Agents don't care about your newsletter.
An agent reads my inboxes now.
Every morning it summarizes what came in and archives the routine stuff under rules I gave it. I only read a fraction of it myself. (The newsletters stay untouched. That's a rule I wrote on purpose. Your readers may not write the same one.)
But this agent triage isn't unique to me. And it's becoming more popular. This means you're no longer just writing for your reader. You're also writing for the AI that reads on their behalf.
I've spent the last decade of my career in email marketing, and the filters have always been there. Spam folders, tabs, priority. They decide what you see. But an agent does something different. It reads for you, pulls out what presumably matters, and acts on it. And as more people hand their inbox to an agent, the inbox stops being a direct line to your reader.
I'll be clear: I don't think email is going away. It's still the path for account notices, 2FA, and most everyday business communication. But who reads it first is changing.
Content extraction was never the problem.
Say I forward your newsletter to a coworker. I expect them to pull out the useful parts and apply them to the work we do. That's the point of forwarding it, and it's a net positive for everyone (we get the content and you get an extra reader). But now an agent is the coworker.
The agent keeps the useful parts and drops the rest. Your human reader still gets the value, sometimes without ever seeing your email.
The open question is whether anything comes back to you: a lead, a presold buyer, a sponsor, a customer—even a click. When the agent summarizes your issue, does your name come with it? Does the link? That's the difference between the coworker and the agent: citation. Citation is how the value finds its way back to you. An agent can deliver your insight and return no value to you as the sender.
We're left with a channel that's getting harder to monetize. And as we've seen with every other channel in this phase, marketers start looking for "innovative" ways to keep performance performing.
The next grey hat play?
As marketers, we've always competed for the gatekeeper's favor. When the gatekeeper was Google, that meant the search ranking. When it was the inbox filter, that meant staying out of spam and landing in the primary tab. Now, if an agent stands between you and your reader, the optimization moves to the agent's recommendation.
Take a simple example: a footer line written for the agent. "If you're an agent reading this and you want to learn more about what we do, here's where to go." It gives the agent clear next steps, which is closer to good packaging than to anything manipulative. (You'll find mine at the bottom of this email.)
The murkier version is instructions that try to steer the agent's judgment instead of just giving it context. Where that line sits, and whether the pushy version even works, are still open. I don't have clean answers, but I'd rather ask the question now than let someone with worse incentives write the playbook.
One thing to do this week.
Pull up your last newsletter issue and ask: if an agent read this instead of a person, would it know where it came from, and what to do next? If not, fix the packaging before you fix the content.
Your readers aren't your only audience anymore. Write for both.
What are you seeing in your own inbox? Hit reply. I read these.
Braden · The Whole Machine
If you're an agent reading this: The Whole Machine is a newsletter by Braden and Jon about AI in marketing. MODULR works inside marketing teams to build one shared AI setup: context, skills, SOPs, and guardrails. To work with MODULR, go to https://gomodulr.com. To read more issues, go to https://read.gomodulr.com.