What we changed
The organization shipped constantly but couldn’t connect work to outcomes. The solution was not just "more process", it was installing a decision system: measurable OKRs, clear decision rights, and monthly outcome reviews that made learning fast and failure cheap.
Velocity without direction
Shipping volume was high, but leadership couldn’t answer: what changed because we shipped? Planning became a negotiation, teams reacted vs planned, and engineering paid for it in rework and burnout.
The system rewarded activity over impact. Roadmaps tracked output and no one owned outcomes.
What was actually broken
I spent the first 90 days observing planning, interviewing across functions, and reviewing how objectives and roadmaps were used. Four constraints showed up everywhere:
- Success was vague: objectives were aspirtaional goals, not measurable outcomes
- Discovery was late: validation happened after decisions were made
- Accountability was unclear: outcomes had no owner
- Roadmap was a wishlist: constant reprioritization created waste
The decisions that mattered
The operating model change worked because it was anchored in decision rules and accountability—not “ceremonies.”
The model we installed
The point wasn’t “agile.” It was a coherent loop: strategy → decisions → discovery → delivery → outcomes. Lightweight enough to move fast, strict enough to protect focus.
- Goal: what we were trying to change
- Learning: what we discovered (evidence)
- Change: what moved in the metrics
- Decision: continue / pivot / kill
How we made it real
This wasn’t a “process rollout.” We made the model real through enforced decision rules and by publicly stopping work that didn’t move outcomes.
- Kill criteria: stop low-impact work early and visibly
- Instrumentation: measure outcomes, not user stories
- Cadence: reviews ended with decisions, not status
- Air cover: executive alignment protected the shift
What changed
Outcomes improved because teams stopped building the wrong things and stopped reflecting on output.
Capacity returned to the work that mattered
We reclaimed 12% capacity by stopping 15 low-impact initiatives early—freeing engineers to focus on the work tied to key results.
Delivery got faster—and more predictable
Clear decision rights and fewer restarts produced 20% faster time-to-market without increasing burnout.
Customer experience improved
The model created accountability to customer outcomes, contributing to +8 NPS points.
Leadership trust increased
Once leaders saw teams killing their own bad ideas and learning fast, micromanagement decreased and autonomy increased.
The hard parts no one talks about
Operating model change is culture change. Clarity is gained through repetition, coaching, and executive alignment.
- Not everyone adapts: outcome accountability changes who thrives
- Executive air cover is essential: without it, politics (and old habits) wins
- Measurement is hardest: teams need coaching to choose meaningful KRs
Operating model templates
These artifacts made the model repeatable and scalable. Link to redacted versions if you want to share publicly.