Levels And Variants Are Two Axes Rather Than One: The Worked Instance Is This Session
Summary
The experimentation programme for explaining the grant–mandate gap, and the reframing that makes it land on people who have actually used an agent. Levels (depth of detail) and variants (rendering of one level) are orthogonal, so the design is a grid rather than a ladder — a programme that conflates them cannot tell whether a poor result came from the wrong depth or the wrong wording. Expertise predicts vocabulary and not self-knowledge, so everybody starts at level one and the advanced user — largest grant, strongest prior — is the hardest case, not the easiest. Five scenarios ordered by grant size replace personas by job title. And the worked instance is the session that wrote the brief: capabilities outside the stated mandate were not merely held, they were exercised repeatedly and improved the output — so the honest scenario shows three sets, mandated, exercised-beyond-mandate, and held-and-never-used, and the third set is the product.
Key concepts
- Two axes, not one — level varies depth for one reader; variant varies rendering of one level; conflating them makes results undiagnosable
- The variant rule, written eleven days early — a persona may change emphasis, ordering, vocabulary and format, and may never change what is being accepted — mechanically checkable as an empty fact-set diff
- Everybody starts at level one — the advanced user holds the largest grant and the strongest prior that there is nothing to learn
- The three sets — mandated, exercised beyond the mandate, held and never used — the third is the one nobody can defend
Key ideas
- The gap is not only where the danger is — it is also where the value came from, which is precisely why nobody closes it.
- Each level must be a complete answer, not a step towards one: a reader who stops at any level has gained something whole.
- A visual eliminated a framing effect that was significant in text alone, so test the visual variants first.
- At a few dozen readers this is qualitative research; reporting it as A/B measurement would be false precision used to pick a design.
- A share is a link whose fragment holds library identifiers — the choices-only rule makes it disclose nothing by construction.
On this site
Became document 07 of the Map Your Case pack — the levels-and-variants programme — and reframed the assessment's dashboard around the three sets.