PLO Product Leadership Offsite May 2026

Four asks. One direction.

We've built the framework for an AI-Enhanced FDLC. The question for this offsite is whether we go forward together. Four asks of leadership turn intent into delivery: alignment on the path, protected time to change how we work, budget for the tooling that actually moves the needle, and a FY27 adoption OKR that holds the org to it.

The leadership question isn't how much AI your teams are using. It's what kind of work they're applying it to. Working principle, PLO May 2026
The four primary asks

What we need from leadership.

Each ask reflects a constraint the framework cannot solve on its own. Without all four, the shift either gets diluted in execution or runs in parallel with the work it's trying to replace.

Ask 01 of 04
Process Alignment
Decision · Endorse the direction

Confirm that the AI-Enhanced FDLC is the path we are taking. While we still need to validate the specific activities and tasks with our pilot, there should be alignment across all PM, UX, and Engineering leaders that we need to embrace an AI first way of building software.

The next pilot phase should be for calibration, not a go/no-go gate. The signal it produces shapes how rollout sequences across the org. Without alignment at this level, teams risk optimizing for capability level outcomes vs. platform driven alignment.

  • Endorse the framework as the org's reference model for how features get built.
  • Treat pilot signal as evidence to calibrate rollout, not as a referendum on whether to proceed.
  • Make it explicit that teams have permission to challenge legacy process where it no longer serves the goal.
Ask 02 of 04
Protected Time
Capacity · For change, not just delivery

Changing how we work takes time. Asking delivery teams to ship the roadmap and change their workflow at the same pace will fail. We need explicit, protected capacity for the change itself: companion file authoring, constitution grounding, spec sufficiency work, gate adoption, and the reflection that compounds into the next iteration.

Without protected time, AI gets bolted onto the existing process. The activity moves to AI, the workflow stays the same, and the outcomes don't change.

  • Carve out explicit capacity for pilot trios across the pilot duration, not implicit "do this on top".
  • Protect time for domain leads to author companion files and constitutions before pilot starts.
  • Build change-management time into rollout cohorts so teams can absorb the shift without sacrificing quality.
Ask 03 of 04
Budget
Investment · Tooling, infra, licensing

The right tooling does not come for free. We have evidence that Copilot on its own is not sufficient for the outcomes we want: constitution-grounded generation, deep context loading, multi-step agentic work, and the persistent context that makes spec-driven development reliable.

The budget ask covers an enterprise Claude license, CI/CD pipeline setup for Track B teams, token budget governance, and the platform infrastructure (Textura MCP, agent landscape, observability) that the workflow depends on.

  • Enterprise Claude license to enable Track B (deep context, persistent grounding, multi-step work).
  • Token budget and governance model owned by EngOps, with cost-per-outcome tracked from day one.
  • Infrastructure investment for Textura MCP, the agent landscape, and the in-flight measurement system.
Ask 04 of 04
FY27 OKR
Measurement · Org-wide adoption commitment

Going into the start of FY27 (July), we need a set Agentic SDLC adoption OKR with named owners. This cannot be an aspirational metric tracked in isolation — it has to be a rollup of team-level OKRs across the organization so every team's progress aggregates into the same org scorecard.

Without a committed OKR at this level, adoption signal stays anecdotal and rollout loses the feedback loop it depends on. With it, the AI FDLC Review has the outcome data it needs to keep calibrating where the framework goes next.

  • Set the org-level Agentic SDLC adoption OKR with named owners before the FY27 planning cycle closes.
  • Require each PM, UX, and Engineering team to contribute a team-level OKR that rolls up into the org scorecard.
  • Measure teams onboarded, gates exercised, and outcome signal returning to the AI FDLC Review — not raw activity counts.
Smaller notes

Themes worth naming alongside the asks.

Not asks in their own right, but adjacent topics where shared understanding at this level keeps the rollout calibrated to the right outcomes.

Note 01

Productivity uplift through AI-assisted development

The point isn't more lines of code. It's a shorter path from problem definition to release-ready, with higher implementation fidelity and less rework. Productivity is measured against cycle-time delta and cost-per-outcome, not activity volume.

Note 02

Mitigating friction in AI adoption across teams

Adoption failures usually look like silence: people quietly working around the new workflow rather than naming what's broken. Surface friction early, treat tooling pain as a signal, and separate workflow friction from identity strain. They need different leadership responses.

Note 03

Developing skills and mindset for an Agentic SDLC

The shift is from doing the work to directing, reviewing, and certifying it. That changes what each discipline is great at. Invest in the skills: spec authorship, prompt and context design, review craft, and judgment under uncertainty.

Context for the room

What's already in motion.

These aren't asks. They're commitments and signals the leadership group should be aware of so the conversation lands in the right place.

Already in motion

01
AI ways of working feedback is shaping the framework We are taking recommendations from the AI Ways of Working group, folding their signal into the Agentic SDLC, and intending to mandate adoption once the pilot calibration is in.
02
FY26 pilot to surface gaps and friction For the rest of FY26 (May / June), we're piloting the day-to-day "on the ground" activities of the updated FDLC with teams across the organization. The goal is to identify gaps and friction points and address them before wider rollout.
03
Permission to challenge the status quo Teams should hear from their leaders that re-imagining process is part of the job. Practices that exist because of constraints AI has removed should not survive the shift by default.
04
Tool friction signals are open for inspection Survey output from the AI Ways of Working group is available. Friction patterns, capability gaps, and "tools don't fit the workflow" comments inform the Track A vs Track B guidance.