top of page

Workflow Performance Indicators Explained

4 days ago
5 min read

What Are Workflow Performance Indicators?

Workflow Performance Indicators (WPIs™) are structured metrics defined and standardized by the Work Management Institute (WMI) to measure how work moves through a workflow, focusing on flow, delay, quality, stability, and predictability rather than outcomes or individual productivity. They fall into three categories: Flow Indicators, Quality Indicators, and Stability Indicators.

WPIs measure the workflow, not the people in it. They show where work waits, where it comes back for rework, and where performance becomes unpredictable. That makes them the measurement layer of workflow architecture: architecture designs how work should move, and WPIs show whether it actually does.

WPIs put into practice Measurable Performance, one of the 7 Workflow Architecture Standards. A workflow without WPIs can be designed, but it cannot be managed or improved with evidence.

The Three Categories of WPIs

WPIs fall into three categories. Each answers a different question about a workflow.

  • Flow Indicators: Is work moving? They measure speed and volume through the workflow.

  • Quality Indicators: Is work moving correctly? They measure how often work has to be redone, clarified, or rejected.

  • Stability Indicators: Is work moving predictably? They measure how consistent performance is over time.

Flow Indicators

Indicator

What it measures

What a problem looks like

Cycle time

Time from start to completion

Simple work takes weeks

Throughput

Items completed per period

Output stays flat as demand grows

Wait time

Time work sits idle between steps

Work sits for days at handoffs

Queue size

Items waiting at a step

Work piles up at one person or team

Flow Indicators are covered in more depth in What Is Flow in Work Management?

Quality Indicators

Indicator

What it measures

What a problem looks like

Rework frequency

Work sent back to a previous step

The same item cycles through review repeatedly

Error or defect rates

Errors found in completed work

Problems surface after work is "done"

Clarification requests

Questions needed before work can proceed

Requests arrive incomplete or unclear

Approval rejections

Work rejected at an approval step

Submitters do not know the criteria

Stability Indicators

Indicator

What it measures

What a problem looks like

Variation in completion time

How much cycle time differs between similar items

Some items take a day, similar ones take a month

Throughput fluctuations

Shifts in output over time

Completions rise and fall with no clear cause

Sudden spikes in workload

Surges in incoming work

One busy week breaks the workflow

Unstable queue growth

Queues growing unevenly or without recovering

A backlog keeps building at one step

The three categories work together. A workflow can be fast but sloppy (good flow, poor quality) or accurate but erratic (good quality, poor stability). Tracking all three prevents improving one at the expense of another.

WPIs vs. KPIs vs. Productivity Metrics

KPIs measure outcomes, productivity metrics measure effort, and WPIs™ measure flow. Organizations need KPIs, but KPIs alone cannot show why results are slipping. WPIs explain the KPI.

Measure

What it tracks

Example

Question it answers

KPI

Outcomes

Customer retention rate

Are we achieving our goals?

Productivity metric

Effort and output

Tasks completed per person

Are people busy?

WPI

Flow

Wait time in client onboarding

Is the work moving well enough to achieve our goals?

Productivity metrics sit at the Activities layer of the Work Value Pyramid, which separates Activities from Progress and Outcomes. WPIs connect the two: they show whether activity is turning into progress.

Other Meanings of "WPI"

The acronym WPI is used for other concepts, which are different from Workflow Performance Indicators:

  • Work Performance Indicators: a set of executive metrics (mix, capacity, velocity, quality, and engagement) promoted by Adobe Workfront for measuring an organization's work at a high level.

  • Work Performance Information: a project management term from the PMBOK Guide for project data analyzed against the project plan.

Workflow Performance Indicators are narrower and more operational than either. They measure a specific workflow, and each one has a named owner who acts on it.

Every WPI Needs a Signal Owner

A WPI without an owner is a number no one acts on. Under the IDEAS Workflow Ownership Model, every WPI has a Signal owner: the person accountable for watching the indicator. Each WPI should also have a monitoring cadence and a clear response when thresholds are exceeded. Without ownership, signals are ignored and issues escalate.

Signal is the final domain of IDEAS (Intent, Design, Execution, Alignment, Signal). It closes the loop: the signal from a WPI informs the next round of Design, so the workflow improves based on evidence rather than opinion.

How to Implement WPIs

Start with one workflow and a few indicators, then expand.

  1. Pick one workflow that matters. Choose a recurring workflow with visible friction, such as client onboarding or content approval.

  2. Define start and end points. Decide exactly when an item enters the workflow and when it counts as complete. Cycle time is meaningless without this.

  3. Choose one indicator per category. A good starting set is wait time (Flow), rework frequency (Quality), and variation in completion time (Stability).

  4. Assign a Signal owner for each. Name the person who reviews the indicator and decides what to do when it moves.

  5. Set a baseline before changing anything. Measure the current state for several weeks so improvements can be proven.

  6. Review on a fixed cadence. Look at WPIs regularly, and change the workflow design, not just the people, when they slip.

WPIs for Workflows With AI Agents

WPIs apply equally when AI agents perform steps in a workflow. They are often the clearest way to tell whether an agent is helping. An agent that cuts cycle time but raises rework or clarification requests has moved cost, not removed it. Stability Indicators also help detect agent drift, a core concern of AI Workflow Governance™, when an agent's output becomes inconsistent over time.

Frequently Asked Questions

What are workflow metrics?

Workflow metrics measure how work moves through a defined process. Workflow Performance Indicators organize them into three categories: Flow (cycle time, throughput, wait time, queue size), Quality (rework frequency, error or defect rates, clarification requests, approval rejections), and Stability (variation in completion time, throughput fluctuations, sudden spikes in workload, unstable queue growth).

What is the most important workflow metric?

For most knowledge work, wait time. It usually accounts for most of a workflow's cycle time and points directly to where work gets stuck.

How many WPIs should a workflow have?

Start with three: one Flow, one Quality, and one Stability indicator. Add more only when someone owns them and will act on them.

Can WPIs be used to evaluate employees?

They should not be. WPIs measure workflow design, and most workflow problems come from structure, not individual effort. Using WPIs to judge people encourages gaming the numbers and hides the real causes.

Do you need software to track WPIs?

No. Most work management tools can report cycle time and queue size, but WPIs can start in a spreadsheet. What matters is a consistent definition of when work starts and ends, and a Signal owner who reviews them.

bottom of page