User Guide

Can this person really do marketing work well — and can you show why you think so?

Market Skill turns “good at marketing” into five competencies graded on the same 0–5 scale as every other skill in WorkDNA. People practise in the training gym, get assessed on real work-sample tasks, and a competency is only certified when there is both a level and a scored piece of work behind it.

Evidence, not a personality score“Not assessed” is a normal answerNo outcome promises

1. Marketing stops being one word

Until now the entire discipline was a single line in the skill library, sitting next to sales. That is fine for a headcount report and useless for the conversation you actually have — whether this person is ready to own a launch, and which part of the job they should be handed next.

Market Skill replaces that one line with five competencies, each graded 0 to 5 on exactly the same scale as every other skill in WorkDNA. A level 3 here means what a level 3 means everywhere else in the product, and it feeds the same skill matrix, the same career-track fit, the same growth view.

The five

  • Knowing when a piece of work applies — reading the situation and choosing the right kind of job before starting, including the times when the right answer is “we do not know enough to start yet”.
  • Knowing which inputs already have an owner in the business, and going to fetch the real number instead of typing a guess. This is the flagship, and it gets its own section below.
  • Building a brief that is actually complete — role, context, goal, constraints, and a stated shape the result has to come back in.
  • Knowing the Thai marketing rules well enough to spot what has to be handled before anything is published.
  • Knowing what can and cannot honestly be measured at your volume — including being right when the honest answer is “there is not enough data here to call this a win”.

2. The one that separates a junior from a senior

Give two people the same brief with seven blanks in it. The junior fills in all seven from memory. The senior notices that five of them already have an owner somewhere in the business — the analytics segment, the brand guideline, the media plan, whoever signed off the budget — goes and gets the real values, and only decides the remaining two, because those genuinely are judgement calls with no source to check.

That is not a personality trait. It is a habit, it is teachable, and WorkDNA teaches it in a place people already look. In Prompt Studio, every field a template needs is a labelled chip on one of two sides. A solid blue chip means the value has an owner and you go and fetch it — and the template will not save unless it names who owns it. A dashed violet chip means nobody owes you this one; it is your call, and claiming an owner for it would be inventing a fact.

Nothing is resolved for you. The chips explain, they do not look up: WorkDNA does not reach into your business systems and fill the blanks in. Going and getting the number is the skill being trained, so the product deliberately does not do it for you.

3. Practise in the gym, get assessed on real work

There are eight practice missions, and they appear in the training-gym queue of anyone in the creative and marketing role family. They are graded against a fixed answer key — no AI decides whether you were right — and each one ends with a debrief in three parts: what the AI gave you, what to watch out for, and what the skill point actually was.

The missions are the practice. The evidence is three work-sample tasks with structured rubrics. Write a creative brief from a synthetic data pack that quietly contradicts itself — the sales report and the brand book disagree about the price, and the brand book is ten months older. Review six pieces of copy queued for tomorrow and say, for each, whether it carries an obligation and what has to happen before it goes out. Read a four-channel budget test, recommend where the next money goes, and list the conclusions that cannot be read at that number of orders.

A submission is scored by a person against the rubric, and every one of these templates is marked blind-scorable and competency-only: no criterion touches a personal attribute. Once a template is published it is frozen. A fix means a new version, never an edit to the evidence somebody was already graded on.

What a certification says, and what it never saysA competency is certified only when two things are true at once: the assessed level is 4 or higher, and at least one scored work-sample run exists against a published template that declares that competency. A level with nothing behind it is a self-report; a scored task with no level is a loose artifact. Neither certifies anything on its own.The badge states what was assessed, when, and on what evidence. It never predicts a result. WorkDNA has no external users yet, so nobody can honestly tell you that a certified marketer ships better campaigns — and a badge that implied it would be a lie dressed as a credential.It is also not a personality score and not a ranking of people. It is not built to sit behind a hiring or firing decision, and it should never be used as one.

4. “Not assessed” is a normal answer, and it is not a bad one

If nobody has assessed someone on a competency, WorkDNA says exactly that. It does not quietly show a zero. There is a real difference between “rated, and rated low” and “nobody has looked at this yet”, and the second is by far the more common state in any real company — so it gets its own words instead of being rounded down into a bad score.

Absence of evidence is not weakness. Expect to see “not assessed” a lot, especially in the first months, and read it as a to-do for the manager rather than a verdict on the person.

5. Compliance: know the obligation, then go and check the current wording

Thai marketing carries real obligations, and this is the easiest place in the whole feature to ship something false. So the training is deliberately narrow about what it grades: that an obligation exists, which piece of work it attaches to, and what has to happen before that work is published.

Work drafted or generated with AI carries a disclosure obligation. Regulated product categories — food, supplements, cosmetics, medical devices — need permission before publication and have claim vocabulary you may not use. Customer data used for marketing has consent limits. A mission grades you on spotting that a piece is carrying one of these and that nobody has handled it yet.

What it deliberately does not grade is the exact required wording. Rules move, and an assessment that tests recall of a string we have not verified against the official source teaches something that may already be wrong. So the training teaches the habit instead: every compliance mission carries the official regulator's link, and going there to read the current wording before publishing is the behaviour being trained. The work-sample task says so out loud — write “check the current wording at the official source”, name which regulator, and you have answered correctly.

And WorkDNA does not inspect your real marketing copy before it goes out. It teaches and tests this competency; it is not a gate on your live work, and it is not legal advice.

Certification is off until an admin turns it onThe certification surface is behind a feature flag called market_skill_enabled, and it ships off. Until an owner or admin turns it on — at /flags, or the Feature flags tab in /admin — the certification endpoints answer as though the feature does not exist. That is deliberate fail-closed behaviour, not a bug, and it is why you may go looking for a button and not find one.Everything else is not gated. The six Marketing skills, the Marketing Specialist career track, the eight missions and the three work-sample templates are catalogue data — they are there for every tenant already, whatever the flag says.

What is not covered yet — said out loud, not hidden

  • Six marketing skills exist in the library and four of them have an assessment instrument today. Brand & Messaging Strategy and Content & Creative Direction do not have one yet, so they stay honestly not-yet-assessable rather than being graded by an instrument that was not built for them.
  • Measurement judgement is graded through the analytical skills WorkDNA already had — experimentation and statistics — on purpose, so the same competency is not measured twice under two different names.
  • Every brand, number and document in the missions and work samples is invented. They are assessment material, not case studies, and nothing in them describes a real company.
  • There is no evidence that any of this improves anybody's marketing. That would take real users over real months, and WorkDNA has none yet. What exists today is the content and the assessment instruments — claiming more than that is exactly the thing this feature was built not to do.