Field note · 8 min read
Marketing in the age of AI slop — earn the right to be heard
When generic copy, generic landing pages, and generic outreach are cheap, volume stops being a moat. The work shifts to earning attention with proof, specificity, and a reason to trust the sender.
Do not compete to produce more marketing. Build evidence and distribution loops that would be difficult for a generic system to imitate.
The signal behind the complaint
An Ask HN post about marketing in an AI-saturated market described a familiar experience: crowded channels, poor-quality outreach, expensive clicks, and more useful conversations in communities where the product could be discussed in context. It is one practitioner’s perspective, but it points to a real operating problem.
Abundance does not remove demand. It raises the cost of attention and lowers the credibility of messages that could have been sent to anyone.
What becomes scarce
| Abundant | Scarce | What to build |
|---|---|---|
| Generic educational content | Original evidence and useful interpretation. | Teardowns, benchmarks, field notes, and customer-informed frameworks. |
| Personalised-looking messages | Real relevance to a current problem. | Signal-led outreach with inspectable evidence. |
| Landing-page variants | Clear positioning and credible proof. | Specific claims, demos, examples, and objection handling. |
| Automated engagement | Trusted participation in a community. | Helpful answers, shared methods, and consistent contribution. |
Move from content volume to evidence volume
Your content should make a claim the reader can inspect. That might be a system diagram, a tested workflow, a de-identified implementation pattern, a before-and-after decision, or an honest failure analysis.
- Show the work: publish the inputs, caveats, and decision rules behind a recommendation.
- Be usefully narrow: answer a specific role’s question at a specific moment, not “everything about AI.”
- Keep receipts: link to sources, label anecdotes, and distinguish a hypothesis from a benchmark.
- Create reusable assets: turn insight into a checklist, template, rubric, or tool the reader can use immediately.
Make distribution a feedback loop
Distribution works best when it creates learning, not just reach. Treat each channel as a way to hear how a real audience describes a problem.
- Choose one audience and one question.
For example: how does a PLG RevOps team decide when a product-qualified account deserves sales attention?
- Publish a useful point of view.
Share a short framework or observation that gives people something concrete to agree with, challenge, or adapt.
- Participate where the question already exists.
Join relevant communities to contribute context — not to paste a link into every conversation.
- Turn feedback into the next asset.
Questions, objections, and implementation gaps should become the next field note, template, or product experiment.
Use automation without becoming indistinguishable from it
Automation is still useful for research, summarisation, drafting, routing, and measurement. The line is simple: use it to prepare better judgment, not to simulate interest you have not earned.
For GTM engineers, that means building controls around message generation, source freshness, approval, volume, and feedback. The work is not less technical in an AI-saturated market — it is more accountable.