Every number on this page is computed by the Chapter 11 notebook's own code. The 200,000 weightings are drawn with seed 26111, as in the book; the breach odds are Worked Example 8.1's, drawn with the book's seed 8. Your notes stay on this page until you leave it; copy the memo before you go.
Meridian's board wants a performance dashboard it can trust through five years of transformation, whichever option it chooses. The chief financial officer wants one number; the transformation office wants the full profile; the divisions want their own margins reported. You are on the performance-management team, and the dashboard you design will be used to judge the program it was built to support.
Use AXIOM to build Meridian's performance dashboard, test what its single number decides, and check that its ratio measures and review windows tell the truth.
Worked Example 11.1: the CFO's scorecard, ¼(Liquidity + Technology + Efficiency − Strategic risk), on the Chapter 1 model's paths for the three options (Online Appendix 1.B.4).
In the dashboard builder, declare six or more dimensions with their supports, one indicator each with its error, and a band for each dimension.
Table 11.4's six dimensions with their supports on Chapter 9's architecture, and the indicators Meridian declares (Table 5.2): today's band is the reading plus or minus its error.
Improving the row's dimension moves the column's with the declared sign, no sooner than the lag floor in quarters: the distance between their supports along Chapter 9's fourteen arrows (Theorem 11.3(i)).
On the weight-sensitivity bench, score the three options, sweep the weights, and find the tie nearest the CFO's equal weights.
Left: a slice of the weights, technology across and liquidity up, with efficiency and strategic risk sharing the rest equally; the dot marks equal weights. Right: the share of 200,000 weightings, drawn uniformly with seed 26111 as in the book, that rank each option first: a measure of the space of weights each option wins, not a probability.
In the normalization panel, switch among index points, best-to-worst scaling, and percentage improvement, and between two reference groups; record every change of rank.
Equal weights throughout. A normalization p ↦ (p − m)/r turns the declared weights wi into effective weights wi/ri (Theorem 11.2(iii); Online Appendix 11.B.3). The two other reference groups are computed from the same Year-5 readings; the book prints the three-option scale.
On the mix bench, shift revenue between divisions and watch every divisional margin improve while the group margin falls; then rebuild the report as a pooled ratio with a rate-and-mix bridge.
Theorem 11.5(i)'s two divisions raise their margins from 20.0 to 22.0 percent and from 15.0 to 16.0 percent. Line: the group margin at this year's divisional margins as the revenue mix moves; dashed: last year's group margin. The book declares no divisional margins for Meridian.
In the frontier explorer, find the Pareto optima and any balanced option that no weighting selects; in the J-curve simulator, slide the evaluation window until the digital program turns from loss to gain.
Left: three indivisible programs in points of margin and months of resilience gained (Theorem 11.4(iii); Online Appendix 11.B.4). Right: Figure 11.2's frontier of EBITDA margin against liquidity cover, the operating point inside it and the point it reaches (Proposition 11.2; Theorem 11.4(iv)).
The board's risk appetite as a gate (Table 11.3): the digital program must breach the $350 million floor in at most 5 percent of paths before it is scored (Worked Example 8.1, seed 8).
Between options that do not dominate each other, the ranking moves with the weights and with the normalization (Theorem 11.2); divisional ratios can all improve while the pooled ratio worsens (Theorem 11.5(i)); and a program's verdict flips at the payback window (Theorem 11.3(iii)).
Dimensions with supports and indicators; weights with their margin from a tie; the yardstick and its reference group; the aggregation rule and any gates; ratio measures reported as pooled ratios with a mix bridge; and the evaluation window for each program.