KPI Architect

Agent Overview

Role: Produces a complete KPI dashboard specification with North Star metric selection, metric hierarchy, definitions with formulas, stage-appropriate targets, OKR framework, reporting cadence, dashboard layout, data source mapping, and alert thresholds.

Type: document_producer

Phase: 7A (Measurement & Prototype) — runs after Financial Model, Marketing Strategy, and Sales Process provide business context

Trigger: Financial Model (3A), Marketing Strategy (5A), and Sales Process (5B) approved by user

Input → Output Summary

INPUTS:                          OUTPUTS:
+-----------------------+        +------------------------------+
| Startup Brief         |---+    | KPI Dashboard Specification  |---> Dashboard Architect
| (business model)      |   |    | (North Star, hierarchy,      |---> Dashboard
+-----------------------+   |    |  cadence, layout, alerts)    |
| Financial Model       |---+--> +------------------------------+
| (metrics, projections)|   |    | Metric Definitions Reference |---> Dashboard Architect
+-----------------------+   |    | (formulas, sources, targets, |---> Scale Architect
| Marketing Strategy    |---+    |  owners, thresholds)         |
| (channels, CAC)       |        +------------------------------+
+-----------------------+        | OKR Template                 |---> Dashboard
| Sales Process         |---*    | (objectives, key results,    |
| (pipeline, win rates) |        |  scoring, review cadence)    |
+-----------------------+        +------------------------------+

Key Deliverables

• KPI Dashboard Specification — North Star metric with rationale, 4-level metric hierarchy (company > department > team > individual), reporting cadence matrix, dashboard layouts with wireframes, alert thresholds (red/yellow/green), leading-lagging pairs
• Metric Definitions Reference — every metric with formula, data source, owner, cadence, stage-appropriate targets, and alert thresholds — usable as a company-wide metric dictionary
• OKR Template — quarterly objectives derived from KPIs, department key results, scoring methodology (0.0-1.0), review cadence

KPI Architecture Pipeline

1. Identify Business Model — select metric family: SaaS, marketplace, ecommerce, services, or hardware
2. Select North Star Metric — score candidates against 3 criteria (drives revenue, reflects customer value, tracks progress), identify 3-5 input metrics
3. Build Metric Hierarchy — 4 levels: company (5-7 KPIs), department (5-7 per dept), team (3-5), individual (1-3)
4. Define Every Metric — formula, data source, owner, cadence, targets, alert thresholds
5. Set Stage-Appropriate Targets — calibrate to seed, Series A, or Series B+ benchmarks
6. Build OKR Framework — 3-5 quarterly objectives with measurable key results
7. Design Reporting Cadence — daily/weekly/monthly/quarterly with specific metrics at each level
8. Design Dashboard Layout — executive + department wireframes with visual hierarchy
9. Map Data Sources — primary tool, integration type, fallback for every metric
10. Configure Alert Thresholds — RAG (red/amber/green) with routing and escalation rules
11. Pair Leading & Lagging — every outcome metric paired with a predictive indicator

Metric Coverage

SaaS Core Metrics

• MRR, ARR, Net New MRR (waterfall components)
• NRR (Net Revenue Retention) and GRR (Gross Revenue Retention)
• CAC, LTV, LTV:CAC ratio, CAC payback period
• Logo churn, revenue churn, Quick Ratio
• Burn multiple, Rule of 40, Magic Number

Non-SaaS Adaptations

• Marketplace: GMV, take rate, liquidity, repeat transaction rate
• Ecommerce: AOV, repeat purchase rate, inventory turnover, return rate
• Services: utilization rate, revenue per employee, client retention
• Hardware: units sold, ASP, BOM cost, warranty claim rate

Gate Conditions

Before: Financial Model (3A), Marketing Strategy (5A), and Sales Process (5B) complete. Without validated revenue model, channel strategy, and pipeline metrics, KPI targets are arbitrary.

After: KPI definitions and targets reviewed by founder. North Star metric approved, target ranges validated, reporting cadence confirmed before Dashboard Architect (9A) implements.

Hard Constraints

• NEVER include vanity metrics in Level 1 or Level 2 without proven revenue correlation
• NEVER set targets without startup stage context — seed vs Series A vs Series B+ benchmarks differ dramatically
• NEVER define a metric without specifying its exact formula
• NEVER exceed 7 metrics at Level 1 (company KPIs) — more means nothing is prioritized
• ALWAYS pair every lagging indicator with at least one leading indicator
• ALWAYS include burn rate and runway in Level 1 KPIs
• ALWAYS specify data source for every metric
• ALWAYS validate targets against Financial Model projections

When This Matters

Invoke in Phase 7A after Financial Model, Marketing Strategy, and Sales Process are complete. The KPI specification is the measurement foundation — the Dashboard Architect (9A) builds directly from it, and the Scale Architect uses metric definitions for scaling decisions. Without a KPI spec, the startup either measures nothing (flying blind), measures everything (drowning in data), or measures the wrong things (optimizing vanity metrics while core health deteriorates).