how the brief is scored.

the standard

we measure product judgment, not product craft. AI made implementation free; what got scarcer is the judgment of what’s worth building. that’s the durable half — and what the mark certifies.

the spine has three tiers. four skills, twelve competencies, four levels.

1. 4 skillsmap · acuity · resolve · know-how — the MARK quadrants. the acronym is the framework.
2. 12 competenciesthree facets per skill. each is a single-word handle (worth, kill, halt, signal, reframe, bet, power, hold, miss, room, user, taste) with an explainer attached wherever it’s shown.
3. 4 levelsdeveloping · competent · proficient · expert— the depth bands on each competency. every competency has a behavioral anchor at each level so “proficient” means the same thing across submissions.

how a submission becomes a score

1. you write a response to the brief. the prompt is the same for every submission — that’s what makes the corpus comparable.
2. the response is read against the rubric. behavioral anchors at each level give the read a fixed reference point; a calibration panel keeps anchor interpretation consistent over time.
3. each of the 12 competencies gets a level read. those roll up into the 4 MARK skills, and into an overall band: developing / competent / proficient / expert.
4. a percentile is reported against the live corpus only once the corpus is large enough to claim one honestly. small-n results show the band but no percentile.

what the rubric is not

it is nota founder’s opinion. talvinder authored the first version of the standard; a calibration panel upholds it now, so your mark means the same whether or not he reads your work.

it is notAI grading the credential. AI assists reading at scale, but the anchors are human-written and every band is anchored to observable behavior, not a model’s confidence. the mark is panel-upheld; that’s the authority.

it is nota measure of PM craft. PRDs, sprints, and coordination still matter — but they’re table stakes now, and they’re not what the mark certifies.

versioning

every result is stamped with the rubric version that scored it. today’s standard is PL Standard v3.2. when the rubric changes, old results stay interpretable under the version they were scored against; new results carry the new stamp.