Breaking Sacred Rules: When PM Best Practices Become the Problem
1. The Setup
I inherited an account in swirl confusion mode. The prescription? Add more project managers to handle increasing workload. The logic seemed sound: more projects = more PMs doing more validation = better control. Except the account was operating at a change frequency that made traditional PM gatekeeping completely counterproductive. By the time we validated, staffed, and got legal sign-off, the client had already pivoted twice and the team was three weeks behind. The question: What if the problem isn’t insufficient process, but the wrong operating model entirely?
2. My Approach + AI Role
I spent the first month confused, assigned to a project without the full picture. Once I realized I was supposed to be the lead, I started pattern hunting with AI:
- Mapped how work was actually organized vs. how our tools assumed it was organized
- Reviewed meeting transcripts for nuances I’d missed in real-time
- Visualized the mismatch between our process-heavy approach and what this relationship-based account needed The patterns pointed to something I’d learned in APAC: when change frequency exceeds a certain threshold, administrative perfection becomes administrative theater. My hypothesis: We didn’t need more PMs doing project-based deployment. We needed domain leads playing zone coverage with strategic oversight, not one-on-one matchup validation for every single task.
3. What Actually Happened
I broke multiple sacred PM rules:
Rule broken #1: Validate before resource allocation. What I did instead: Strategic pre-positioning based on pattern recognition, knowing we’d be 2-3 weeks behind if we waited for signed paperwork.
Rule broken #2: Keep forecasts updated as source of truth. What I did instead: Built a mode of operating where team members speak up when something’s going off the tracks. Portfolio-level hygiene checks and continuous signal-sharing replaced perfect tracking.
Rule broken #3: Quick sizing charts are for speed, detailed estimates are for resource management. What I did instead: Used quick sizing for directional planning, stopped pretending it could become a foundational capacity management system.
The client context mattered: organizational shifts meant internal coordination was still taking shape. They needed strategic partnership during transition, not rigid gatekeeping.
The shift required patience to let the old system show its limitations while building evidence for the new approach.
4. The Real Insight
AI amplifies clarity, but team members need permission to break rules that no longer serve the situation. The most valuable thing I did wasn’t my billable hours on a deliverable (that would change after 1 week) or frameworks. It was giving the team permission to operate differently by articulating why the old rules were creating the chaos they were trying to prevent. The breakthrough: Helping PMs understand when to play zone coverage (domain-level oversight across shifting initiatives) vs. one-on-one matchup (integrated campaigns needing dedicated attention). Context-aware judgment beats universal process.
5. Try This If…
If your team is drowning in process designed for a different operating reality:
- Use AI to review transcripts and identify what’s actually driving change in the work
- Map where validation checkpoints create delays vs. add value
- Articulate clearly: What’s outside your control? What’s in your control?
- Build a culture where speaking up about off-track signals is expected
- Give your team language for when to shift between operating modes
The forcing function: Strategic risk taken with eyes open beats false certainty from outdated forecasts.
6. Systems Lens
Process is a cultural artifact, not a universal truth. Traditional PM methodology optimizes for predictable workflows and clear contractual boundaries. Relationship-based account management optimizes for trust and strategic flexibility in ambiguous conditions. Neither is wrong, but applying the first to the second creates expensive theater.
The deeper pattern: We often add more gatekeeping when what we really need is better integration. More PMs doing validation didn’t solve the problem. Domain leads with portfolio visibility and strategic judgment did.
AI helped me see patterns across dozens of conversations that my brain couldn’t hold simultaneously. The result? A team operating with clarity about priorities, integration about signals across workstreams, and confidence that someone is watching the portfolio even when individual projects are chaos.
The meta-lesson: Sometimes the best PM work is knowing which PM rules to break and having the courage to break them visibly.