An underperforming PM needs a diagnosis, not a verdict. Separate whether the problem is a skill gap (can't yet), a will gap (won't), or a fit gap (wrong seat) — each demands a different intervention. Run a structured 90-day window with clear expectations, weekly evidence-based feedback, and an honest decision gate before you coach, reassign, or exit.
Quick Answer: Diagnose skill vs. will vs. fit before building a plan. Skill gaps get coaching and practice reps. Will gaps get a direct conversation about commitment, not more training. Fit gaps get reassignment or exit. Run all three through a 90-day structure — expectations set in week one, weekly evidence logged throughout, a firm decision on day 90.
Skill, Will, or Fit: The Diagnostic Before Any Plan
Most performance plans fail because they treat every underperforming PM the same way — more documents, more process, more meetings. The fix starts earlier: name which of three distinct problems you're actually looking at, because coaching a will problem wastes months, and firing a skill problem you never actually tried to fix wastes a PM you could have kept.
The three-way split borrows from Situational Leadership II (Hersey and Blanchard), which plots people on two independent axes: competence (can they do the task) and commitment (do they want to). Andy Grove extended this in High Output Management as "task-relevant maturity" — the same person can be highly capable on roadmap reviews and genuinely under-skilled at cross-functional negotiation. A fit gap is the third axis Grove's model doesn't fully cover: the role itself may no longer match what the person is actually good at, regardless of effort or ability.
| Skill Gap | Will Gap | Fit Gap | |
|---|---|---|---|
| What it looks like | Inconsistent output quality; struggles with a specific competency (data modeling, stakeholder facing, roadmap synthesis) | Meets the bar when watched, coasts otherwise; disengaged in 1:1s; misses self-directed deadlines | Technically capable but chronically misaligned with the role's actual demands (e.g., a builder PM in a pure stakeholder-management seat) |
| Root cause | Never taught, never coached, or promoted before ready | Motivation, trust, or burnout — not capability | Role-person mismatch, not a performance defect |
| Right response | Structured coaching, paired work, explicit skill-building reps | Direct conversation about commitment; clarify consequences | Reassignment to a better-fit role, or exit if none exists |
| Wrong response | Performance-managing them out for something never taught | More training on a skill they already have | Coaching harder at a mismatch that coaching can't fix |
| Typical resolution window | 30–60 days with deliberate practice | 2–3 weeks — clarity forces a visible choice | Immediate reassignment conversation; exit if no seat exists |
The table above isn't diagnostic on its own — it's a hypothesis-generator. You still need evidence, which is what the 90-day structure below is built to produce.
Why "Not Meeting the Bar" Is a Symptom, Not a Diagnosis
Vague feedback is the single biggest reason turnaround attempts fail before they start. "Needs to level up" or "not meeting the bar" gives a PM nothing to act on and gives you nothing to measure against 90 days later — it's a verdict dressed up as feedback.
Gallup's manager research is blunt about the stakes here: manager behavior accounts for roughly 70 percent of the variance in team engagement scores. When the manager's own read on performance is fuzzy, that fuzziness propagates downward — the PM can't tell if they're failing at prioritization, at stakeholder communication, or at something else entirely.
Two disciplines fix this before day one of any formal plan:
- Write down specific evidence, not impressions. "Missed the roadmap review deadline twice" beats "seems disorganized."
- Calibrate your own judgment before you calibrate theirs. A structured decision journal is built exactly for this — logging what you predicted about a PM's trajectory against what actually happened catches your own recency bias and halo effects before they contaminate a performance conversation.
The Center for Creative Leadership's SBI model — Situation, Behavior, Impact — is a useful format for turning an impression into evidence: name the specific situation, describe the observable behavior, then state the impact it had. "In yesterday's roadmap review (situation), the update arrived after engineering had already escalated (behavior), which meant the VP heard about the slip from support, not from us (impact)." That structure is repeatable enough to use in every weekly check-in without reinventing it each time.
Without that discipline, you risk two failure modes: coaching someone who's actually disengaged (a will problem disguised as a skill problem), or writing off someone who was simply never taught a specific competency, like translating messy qualitative interviews into a usable customer journey map. That second case is common and fixable — it's a named skill gap, not a character flaw.
The 90-Day Turnaround Structure
A 90-day window works because it's long enough to see a real pattern (not one bad sprint) and short enough that the ambiguity doesn't quietly become permanent. The structure has three phases: a clear-expectations contract in week one, weekly evidence-based feedback for the following eleven weeks, and a firm decision gate on day 90 — not a soft "let's see how it goes."
