Hybrid Organization
Neither pure functional (group by expertise) nor pure mission-oriented (group by objective) structures work at scale. Grove argues for combining both:
- Mission-oriented units handle customer-facing work — speed and autonomy to respond to markets
- Functional units provide centralized services where scale matters — HR, legal, specialized expertise
The question isn’t “which structure?” but “which structure for which work?”
Dual Reporting
The hybrid creates matrix management — individuals reporting to both a functional manager (expertise, standards, career) and a mission manager (priorities, delivery, outcomes).
This friction is a feature. It forces trade-off conversations to happen explicitly rather than letting one priority silently dominate. The matrix doesn’t eliminate conflict — it institutionalizes it. The alternative isn’t harmony, it’s hidden dysfunction.
Task-Relevant Maturity
Adjust management style based on someone’s experience with a specific task — not their general seniority.
- Low TRM: Structured, directive. Tell them what, when, how. Monitor closely.
- Medium TRM: Coaching. Two-way communication, set goals together, regular but less intensive monitoring.
- High TRM: Delegative. Set objectives and constraints, then step back. Monitor at milestones.
A senior engineer might need high TRM style for backend architecture but low TRM style for their first people management responsibility. Mismatching style to TRM fails in both directions.
Pairing Indicators
Any single metric creates incentive to game it. Pair complementary indicators that create natural tension:
| Quantity | Quality |
|---|---|
| Units produced | Defect rate |
| Tickets resolved | Reopened rate |
| Features shipped | Bug count |
Neither metric can be gamed in isolation because optimizing one at the expense of the other becomes immediately visible.
Checkpoint vs Outcome KRs
Grove’s original Key Results were checkpoints — concrete milestones. Modern OKR literature reframes them as outcome measurements — metrics indicating progress.
Different failure modes:
- Checkpoints force sequencing discipline but can become output-focused without asking if it mattered
- Outcomes keep focus on impact but are harder to pace against
Execution mode (known path) favors checkpoints. Innovation mode (still searching) favors outcomes. Mismatch between style and context is why most OKR implementations fail.
Tension with Theory of Constraints
Grove measures productivity at the individual work unit level — improving any unit’s productivity improves the whole.
Goldratt would disagree. In any system with dependencies, only the constraint determines throughput. Optimizing non-constraints creates waste: inventory piles up, WIP increases, resources consumed producing things that can’t flow through.
Grove’s indicator pairing partially guards against this, but he never asks “is this work unit even the bottleneck?” before optimizing it.