Your follow button is broken.
Not broken like a bug. Broken like a payphone. Still standing there, technically functional, but nobody's routing calls through it anymore.
You've seen the evidence yourself. You've got 1,200 friends on Facebook, you post something, and you get 3 comments and 14 likes. Friends tell you to your face, "I never see your posts anymore."
My own wife has to WORK at seeing my posts. Chaunna likes and comments on purpose, because if she doesn't, my posts drop out of her feed within a few days. She's married to me and the algorithm still won't show her my content.
The follow doesn't matter. The subscribe button doesn't matter.
Social media stopped being social a few years ago, and most people never noticed. We still call it social media the same way we still say "roll down the window." The name outlived the machine.
What we actually have now is a discovery layer.
Write this definition down: a discovery layer is a recommendation engine that decides what content gets seen based on engagement data, not relationships. Who you follow is a suggestion. What you watch is a command.
Every major platform made the same switch. TikTok, YouTube, Instagram, Facebook, LinkedIn, X. The tell is the FYP, the "For You Page." Not the "From Your Friends Page." For YOU, as calculated by a machine that's been watching you scroll.
How The Algorithms Actually Decide
Here's where it gets useful for us marketers.
These algorithms run on engagement signals. Likes, comments, shares, saves. But the king metric, the one that rules them all, is dwell time. How long a piece of content holds a human's eyeballs before they scroll on.
YouTube even has a name for it: AVD, average view duration.
And once you know that, everything weird about the modern internet suddenly makes sense. It's why every creator now has a 1 to 3 hour podcast-style channel. It's why captions say "wait for it." Longer watch time means more ad slots, and more ad slots means the algorithm pushes you to more people.
Now the good news...
These algorithms are elementary. They're not reading minds. They're counting seconds and signals. And anything that elementary can be put to work FOR you instead of against you.
So we built an agent that does exactly that.
I call it the Content Spy Agent.
It watches the top 5-10 profiles in your space, scrapes their engagement numbers, marks which content performed best, and then emails you a designed HTML report: a summary of what's working, why it's working, and strategic recommendations for your weekly content plan.
You can run it standalone just to stay on top of your competitors. Or you can chain it to other agents we've built (like the daily LinkedIn writer) so the spy finds the winning topics and the writer drafts your take before you've had coffee.
The Skill Hiding Inside This Build
And that brings me to what makes this issue's build special. The Content Spy Agent isn't one workflow. It's THREE.
There's a main workflow acting as the orchestrator (think: the manager), and two spy workflows acting as its subagents. One watches YouTube, one watches X. The orchestrator sends them out, collects their reports, and compiles your strategy.
Why does that matter? Because orchestration is the difference between owning an AI toy and running an AI department. Once you can make one workflow command others, you can run multiple blogs, multiple newsletters, multiple social profiles... even one big orchestrator running your entire marketing operation.
And you're going to learn it today on a build small enough to finish before lunch.
Why I Bother With Any Of This
I'll be straight with you. I hate this new version of social media.
But the leads are too good and the reach is too good to sit out. I've had single posts seen by 170,000 people. I've had CEOs of large companies retweet me. You don't walk away from that kind of distribution because you miss 2012.
Because it's not social media anymore, it's TV channel surfing. People scroll to see what grabs their attention, and THEN, once you have their attention, you move them onto your email list.
That's the whole strategy in one sentence: win attention on the discovery layer, build the relationship on email. Every $1 spent on email marketing turns into $36 in revenue on average. Email is the highest converting medium on the planet, and it's the only true subscribe left. (More on that in a future issue.)
So the spy works the algorithm, and my newsletter does the relationship building.
My version watches 6 YouTube channels and 5 X profiles at the same time, runs itself on a schedule, and lands in my inbox every Friday looking like a report I used to pay an agency for. The full step-by-step build is available to Build Notes+ members.
Continue with Build Notes+
This is the free preview. The complete step-by-step guide and supporting files are available to Build Notes+ members.
