If you're short on time, here's what matters in this issue:

The paradox: 95% of B2B marketers now use AI for content creation. 71% of B2B buyers say engagement is harder than ever. More output, lower signal.

The deeper problem isn't AI: Volume was never meant to be the strategy. AI just made it easier to publish more of the wrong things faster.

Specificity breaks through: Intentional POV, genuine expertise, content that solves something real for a named audience. The CTO blog. The technical deep-dive. Not the "5 Trends" roundup. That's, at times, necessary filler for the content calendar, but don't expect any home runs.

Get the details below.

FamousFolks bridges the gap between management consulting and design agencies, giving you access to expertise in operations, marketing and branding, to move your business forward.

The numbers landed this month from the Content Marketing Institute's B2B Content and Marketing Trends: Insights for 2026, a survey of over 1,000 B2B marketers.

The headline is… uncomfortable: 95% of B2B marketers are now using AI for content creation. In the same research period, 71% of B2B buyers report that engagement is harder than ever.

Near-universal AI adoption for content. Near-universal buyer disengagement.

That's suggests a particular type of flood in the content market, or at the least, very obvious, very poor performance.

A point we keep coming back to here, seemly every week, is that when everyone has access to the same tool producing the same outputs at the same speed, the content itself stops being a differentiator. Why you publish, the specificity, the POV, the expertise that can't be replicated by someone else running the same prompt, becomes more important than what or how often.

The CMI research is blunt about what separates the good from the bad. The pacesetters, their term for the marketers showing measurable results, are treating AI as a production tool while investing harder in the thinking that production can't do.

Forrester is forecasting that by end of year, a Fortune 500 company will sue a B2B vendor for AI-generated misrepresentation, a precedent that will force legal and marketing teams to rethink the economics of "publishing fast."

The buying environment is already skeptical. Regardless of the clear improvements most major LLMs have made in accuracy and sourcing, past and current experiences with AI-generated misinformation is conditioning buyers to treat all content as suspect until proven otherwise.

The same data that shows content engagement collapsing also shows what's holding up. Analyst-backed content. Deep technical expertise. External subject matter voices. Forrester found that 75% of enterprise B2B companies are planning to increase budgets for influencer and analyst relations in 2026, because third-party credibility is one of the few credibility markers buyers haven't tuned out yet.

The CTO who writes the blog that still gets 1,000 visits a month isn't special because he's a CTO. He's special because he knew something specific, said it plainly, and put his name on it.

That's the asset.

Cloudflare's blog is one of the most cited examples of B2B content done right. It's technically rigorous, written by practitioners, and built to serve developers and security professionals who will immediately know if you're bluffing. It draws millions of monthly organic visits not through volume, but through depth and specificity. And it deviates quite a bit from the accepted norm in B2B.

Here's what that difference looks like at the content level:

AI-generic version:

What's wrong with it: It could have been written by anyone about anything. There's no POV, no specificity, no data, no named audience. It gestures at expertise without demonstrating any. A developer reads the first sentence and closes the tab. For a less technically savvy audience, it also presents a number of telltale AI crutches like em dashes and overly general weasel words like 'leading organizations'.

High-trust version (Cloudflare approach):

What works: It opens with a specific claim. It names a failure. It promises a mechanism, not a trend. The audience, a security engineer evaluating vendors, knows immediately that the person writing this has seen the problem firsthand. You can write this with AI (we did), but for real content telling a real, verifiable story, you can't write it with AI only.

Before you publish, ask: could this have been written by someone who has never actually done this work? If the answer is yes, well, it probably was.

Specificity is the tell. The number, the named failure, the niche audience, the honest acknowledgment of what didn't work. These make content trustworthy. AI can approximate them. It can't generate them from nothing.

We use LLMs for most of the content we produce, including this newsletter. But there's a meaningful difference between AI as a tool in the toolbox and AI as the manufacturing plant.

Our approach comes down to three principles:

This basic principle underneath all three principles is one most marketers already know from other production processes. AI amplifies what you bring to it. If you bring volume thinking, you get volume output. If you bring a real POV, genuine expertise, and editorial discipline, the tool becomes a genuinely useful difference maker.

—Jay