Days 1–7: The Expectations Contract
Before any clock starts, put the standard in writing and get explicit agreement that it's fair and understood. This isn't a legal document — it's a shared reference point neither of you can later dispute.
- Name the specific gap in behavioral, observable terms ("stakeholder updates arrive reactively, after escalation" not "poor communication").
- Define what "met" looks like with 2–3 concrete, observable markers per area.
- Set the cadence — weekly 1:1s, what evidence gets reviewed, who else is informed.
- State the stakes plainly. The PM should hear, in this meeting, that day 90 ends in one of three outcomes: continued growth in the role, reassignment, or separation.
As Andy Grove frames the underlying test in High Output Management: before deciding someone can't do a job, make sure you've actually established whether they could do it — under real support, with real stakes, in a real attempt.
Days 8–75: Weekly Evidence-Based Feedback
Every week, review the same three things: what was committed to, what actually happened, and what the gap suggests about skill, will, or fit. Keep the record short — a few bullet points, not a narrative — because the goal is a pattern across eleven data points, not a single compelling week.
The leadership style you bring to these check-ins matters as much as the content — using a coaching tone on a will problem reads as permission to keep coasting. Our breakdown of Goleman's six leadership styles maps cleanly onto the diagnosis:
- Suspected skill gap → the coaching style — patient, developmental, paired work on real artifacts.
- Suspected will gap → something closer to the pacesetting or commanding style — naming the gap directly and letting the consequence land.
Kim Scott's Radical Candor framework is the other half of this: challenge directly, but only from a foundation of caring personally. Weekly feedback that's all challenge reads as building a case file; all care with no challenge reads as tolerance. You need both, every week, on the record.
Day 90: The Decision Gate
By day 90 you should have eleven weeks of evidence, not a gut feeling. The decision gate exists precisely so day 90 doesn't quietly slide into day 120 because the conversation is uncomfortable — an indefinite PIP is worse for the PM and the team than a firm, evidence-backed call.
Common Mistakes That Derail the 90 Days
Most failed turnarounds don't fail because the diagnosis was wrong — they fail because the structure around it collapsed. Watch for these five patterns:
- Skipping the written contract. Verbal expectations get remembered differently by both sides; by week six you're arguing about what was actually agreed.
- Letting week four go unrecorded because things seemed fine that week. A gap in the evidence log is exactly the week a pattern reverses.
- Moving the goalposts mid-plan. Adding new expectations at week five that weren't in the original contract makes the whole process feel like a moving target, not a fair test.
- Softening the day-90 conversation because eleven weeks of relationship-building makes the direct message harder to deliver. This is the moment the earlier discipline is supposed to make easier, not harder.
- Treating the plan as private. A PM with no visibility into how the plan is going, week to week, experiences day 90 as an ambush rather than a conclusion they saw coming.
Coaching a Skill Gap vs. Confronting a Will Gap
A skill gap and a will gap can look identical from the outside — missed deadlines, inconsistent quality — which is exactly why the diagnosis has to happen before the response, not after. Get this wrong and you'll coach someone who was never going to change, or confront someone who genuinely didn't know how.
For a skill gap, the coaching plan needs specificity a generic "shadow a senior PM" arrangement rarely delivers:
- Pair the PM on the exact artifact type they struggle with — a spec, a stakeholder update, a prioritization call — not a generic mentorship hour.
- Give feedback on the work itself within 48 hours, while the context is still fresh.
- Track visible improvement across 3–4 reps of the same task type, not just elapsed time.
For a will gap, self-determination theory (Deci and Ryan) offers a useful lens: disengagement usually traces to a missing sense of autonomy, mastery, or purpose in the role as currently structured. Sometimes the fix is structural — a scope change, a new project, more decision rights. Sometimes the direct conversation itself is the intervention: naming that the pattern has been noticed, and that the next 90 days are the chance to show what changes.
The Day-90 Decision: Coach Further, Reassign, or Part Ways
The decision gate isn't a fourth option to keep extending — it's a forced read of the eleven weeks of evidence you've been building. Three outcomes are legitimate; "let's give it another quarter" without new evidence isn't one of them.
| Evidence pattern across 11 weeks | Likely diagnosis | Recommended path |
|---|---|---|
| Steady, measurable improvement toward the defined markers | Skill gap, resolving | Continue in role; close formal review, keep informal coaching |
| Improvement only when closely supervised; reverts otherwise | Will gap, or a fit gap masquerading as one | Direct conversation about commitment; consider reassignment |
| Strong output on some competencies, no movement on others despite real coaching | Fit gap | Explore a different seat — different product area, different scope |
| No meaningful movement despite clear expectations and real support | Will gap, confirmed | Begin separation process |
A fit gap deserves its own honest look before you default to exit. Borrowing a Jobs to be Done lens on the role itself is useful here: our JTBD framework guide is written for products, but the same question applies to a seat — what job did you actually hire this PM to do, and is that still the job the role requires today? Roles drift as companies scale; sometimes the person didn't change, the job did.
