The 10 KPIs Every Project Controller Should Track (But Almost Nobody Does)

Ask most Project Controllers what KPIs they track, and you'll hear variations of the same list: schedule variance, budget variance, milestone completion rate, risk count.

These are fine. They're also fundamentally lagging indicators — they tell you what has already happened, not what is about to happen.

The 10 KPIs below are different. They're leading indicators: metrics that predict future outcomes rather than report past ones. Master these, and you'll spend far less time managing crises and far more time preventing them.

1. Schedule Performance Index (SPI) Trend

What it measures: Not just your current SPI, but how it has changed over the past 3–4 reporting periods.

Why it matters: A project with SPI = 0.92 is slightly behind. A project with SPI trending from 1.0 → 0.97 → 0.94 → 0.92 is in accelerating decline. The trend tells you far more than the point-in-time value.

Threshold: Alert when SPI drops more than 0.05 in two consecutive periods.

2. Milestone Risk Score

What it measures: The probability that each upcoming milestone will be achieved on time, based on current task progress, velocity, and dependency status.

Why it matters: This is the most actionable forward-looking metric in project control. A milestone with 80% on-time probability needs monitoring. One with 40% probability needs immediate intervention.

How to calculate: Weight three factors — predecessor task completion rate, remaining float on critical path tasks, and historical team velocity — to produce a 0–100 risk score per milestone.

3. Resource Concentration Index

What it measures: The percentage of your critical path tasks assigned to your 3 most utilized team members.

Why it matters: If 3 people own 70%+ of your critical path, your project's fate depends on the health and availability of 3 individuals. This is a hidden risk that rarely appears in standard risk registers.

Threshold: Alert when top-3 resource concentration exceeds 60% of critical path.

4. Decision Debt

What it measures: The number of decisions that are pending executive or steering committee approval, and how long each has been waiting.

Why it matters: Pending decisions create frozen tasks. Teams can't progress on work that requires a decision. High decision debt is a leading indicator of upcoming delays.

Threshold: Alert when any critical-path decision has been pending for more than 5 business days.

5. Scope Creep Rate

What it measures: The percentage increase in scope (task count, estimated hours, or deliverable count) since project baseline.

Why it matters: Every unbudgeted task added to the scope is a hidden overrun. Most PMOs track this reactively in change requests; tracking it continuously reveals unauthorized scope creep before it's formalized.

Threshold: Alert when scope has increased more than 5% from baseline without a corresponding approved change request.

6. Action Item Resolution Rate

What it measures: The percentage of action items from steering committee and project meetings resolved within the agreed timeframe.

Why it matters: Low resolution rates are a governance health indicator. When action items pile up unresolved, it signals organizational dysfunction that will eventually manifest in project failures.

Threshold: Alert when resolution rate drops below 70% in any two-week period.

7. Budget Burn Rate Acceleration

What it measures: Not the current burn rate vs. plan, but whether the gap between actual and planned burn is growing week over week.

Why it matters: A consistent 8% overrun is a known, manageable problem. An overrun growing by 2% per week is an exponential trajectory toward disaster. The acceleration is the signal.

Threshold: Alert when weekly burn rate variance increases by more than 2 percentage points in two consecutive weeks.

8. Team Velocity Consistency

What it measures: The standard deviation of weekly task completion rate over the past 4 weeks.

Why it matters: High variance in velocity (some weeks completing 20 tasks, others completing 4) is a signal of underlying instability — unclear priorities, blocked work, team morale issues, or technical debt. Low variance means predictable delivery.

Threshold: Alert when weekly velocity standard deviation exceeds 30% of mean velocity.

9. Stakeholder Engagement Score

What it measures: Frequency and responsiveness of key stakeholder communications — how often are they participating in meetings, responding to status updates, reviewing deliverables.

Why it matters: Disengaging stakeholders are a leading indicator of project abandonment risk. Catch the disengagement early and you can re-engage. Catch it late and you're managing a political crisis.

Threshold: Alert when a key stakeholder has missed 2+ consecutive touchpoints or has a response latency exceeding 5 business days.

10. AI Confidence Score Delta

What it measures: The week-over-week change in your project's AI health score.

Why it matters: The absolute score matters less than the direction and rate of change. A project at 65 (yellow) but trending upward is recovering. A project at 72 (green) but dropping 5 points per week is deteriorating. The delta is the signal.

Threshold: Alert when AI health score drops more than 8 points in a single week or more than 15 points over two weeks.

Making This Practical

Tracking 10 KPIs manually across a portfolio of projects is not realistic. The value of this list isn't in creating 10 new columns in your Excel spreadsheet.

The value is in knowing which signals to automate.

A good AI-powered project control platform monitors all 10 of these continuously, surfaces the ones that require attention, and suppresses the noise from the ones that are healthy.

That's the difference between a Project Controller who spends their day reacting to problems and one who spends their day preventing them.

Start tracking the KPIs that actually predict outcomes →