
For you
Role Checks
Module · Multiplier for your craft, or a shortcut to average
The Marketer's AI Check
AI can write a week of content in an afternoon, all of it competent and none of it yours. That is the quiet cost: fluent, on-trend, and indistinguishable from what every competitor using the same tools is publishing. This module checks the five habits that keep AI a multiplier rather than a leveller: your voice surviving the edit, originality over recycled patterns, measuring what the content actually does, real command of your tools, and respect for the audience on the other end.
What the five levels look like
Every dimension in this assessment is scored 1 to 5. This is what the levels mean, dimension by dimension. The graded report diagnoses where your own answers land and what to do about it.
Your voice survives
- 1Publish AI as-is
- 2Light polish
- 3Edit toward voice
- 4Voice guide in the prompt
- 5Voice guarded, every piece
At the low end: Publishing raw model output means your brand now sounds like everyone else's. Read a piece aloud: if it could carry any competitor's logo, it is not yet yours. What good looks like: Guarding your voice on every piece is what keeps AI a tool rather than a homogeniser. Keep the voice guide current; a brand voice is a living thing, not a one-time brief.
You resist the average
- 1Whatever it suggests
- 2Mostly the safe angle
- 3I push past the first draft
- 4Original angle, AI executes
- 5Distinct point of view throughout
At the low end: Taking the model's first angle gets you the industry average with your name on it. Use AI to draft, then reject the obvious take and find the one it did not suggest. What good looks like: A distinct point of view carried through the work is what AI cannot generate for you. Protect the time to find it; it is the whole reason the audience picks you.
You measure what it does
- 1Never checked
- 2More output feels good
- 3Watch vanity metrics
- 4Track real outcomes
- 5A vs B on performance
At the low end: Shipping more content without measuring it is motion, not progress. Pick one real metric per piece and see whether the AI-assisted work actually moves it. What good looks like: Testing AI-assisted against human-led on real performance is how you learn where the tool helps and where it dilutes. Keep the tests running; the answer changes as the channel does.
You command your tools
- 1One tool, basic prompts
- 2A few tools, hit or miss
- 3Competent with several
- 4Fluent, right tool per job
- 5Building repeatable systems
At the low end: One tool and basic prompts leaves most of the leverage on the table. Invest a few hours a week in learning what the tools can actually do; the compounding is real. What good looks like: Building repeatable systems out of your tools turns personal skill into scalable leverage. Keep learning; the tool landscape shifts monthly and today's edge is next quarter's baseline.
You respect the audience
- 1Volume for its own sake
- 2Quantity over quality
- 3Trying to hold the line
- 4Quality gate on everything
- 5Less, but genuinely better
At the low end: Flooding the channel because AI made it cheap is how you teach an audience to tune you out. Ask of each piece whether it earns the attention it asks for. What good looks like: Choosing less but genuinely better respects the audience's attention and compounds their trust. That trust is the asset AI cannot manufacture and competitors cannot copy.