Chapter 41 — Data governance: system optimization
Lesson map
Lesson 41 published. Map, example, EX121.
flowchart TD
N0["Learning outcomes"]
N1["41.1 Whole-chain view"]
N2["41.2 Balancing objectives"]
N3["41.3 Bottlenecks"]
N4["41.4 Measuring improvement"]
N0 --> N1 --> N2 --> N3 --> N4
N4 --> EX["Worked example"]
EX --> Q["Exercise EX121"]
Logical drawing
flowchart LR
N0["Learning outcomes"]
N1["41.1 Whole-chain view"]
N2["41.2 Balancing objectives"]
N3["41.3 Bottlenecks"]
N4["41.4 Measuring improvement"]
N0 --> N1 --> N2 --> N3 --> N4 --> Q["EX121"]
[Learning outcomes]
|
v
[41.1 Whole-chain view]
|
v
[41.2 Balancing objectives]
|
v
[41.3 Bottlenecks]
|
v
[41.4 Measuring improvement]
|
v
[Worked example] --> [Exercise EX121]
Learning outcomes
You will (1) optimise the governed chain as a whole; (2) balance quality, cost and timeliness; (3) avoid optimising one part at the expense of others.
41.1 Whole-chain view
Kofi looks for improvements across the whole leakage chain rather than within single teams. Billing had tuned its extract to run faster by skipping a validation step, which pushed errors downstream to reconciliation, where fixing them took far longer. He shows the trade-off with timings and break counts. The whole chain performs better when billing restores the validation, even though the billing step itself becomes slower. Local optimisation had made the overall system worse while every team believed it was helping. Billing agrees.
41.2 Balancing objectives
Quality, cost and timeliness pull in different directions. Earlier deadlines leave less time for checks; more checks cost more computing and staff time. Kofi builds a simple table showing, for several schedule options, expected on-time rates, check coverage and cost. Owners choose an option that keeps full checks for high-risk data while moving low-risk checks later in the day. He explains that making trade-offs explicit lets owners choose consciously rather than discovering them when something breaks unexpectedly. The chosen option is recorded with reasons.
41.3 Bottlenecks
Kofi identifies the bottleneck in the chain: manual approval of reading corrections, which waits for one supervisor. Corrections pile up before month end, delaying reconciliation and the leakage figure. He proposes delegating approvals for small corrections to trained stewards, with supervisors approving large ones and reviewing a sample of small ones. The queue clears faster, and audit sampling shows no increase in incorrect corrections. Improving the bottleneck helps the whole chain far more than speeding up steps that were already fast. Supervisors agree.
41.4 Measuring improvement
Kofi measures the chain before and after changes: time from meter reading to leakage figure, share of figures published on schedule and number of restatements. All three improve over the following quarter. Prof. Varga notes that system-level measures prevent teams from claiming success on local measures while the overall outcome stagnates, and that the leakage figure's timeliness and stability are what the regulator and network engineers actually experience. Kofi adds the three chain measures to the governance board's monthly dashboard, so that any slipping trend is visible to everyone early.
Worked example
Kofi reverses a local billing shortcut that harmed the chain, presents explicit trade-offs, relieves a correction approval bottleneck with delegated steward approvals and measures chain-level improvement.
Synthetic case. Timings are synthetic.
Errors and limits
Optimising one team's measure can hurt the system; measure the chain.
Exercises
- EX121 — Billing shortcut. What did billing's shortcut cause? Rubric: 3 = errors pushed to slower reconciliation;; 2 = partly correct;; 1 = vague;; 0 = none.
- EX122 — Bottleneck. What was the bottleneck? Rubric: 3 = manual approval of corrections by one supervisor;; 2 = partly correct;; 1 = vague;; 0 = none.
- EX123 — Chain measures. Name the three chain measures. Rubric: 3 = reading-to-figure time, on-schedule share and restatements;; 2 = all;; 1 = some;; 0 = none.
Continue studying — lesson 41 (chapter 41 of 50)
← Chapter 40 · Course chapter index · Chapter 42 →
Continue to the next authored lesson.