Analytics Catalog/Workday/Talent/Nine-box
Explore the catalogModulesTalentThe talent starRating distributionDelivered vs owned
Workday · Talent · Report

Nine-box placement, grids that sum

The nine-box is a join: performance against potential, per worker, per cycle. Delivered reporting keeps the two ratings in separate sources, so the grid gets built by hand each cycle and history gets lost when the next one starts.

RuleBuild the grid from the rating fact: two rating types joined on worker and cycle, cells summing exactly to the rated population.
Neverkeep only the current grid. Placement movement between cycles is the succession signal, and it needs last cycle’s grid to exist.
Potential High Medium Low High The bench Keep engaged Reward in place Medium Develop next The solid core Steady hands Low Wrong seat? Support closely The hard talk Performance
What the grid decides: where each worker stands, and what to do about it. The counts are in the table below; the conversations are the point.
The grid, mid-year 2026296 rated workers, nine cells, sums shown.
Performance \ PotentialHighMediumLowRow total
High1427849
Medium3111822171
Low9412676
Column total5418656296

The 14 in the top-left cell are a list with names on the report itself, because a succession conversation starts from people, not counts. Sample values are illustrative, never client data.

Why delivered grids get rebuilt by handtwo ratings, two sources, one wall.

Performance and potential are separate rating types in separate delivered data sources, and one report reads one source: the single-source wall documented on delivered vs owned. So the grid becomes an export-and-spreadsheet exercise each cycle, and each cycle’s spreadsheet replaces the last. In the rating fact the join is one line, and per-cycle rows mean every grid that ever existed still does.

Movement between cyclesthe grid’s real payload.

With two cycles stored, movement is a self-join: who rose into the top band, who slid, who has sat in the same cell for three cycles. Those trajectories, not the snapshot, are what succession planning and flight-risk conversations actually need, and they only exist because the model keeps every cycle. Band boundaries are configured on the scale dimension, versioned like everything else.

Use case
Problem
The nine-box is rebuilt in spreadsheets each cycle because the two ratings live in separate sources, and history dies with each rebuild.
What we build
The grid as a join on the rating fact, per cycle, with banded scales versioned and every historical grid retained.
What you get
Grids that sum to the rated population, named lists in every cell, and cycle-over-cycle movement for succession planning.
Who moved out of your top-right cell since last cycle?
We ship the grid with history, names, and movement, in your cloud.
Talk to us
Terms on this page
battery
The automated set of checks, including the crossfoot arithmetic, run on every build of this catalog.
nine-box
Performance against potential, three bands each, nine cells.
band
A range of the scale mapped to high, medium, or low.
placement
A worker’s cell in one cycle’s grid.
movement
Cell changes between cycles. The succession signal.
succession planning
Deciding who is ready for what, next.
self-join
Joining a fact to itself across cycles to see change.
rated population
Workers with both ratings in the cycle. The grid’s total.