Project Management KPIs: Metrics Every Manager Should Track

Most project reports are full of activity, not insight. Teams track “percent complete” on tasks, hours worked, and open issues, but nobody can answer the only question that matters: is this project still heading for the outcome we promised, or quietly drifting off course? The difference between activity and progress is measurement. Project management KPIs are the specific, quantified measures that tell a manager whether a project is on schedule, on budget, and on target — before small problems become large ones.

This guide covers the KPIs that actually matter across the four core dimensions of a project — schedule, cost, scope and quality, and resource/process health — with the formulas you need (including earned value metrics), how to select a small set of KPIs for your project, how to set targets, and the tools that make tracking realistic. You will leave with a concrete KPI set you can start using this week, not a list of fifty metrics nobody will read.

Quick Answer: What Are Project Management KPIs?

Project management KPIs (key performance indicators) are the specific, quantified measures that track how a project is performing against its approved baselines — schedule, budget, scope, quality, resources, and risk. Unlike generic metrics, each KPI has a target and a threshold that tells you whether the project is healthy, drifting, or in trouble.

The nuance that matters: a KPI is not just a number on a dashboard. It is a management contract. A good KPI answers a specific question — “are we on budget?” — has a target (“within ±5% of the cost baseline”), a trigger for action (“flag anything above ±10% variance”), and an owner who is responsible for acting on it. Without those four elements, what you are looking at is data, not a performance indicator.

Why KPI Tracking Fails (And How to Make It Stick)

Before listing KPIs, it is worth understanding why most KPI programs die. The usual reasons:

• Too many metrics. A dashboard with forty indicators is not informative; it is noise. People cannot tell which number should drive a decision today.
• Measuring activity instead of progress. Hours logged and tasks moved tell you the team is busy. They do not tell you if the project will finish on time.
• No targets. A metric without a target is a number without a judgment attached. You cannot say “green” or “red” without a threshold.
• Data that is always late. If the KPI report reflects last month’s reality, the decisions made from it are already outdated.
• KPIs nobody owns. When nobody is responsible for a KPI, an amber reading sits in the report for weeks while everyone assumes someone else is handling it.

The fix is discipline, not software: choose a small set, define targets and thresholds, assign owners, and review on a fixed cadence. That is the entire secret, and it is worth repeating throughout this guide.

The Four Core Project Management KPI Categories

The direct answer: the KPIs that matter most fall into four categories — schedule performance, cost performance, scope and quality, and resource and process health. Covering all four gives you a full picture; covering only one gives you a misleading one.

1. Schedule KPIs: Are we on time?

Schedule KPIs answer the most visible project question. The essential ones:

• Schedule Variance (SV) — the difference between earned value and planned value. SV = EV − PV. A negative SV means you are behind schedule in terms of value delivered.
• Schedule Performance Index (SPI) — SPI = EV / PV. An SPI of 1.0 means you are exactly on plan; below 1.0 means behind; above means ahead.
• Milestone hit rate — the percentage of planned milestones delivered on or before their target date.
• On-time task completion — the percentage of tasks finished by their due date, useful as an early operational signal.
• Critical path slippage — days of delay on the tasks that actually drive the finish date, not the tasks with slack.

2. Cost KPIs: Are we on budget?

Cost KPIs measure the money side of the iron triangle:

• Cost Variance (CV) — the difference between earned value and actual cost. CV = EV − AC. Negative means you are spending more than the value you have produced.
• Cost Performance Index (CPI) — CPI = EV / AC. Above 1.0 means you are getting more value per dollar than planned; below 1.0 means you are overspending.
• Budget variance — the simple difference between planned budget and actual spend at a point in time (budgeted cost of work performed vs actual).
• Estimate at Completion (EAC) — EAC = BAC / CPI, the forecast final cost if the current cost efficiency continues. This single number is often the most useful budget KPI in a report.
• Cost of rework — spend on redoing work that was done wrong, an underrated indicator of quality problems.

