Sourced writing onAI and marketing.
Short pieces on what is actually measurable, what an agent can and cannot be trusted with, and what the new disclosure rules ask of advertisers. Every claim links to the source it came from.
Few pieces. Each one sourced.
Troy publishes when it has something checkable to say, which is why this list is short. Filter by topic; the count updates as you do.
Your ad platform’s score is not grading your marketing
Meta’s opportunity score and Google’s optimization score are honest numbers about a real thing. That thing is how much of the platform’s own advice you have taken.
What marketing agents can reliably do in 2026 — and what they cannot
The platforms have handed out the keys. Access is no longer the hard part; judgement and control are. Here is the split, with the sources.
AI-made ads now have to say so
Article 50 of the EU AI Act has applied since 2 August 2026, and it is not the only rule pointing at AI-made creative. Here is what the sources say, and where they stop.
How these are written.
The same rule the Troy Score works under: if it cannot be shown, it is not claimed.
- Every claim carries a numbered link to the source it came from.
- Each source lists its publisher, its own date, and the day Troy last opened it.
- At most one short quote per piece. The rest is Troy's own wording.
- The byline is Troy. No invented authors, no guest names, no head-shots.
- Nothing here is a case study. Troy is pre-launch and has no client results to show.
- Where a number could not be sourced, it is left out rather than estimated.