World Model Readiness
Engraved role-check instrument

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.

Question 1 of 5 · Your voice survives

Does AI-assisted content still sound like your brand, or like a model's default?

Every model has a house style: safe, balanced, faintly corporate. Left unedited it flattens your voice into that default, and the thing that made your brand recognisable is the first casualty.

Question 2 of 5 · You resist the average

Does your AI-assisted work say something new, or recycle the obvious angle?

A model predicts the most likely next sentence, which is by definition the most average one. Lean on it uncritically and you produce the campaign everyone else would have produced, competent and forgettable.

Question 3 of 5 · You measure what it does

Do you know whether AI-assisted content performs, or just that you ship more of it?

More output is not more results. Volume feels like productivity, but if engagement, conversion or reach did not move, you are just filling the channel faster. The metric that matters is the outcome, not the word count.

Question 4 of 5 · You command your tools

Do you use your AI tools with real skill, or type a prompt and hope?

The gap between marketers now is not access to AI, it is fluency with it: knowing which tool for which job, how to brief it, where it fails. Prompt-and-hope produces mediocrity at speed; mastery produces leverage.

Question 5 of 5 · You respect the audience

Does AI let you serve the audience better, or just flood them more cheaply?

Cheap content is a temptation to make more of it. But the audience's attention did not get cheaper, and a channel full of adequate AI filler trains them to ignore you. Respect is the long game; volume is the short one.

For the statistics · one click each

Three questions for the public picture

These do not affect your score. They feed the anonymised, aggregated statistics; groups under 8 respondents are never shown.

What share of your content now starts as AI output?

None
Under a quarter
About half
Most of it
Nearly all

Do you measure whether AI-assisted content performs?

No
Only output volume
Vanity metrics
Real outcomes
Tested against human-led

How do you keep AI content in your brand voice?

I do not
Manual editing
Voice guide in prompts
A defined system
I do not use AI for content

Your context

Used to calibrate the report. Company size and sector remain in the anonymized dataset; your email does not.

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

  1. 1Publish AI as-is
  2. 2Light polish
  3. 3Edit toward voice
  4. 4Voice guide in the prompt
  5. 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

  1. 1Whatever it suggests
  2. 2Mostly the safe angle
  3. 3I push past the first draft
  4. 4Original angle, AI executes
  5. 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

  1. 1Never checked
  2. 2More output feels good
  3. 3Watch vanity metrics
  4. 4Track real outcomes
  5. 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

  1. 1One tool, basic prompts
  2. 2A few tools, hit or miss
  3. 3Competent with several
  4. 4Fluent, right tool per job
  5. 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

  1. 1Volume for its own sake
  2. 2Quantity over quality
  3. 3Trying to hold the line
  4. 4Quality gate on everything
  5. 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.