KPI Design Generator

Problem It Solves

Many companies struggle not because they lack metrics, but because goals cannot be decomposed cleanly. Company-level targets often become slogans at department level, while role-level KPIs miss weights, baselines, data sources, and ownership. Metrics can also conflict with each other: sales may optimize price, while marketing optimizes volume. This skill turns the “metric design” phase into a structured drafting workflow: from company goals, org structure, and industry context, it produces a reviewable KPI proposal instead of a generic performance handbook.

How It Works

The input usually includes company annual goals (ideally 3–5), org structure with key roles, and industry type, such as SaaS, e-commerce, finance, manufacturing, or professional services. The workflow covers:
- Company goal overview: checks goal count, maps owning and supporting departments, and flags initial conflicts.
- Department KPIs: breaks goals down by department, with metric names, weights, targets, measurement methods, and traceability.
- Role KPIs: creates 2–4 metric cards for key roles such as sales, product, and engineering, with SMART validation.
- Conflict diagnosis: marks cross-department or intra-metric conflicts as fatal, severe, or minor, and suggests fixes.
- Metric dictionary: outputs definitions, formulas, data sources, and update frequency for downstream import or review.

It includes 300+ industry metric templates and supports a KPI/OKR hybrid mode: company-level OKRs can be converted into quantifiable department-level KPIs.

Boundaries

It fits early-stage metric design, HRBP and business leader alignment, and cross-department KPI conflict scanning. It does not replace human alignment meetings, does not collect live metric values, and does not integrate directly with systems such as Beisen or Lark Performance. The output remains text and tables that need manual review before import. It also does not score employees, allocate bonuses, or decide salary adjustments; those decisions should come after metrics are live and supported by actual performance data.