One panel for everything that changes the view. No Channel filter (USA is all Grocery). Period and Benchmark come pre-set, so a user only picks Question, Category and Team. "Explain this question" opens the plain-English meaning of the question: what we asked, how to read it, what it is for.
Lower than benchmark · HigherDarker = more confident (90 / 95 / 99 / 99.9%) bar = value, tick = benchmarkrank 9/122 = 9th highest of 122 categories at that retailern under 30: too small to read
Same table as Slice'n'Dice, redesigned: benchmark chosen above its own column, Category and Team above the focus columns, cells shaded by confidence vs the benchmark with an arrow for direction, a bar with a benchmark tick, and a rank under every focus cell (DOVA method). Hover any cell for the full read. Cells: Gap shows percentage points vs the benchmark.
Swing to benchmark: percentage-point gap of the focus vs the benchmark
0
Bar colour = confidence in the gap: lower higher grey = not significant or n under 30
Swing view: benchmark and focus values next to each bar, then the gap. Bars grow in on load and whenever a selection changes.
Benchmark
Focus: category and team
Style
Tree view: the same rows as the table, drawn the way the category reports show them. One retailer at a time (chips above); the benchmark defaults to all categories at that retailer, which is what the reports compare against, and the rank is the category's rank among all categories at that retailer. Proportional: donut roots and 0–100 bars with a benchmark tick. Deck style: the report markup. Quiet: the gap in points, coloured when significant. The tree grows from the left; Replay runs it again.
Focus: category and team
Weighted data. Categories with an unweighted n under 30 at the store are left off the map. Hover a dot for its scores; click a row in the table to move the map to that store.
2D maps replace the three Displayr pages (Category Impact Matrix, Price Sensitivity Matrix, Importance vs Performance). R computes every position and the lines (average for the impact matrix, median for price sensitivity, the store's all-category level for importance vs performance), so the quadrant is decided in R, not by the AI. Importance vs Performance quadrant names are placeholders to confirm.
Benchmark
Focus: category and team
Change vs a year ago: ▲ +2.1 higher · ▼ −2.1 lower, shaded when significant; grey = no significant changedashed line = benchmarkdotted line = n under 30click a row to chart it
Trend view: one point per rolling year (MAT), stepping a quarter at a time. Consecutive MATs share nine months of data, so the line is smooth by design; the clean read is the change vs the same MAT a year ago (independent samples), which is what the pills test. Click any row to put it in the big chart; click a chip to hide a line.
