Digiday Says Brands Are Paying a Premium for What AI Can't Fake. Nine of My Last Twelve Posts Would Fail the Test.

Digiday Says Brands Are Paying a Premium for What AI Can't Fake. Nine of My Last Twelve Posts Would Fail the Test.

Digiday published a piece on September 2 with a headline that should make anybody who publishes under their own name a little nervous. Brands are shifting real budget toward Gen X and elder millennial creators, not the twenty-two-year-olds with the ring lights. Alyssa Mercante's reporting cites a WordPress VIP survey that found 60 percent of U.S. consumers call visible AI use in brand messaging a turnoff. The pitch for the older creator isn't nostalgia. It's that a person whose feed carries a mortgage, a health scare, or a mid-career pivot is reporting something a model can't fabricate.

Lily Comba, who founded the influencer agency Superbloom, put the shift better than anybody else Mercante talked to. Authenticity used to mean who could hit the talking points hardest in their own voice. Now, Comba says, it's closer to whether the person feels safe enough to believe. Jo Wong, chief revenue officer at the creator commerce platform POP.STORE, called it a return to basics, the original roots of the creator economy before anybody with twenty bucks a month could generate a flawless talking head.

That's the read. I turned that same test on my own site, and it did not go the way I wanted.

The test

The premium Digiday is describing isn't really about age. It's about whether the person on the page actually paid for what they're describing, a number, a name, a date somebody could go check. So I pulled the last twelve posts live on gallucci.net, not the daily drafts still sitting in review, and read every one against a single question. Does this post contain a named, dated, verifiable fact that only somebody who did the thing could report? Or is it a framework wrapped around a case study with the serial numbers filed off?

Three passed. A review of Amanda Hess's memoir Second Life that cites her actual Goodreads rating, 4.06 across 1,823 ratings, and quotes specific reviewers by name. A breakdown of Super Coffee that names founder Jordan DeCicco, a former Philadelphia University basketball player, and stacks his 20 grams of protein against Fairlife Core Power Elite's 42. A read on Adam Aleksic's book Algospeak that opens with its exact Goodreads number, 3.70 stars across 5,375 ratings, and argues with named reviewers using their own quotes.

Nine failed. And they failed the same way every time.

Nine ways to say the same thing

Why Most Food Branding Agencies Miss the Real Problem, from July 15, opens with "a direct-to-consumer natural chip brand" that dropped $120,000 on a rebrand, name never given, and calls its own three-layer framework the Authenticity Audit. Three days earlier, a different post about a CPG marketing agency built a completely different four-phase framework and named its second phase, no joke, Authenticity Audit. Same two words, two different frameworks, three days apart, and neither post mentions the other.

Between July 10 and July 27 I named seven proprietary-sounding frameworks across nine posts. Founder Velocity. The Founder-First Framework. The Authenticity-First Rule. The Authenticity Audit, twice, and it meant two different things. The Truth Stack. The Transparency-to-Trust Framework. The Transparency Ladder.

Every one got introduced once, argued for close to a thousand words, then never showed up again on the site.

A framework you actually use gets reused. It's how you think.

A framework that appears once and vanishes is a paragraph pretending to be a system.

Most of the nine end in an FAQ block answering questions nobody actually asked me. All of them lean on a client I "worked with" who never gets named, a dollar figure that never gets receipted, a percentage lift that never gets sourced. AI Haters on LinkedIn is the one exception with zero framework, and it still fails the test from the other direction. It has plenty of voice and a tractor metaphor, but no name, no date, no number anybody could go check. Voice without a receipt is still a post a model could write if you fed it my last twenty captions.

What changed on August 6

Something shifted after that run.

It started with the Algospeak review on August 6. Super Coffee on August 23. The Amanda Hess piece two days later.

Every post since ties back to a thing that already existed before I wrote a word. A book with a Goodreads page. A company with named founders and a named competitor set.

The lesson still shows up. It just shows up attached to a receipt now, not a framework I invented on a Tuesday afternoon.

Digiday made that case for video. Same math hits text.

The market doesn't pay for age. It pays for proof you touched something real. Faking that used to cost nothing. Anybody can see the seam now.

A blog with no ad budget has an organic social media growth problem before it has anything else.

Nobody links to a framework with the serial numbers filed off. They link to the post that names the actual company. It's the only version you can point a client at without getting caught flat when somebody checks.

What I'm doing with mine

Nine of twelve is a bad score when you write about this for a living.

I'm not going back to rewrite the old nine. They stay up. That's the record of what the shift looked like from the inside, framework names and all.

Every post from here forward gets the same question Digiday is putting to advertisers right now. Who paid for this, and can you prove it.

If the honest answer is a framework and an anonymous client, it doesn't ship. That discipline is the only organic social media growth strategy that survives the next platform update hunting for the same tell.

Somebody should run that test against your feed before a reporter or a platform does it for you. That's the real job when you hire a food and beverage marketing agency. Get the one that already got caught failing its own test and fixed it in public.