Reed Hastings and Patty McCord's Netflix "keeper test" — would you fight to keep this person if they told you they were leaving — is a blunt but useful gut-check at this stage, precisely because it cuts through sunk cost. If the honest answer is no, and the eleven weeks of evidence back that up, exit is the fairer call for everyone, including the PM.
This article stops at the fork. If your 90 days have already resolved toward exit, the mechanics of doing it well — timing, communication, protecting the person's dignity and the team's trust — are covered separately in managing someone out with dignity. That piece assumes the decision is made; this one is about not making it too early or too late.
Naming the Gap: Where Prodinja Fits
A lot of turnaround attempts stall on the same problem the diagnosis table above is meant to solve: "not meeting the bar" is too vague for either party to act on. This is one narrow place a structured tool can help, without pretending to replace the judgment above.
- Stakeholder communication
- Prioritization rigor
- Spec and requirements quality
- Delivery predictability
Used at the start of a 90-day window, this can turn a vague "not meeting the bar" into a specific, coachable line item you can both point to in week one and re-check in week eleven. It won't run the conversation for you, and it doesn't replace the weekly evidence log — it's designed to make sure the gap you're coaching is the real one.
For the fuller picture of how a moment like this fits into the rest of people leadership — hiring, delegation, difficult conversations, team design — our complete guide to PM leadership is the broader map this article sits inside.
Key Takeaways
- Diagnose before you plan. Skill, will, and fit gaps look similar from the outside but need entirely different responses — coaching a will problem or confronting a skill problem both fail.
- Vague feedback is the most common root cause of a failed turnaround. "Not meeting the bar" gives neither party anything to act on or measure.
- The 90-day structure has three non-negotiable parts: a written expectations contract in week one, weekly evidence-based feedback for eleven weeks, and a firm decision gate on day 90.
- Match your leadership style to the diagnosis, not to your default habit — a coaching style suits a skill gap; directness suits a will gap.
- A fit gap is not a performance failure. It's a role-person mismatch that reassignment often solves faster and more fairly than continued coaching.
- The day-90 decision gate has three legitimate outcomes — continue, reassign, or exit — and "extend the deadline again" isn't a fourth.
- This is the fork, not the exit. Once the decision is made, a separate, dignity-focused process handles the departure itself.
Frequently Asked Questions
How do you know if a PM is underperforming or just under-supported?
Check whether expectations were ever made explicit and whether they had the tools, access, or authority to meet them. If a PM was never given clear markers of "good" or was blocked by scope they didn't control, that's a support gap, not underperformance — fix the support first, then re-evaluate with a real 90-day baseline. Confusing the two is the fastest way to run a plan against the wrong problem.
What is a fair timeline for a PM performance improvement plan?
Ninety days is a common, defensible window: long enough to see a real pattern across roughly eleven weekly check-ins, short enough that ambiguity doesn't quietly become permanent. Shorter plans (30 days) rarely produce enough evidence for a fair decision; open-ended plans tend to drift and demoralize the whole team. Whatever the window, fix it in writing at the start so neither side can quietly extend it later.
Should you start with a formal PIP or an informal conversation first?
An informal, direct conversation should almost always come first — a PIP is the formal record of a problem that should already be familiar to the PM by the time it's written down. If a documented performance plan is the first time someone hears the concern is serious, the process has skipped a step, and it will feel punitive rather than corrective regardless of how fairly it's actually written.
Can a skill gap turn into a will gap over time?
Yes, and it's one of the more common failure patterns. A PM who never gets useful coaching on a real skill gap tends to disengage — the original problem was capability, but by week eight it presents as motivation. This is why the weekly cadence matters: it catches the shift before day 90 instead of discovering it too late, when the will gap has become the dominant problem.
How is turning around a struggling PM different from managing someone out?
Turnaround is the decision that comes before an exit decision — the diagnostic work and 90-day structure that determine whether exit is even the right call. Managing someone out with dignity assumes that decision has already been made and focuses on executing it well; this article focuses on not reaching that fork too early or too late, since both mistakes cost the team credibility.