3. Scope and quality KPIs: Are we delivering the right thing?

A project can be on time and on budget and still fail if the deliverable is wrong or broken:

• Requirements completion — the percentage of approved scope items delivered and accepted.
• Defect density — defects per unit of deliverable (e.g., per module, per 1,000 lines of code, per batch).
• Defect resolution time — average time from a defect being reported to it being resolved.
• Acceptance rate — the share of deliverables accepted by stakeholders on first review.
• Scope change count — the number of approved changes; a high count is not automatically bad, but a high count with no change control is a warning sign.

4. Resource and process KPIs: Are we set up to succeed?

The people and process dimension is the one most managers under-measure:

• Resource utilization — the percentage of a person’s available time that is booked on project work. Sustained utilization above ~85% typically leads to burnout; very low utilization means wasted capacity.
• Cycle time — average time from a task starting to it being done, the most useful lean/agile measure.
• Lead time — average time from a task being requested to it being delivered, including waiting.
• Work-in-progress (WIP) — how many tasks are open at once. Too much WIP slows everything down.
• Team load / workload balance — the distribution of work across the team; one overloaded person and three idle people is a process problem, not a people problem.

Targets above are examples to adapt, not universal standards. The value is the structure: each category has a question, a measure, and a threshold.

What Is Earned Value Management and Why Do PMs Love It?

The direct answer: earned value management (EVM) ties schedule and cost together into one consistent set of numbers, so you can answer “are we behind schedule” and “are we over budget” with the same units and the same data — value earned against value planned.

The core idea is elegant. Instead of comparing hours to hours, EVM converts work into value. Three values drive everything:

• Planned Value (PV) — the value of the work you planned to have done by now.
• Earned Value (EV) — the value of the work you actually completed by now.
• Actual Cost (AC) — what you actually spent by now.

From those three you compute the four indicators above: SV, CV, SPI, CPI. The beauty is that SPI and CPI are ratios, so they work across projects of any size, and EAC = BAC / CPI gives you a forecast.

A worked example

Say a project has a budget at completion (BAC) of $100,000 and a 10-week schedule. At week 5, you planned to have done 50% of the work — planned value $50,000. In reality you have completed 40% — earned value $40,000 — and spent $45,000.

• SV = 40,000 − 50,000 = −10,000. You are $10,000 of value behind schedule.
• SPI = 40,000 / 50,000 = 0.80. You are delivering value at 80% of the planned rate.
• CV = 40,000 − 45,000 = −5,000. You are $5,000 over budget for the value produced.
• CPI = 40,000 / 45,000 = 0.89. You are getting $0.89 of value for every dollar spent.
• EAC = 100,000 / 0.89 ≈ $112,400. If nothing changes, the project finishes about $12,400 over budget.

This is not a theoretical curiosity — it is the fastest way to see the truth. A manager who only tracks “spent so far” sees $45,000 and thinks the budget is fine. The EVM view shows the same project is already behind schedule and over budget, with a concrete forecast of the final overrun.

How Many KPIs Should You Track? (And How to Choose Them)

The direct answer: track 5–8 KPIs maximum, chosen so that each one answers a different core question. If two KPIs answer the same question, keep the one that is more actionable.

A practical selection method is the “question-first” approach. For each KPI you are considering, ask: what decision does this number change? If nothing changes, drop it. Then apply these filters:

• Relevance to your project type. Agile projects lean on cycle time, WIP, and velocity; fixed-bid waterfall projects lean on SPI, CPI, and milestone hit rate. A marketing campaign needs different KPIs than a construction project.
• Data availability. A KPI you can only produce once a month because the data is manual is less useful than one you can produce weekly. If you cannot keep it current, do not promise it.
• Mix of leading and lagging. Lagging indicators (SV, CV, milestone hit rate) tell you what already happened. Leading indicators (EAC trend, resource availability, defect trends, risk exposure) tell you what is coming. You want both.
• Stakeholder appetite. If sponsors only want two numbers, deliver the two best — typically SPI/CPI or their simpler equivalents, plus forecast finish date and forecast cost.

