked before publication rather than after, is most of what separates the firms using AI to get better from the firms using it to get louder. It is the difference between a Digital strategy and a Digital output. It is, not incidentally, what our PIE Framework exists to force: a deliberate check on whether the work is worth the audience’s time before it ever reaches them.
The tools got cheaper but, attention did not.
Essentiate brings the judgment that volume cannot buy. We treat AI as leverage on good thinking, never a substitute for it, which is why the work that leaves our desk is built to earn regard rather than fill a calendar. If your marketing has more output than it has meaning, that is the conversation we would like to have.
