Description
A template that builds a quantitative finding up from subject and value through time, change, goal assessment, cause, forecast and recommended action.
Armitage treats a metric as a grammatical statement whose elements stack in nine tiers, in the order quantitative systems are typically built. Entity names the subject (dimension and member); Value adds the measure and number, the entity-measure-value triad underlying all business intelligence; Time fixes the point; Change compares with another point; Period aggregates over a span with change from the comparison period; Assessment compares with plan or goal, as a flag or scaled gap; Interpretation adds a peer comparison or root cause; Prediction adds a forecast with explicit confidence; Action proposes a response with a link to carry it out. Higher tiers can be shown without lower ones, and the full chain can be read out as plain sentences for text or voice delivery. It makes explicit what a reader needs to act on a number.
Attributes
- Discipline
- Product Analytics & Strategy
- Loss type
- Latent
- Stage
- Design, Post-launch
- Data
- Analytical / expert
- Effort
- Hours
Steps
- Name the entity: dimension and member, with hierarchy or extra dimensions if needed.
- Attach the measure and its value, and fix the time point.
- Add change against a reference point and express the value over a period with change from the comparison period.
- Assess against plan or goal, as a binary flag or a scaled gap.
- Add interpretation: comparison with peers or a known, evidenced cause.
- Add a prediction with explicit confidence, then propose an action with a link to where it can be taken.
- Render the chain as plain sentences for reports, alerts or voice, omitting tiers the audience does not need.
Sources
- Bringing Numbers to Life (Armitage, 2015)
Also known as
- LAVA Metric System
- Metric grammar
- Tiered metric statement
Categories
- Analytics & Behaviour
- Product Analytics & Strategy