Learning objective
Build a compact strategy map and scorecard that distinguishes objectives, measures, targets and initiatives, then use conflicting signals to review causal assumptions.
Why it matters
A dashboard can report activity perfectly while the strategy fails. The scorecard is useful only when measures test an explicit account of how capability and process change should create customer and financial or mission results.
Core concept: objective → hypothesis → measure → decision
An objective is a desired strategic outcome. A measure is an operationally defined signal. A target is a desired level by a date. An initiative is work intended to contribute. Confusing these turns deployment into success by definition.
Scenario
An insurer reports that AI-answer adoption reached 82%. Its customer objective is “trusted resolution,” but complaint recurrence and exception backlog are rising.
Worked decision
Do not promote adoption into the customer perspective. Treat it as an initiative signal. Define trusted resolution through observable outcomes, connect it to correct exception handling and capability, then investigate why the leading and lagging signals disagree.
Decision rule
Keep a measure only when the team can answer: which objective does it inform, how is it defined, who can act, what would contradictory evidence mean and which guardrail prevents gaming?
AI lens
Use AI to reconcile definitions and surface inconsistent signals with source links. Do not allow it to invent causality, change denominators or decide acceptable trade-offs.
Quick check
Which is an initiative rather than a strategic outcome?
A. Reduce validated repeat complaints.
B. Improve first-contact resolution for complex cases.
C. Deploy an AI assistant to all service teams.
D. Sustain segment-level retention.
Answer: C. Deployment is work whose contribution must be tested.
Next step
Complete the Balanced Scorecard workspace and then practise on the AI service scorecard case.