KPI vs Metric: What Is the Difference?

The direct answer: a metric is any measurement; a KPI is a metric that is critical to the project’s success and carries a target, an owner, and a consequence.

For example, “number of tasks completed” is a metric — interesting, but not critical on its own. “Tasks completed versus tasks planned this sprint, with a 90% completion target and a rule that any sprint below 70% triggers a retrospective” is a KPI. The same number, but one has a judgment attached and a decision connected to it. When someone says “we track KPIs” but can point to no target or threshold, what they actually track is metrics.

Scenario: A $250K CRM Implementation (Cost-Heavy Project)

A mid-size services firm is 16 weeks into a 30-week CRM implementation with a $250,000 budget and a fixed price for the client. The PM runs the earned value numbers at week 16:

• PV at week 16: $125,000 (half the project by time).
• EV: $110,000 (44% of value delivered).
• AC: $132,000.

The readout: SV = −$15,000, SPI = 0.88, CV = −$22,000, CPI = 0.83, EAC ≈ $301,000. The project is both late and over budget, and the forecast says it will exceed the fixed price by roughly $50,000 — profit on the deal has essentially disappeared.

The PM’s KPI set (SPI, CPI, EAC, milestone hit rate, defect density, rework cost) makes the rescue decision concrete: add two resources for three weeks to recover schedule (projected cost $18,000), tighten the change control board, and renegotiate the client milestone schedule. Without the EVM KPIs, the team would have discovered the overrun only at the final invoice. With them, they act at week 16 instead of week 30.

Scenario: An Agile Mobile App (Flow-Heavy Project)

A product team builds a mobile app for a 12-week launch. Their KPI set is agile-native: cycle time, WIP, sprint burndown/velocity, and defect escape rate. In week 3, cycle time for features has climbed from 4 days to 6 days and WIP sits at 14 open items against a limit of 8.

The manager treats these as leading indicators. They do not wait for the launch date to slip; they investigate now. The root cause: the design resource is split across two other projects, so features queue behind design. The fix is a WIP-cap on design work and a protected design block in the calendar. Cycle time returns to 4 days by week 5, and the launch stays on the original date. Here the KPIs did not measure failure — they prevented it.

Scenario: A Marketing Campaign (Scope-Heavy Project)

An in-house marketing team runs a product-launch campaign with 24 deliverables, a 6-week timeline, and a $40,000 budget. Their KPI set: milestone hit rate, scope-change count, and cost variance. By week 3, nine of the 24 deliverables are late, and four scope changes have been approved without formal cost impact assessment.

The KPI readout is unambiguous: milestone hit rate at 63% and scope changes climbing. The PM calls a stop on new scope requests, runs an impact assessment on the four approved changes, and discovers two of them add $6,000 in unbudgeted production cost. The campaign is re-planned with the real costs, and the sponsor decides to cut two deliverables instead of blowing the budget. The KPI set made a scope negotiation concrete instead of emotional.

What Tools Should You Use to Track Project Management KPIs?

The direct answer: start with a spreadsheet if you are small, but move to a project management platform with live reporting once the project has more than a handful of people — manual updates do not survive contact with a busy team.

Real options, with honest trade-offs:

Excel / Google Sheets. Pros: free, infinitely flexible, everyone knows it. Cons: manual data entry goes stale fast, no live link to tasks, and formulas break silently. Good for a single small project or a portfolio-level summary you update weekly.

Microsoft Project. Pros: strong scheduling engine, built-in earned value and baseline tracking for classic waterfall projects. Cons: desktop-focused, steeper learning curve, and its EVM reports require disciplined baseline discipline. Good for construction, engineering, and heavy scheduling work.

Jira. Pros: excellent for agile teams with velocity, sprint burndown, cumulative flow, and cycle-time reports out of the box; deep ecosystem. Cons: setup takes effort, and it is not the natural home for budget/cost KPIs. Good for software delivery teams.

Asana. Pros: easy for cross-functional teams, clean dashboards, workload view for resource balance. Cons: EVM-style cost metrics are not native; you track them via custom fields or integrations. Good for marketing, operations, and general business projects.

Smartsheet. Pros: spreadsheet familiarity plus live dashboards, useful for reporting-heavy teams and PMOs. Cons: license cost grows with seats, and some advanced cost logic needs building. Good for PMOs that live in spreadsheet culture.

Power BI / Tableau. Pros: the gold standard for beautiful, portfolio-wide dashboards pulling from multiple systems. Cons: needs a data model and someone to maintain it; overkill for a single small project. Good for enterprise portfolios.

The honest advice: the KPI system is more important than the tool. A clean spreadsheet updated weekly beats a beautiful dashboard nobody updates. But the moment a project has cross-functional teams and weekly reporting, a platform where KPIs update from live task data saves hours and keeps the numbers trustworthy.

To be transparent: Doitify is our product, which is why we know its capabilities from the inside — and it is designed for exactly this workflow. It combines Kanban boards, Gantt charts, sprints, task owners and due dates, quality control, resource and workload management, and work and performance reports in one workspace, so the KPIs in this guide update from the same place the team actually works. For a solo project or a small team that only needs a weekly spreadsheet, a lighter tool is the more honest recommendation; for a team that wants KPIs tied to live execution, Doitify’s unified reporting is worth evaluating.

Common Mistakes

These are the mistakes that make KPI tracking useless or harmful:

• Tracking thirty KPIs and reading none. Each extra KPI dilutes attention. Cut the list until each number could change a decision this week.
• Using “percent complete” as the primary schedule KPI. Percent complete is optimism disguised as data. Use value-based or deliverable-based measures (earned value, milestone hit rate) instead.
• No baseline. KPIs compare against a baseline. If you never froze scope, schedule, and budget, you have nothing to measure against and every number is meaningless.
• Only lagging indicators. Reporting last week’s variance is fine; reporting only last week’s variance means you will always react late. Add forecasts and trends.
• KPIs with no owner. Assign one owner per KPI. An amber KPI with no owner stays amber forever.
• Moving targets. Changing targets mid-project to make the dashboard look better is the fastest way to destroy trust in the numbers. Change the plan, then change the baseline — with approval.
• Ignoring the data quality problem. Garbage in, garbage out. If time entries and task updates are unreliable, every KPI built on them is fiction. Fix the data capture first.
• No review cadence. A KPI dashboard that is looked at once a month for a project that needs weekly steering is theater. Match the review frequency to the project’s pace.

Know This Before You Choose

Before you finalize your KPI set, run through this checklist:

• Can I name the decision each KPI would change? If not, remove it.
• Does my set cover all four dimensions — schedule, cost, scope/quality, and resources? One dimension missing is a blind spot.
• Do I have both leading and lagging indicators?
• Do I have an approved baseline to measure against, and a change-control process to keep it honest?
• Have I defined a target and a trigger threshold for each KPI, with an owner?
• Can my team produce the data on a cadence that matches the project’s pace (weekly for active projects)?
• Have I picked KPIs that fit the methodology — earned value for predictive projects, flow metrics for agile?
• Does my tool actually keep the data current, or will I be chasing manual updates every week?

Conclusion

Project management KPIs are not about collecting numbers; they are about making the invisible visible before it costs you the project. The practical system is simple: choose 5–8 KPIs that each answer a different core question, cover the four dimensions of schedule, cost, scope/quality, and resources, set targets and thresholds with owners, and review them on a cadence that matches the project’s pace. Use earned value metrics (SV, CV, SPI, CPI, EAC) when you need schedule and cost in one coherent view, and flow metrics when you run agile.

Start small. Pick two schedule KPIs and two cost KPIs, add one quality and one resource KPI, and run them for two weeks. The numbers will force conversations you have been avoiding — about estimates, about scope, about who is actually overloaded. That is the point. A project that can be measured honestly is a project that can be steered. Once you have the KPI set working in a spreadsheet, moving it into a project management platform where it updates from live execution is a natural next step for teams that need the truth on a weekly basis.