Scope Is Where Hidden Rework Begins

A delivery executive's best number said her teams were productive. Two weeks inside the work showed they were guessing at scope, building, and rebuilding once stakeholders finally looked.

02 SCOPE
The condition

Throughput looked healthy while the team was quietly doing much of its work twice.

A Good Number and a Tired Team

A senior delivery executive at a global research and advisory firm sent us a short note. Her throughput numbers looked good. Work was going out the door. But she was hearing more complaints from inside her teams than she used to, and people said they felt exhausted. She could not tell whether she had a delivery problem, a stakeholder problem, or something else. She asked us to spend a couple of weeks with the teams and tell her what we saw.

That combination deserves attention. When the metric leaders watch says healthy and the people doing the work say tired, one of them is measuring the wrong thing.

Two Weeks Inside the Work

We embedded with the teams she was most concerned about. We sat in refinement sessions, planning meetings, stand-ups, and retrospectives, and watched how work moved from request to delivery. By the end of the second week, three pieces of the picture had come into focus.

The first was the team's Scrum Master. On paper she facilitated the team. In practice she was acting as project manager for four separate streams of work: chasing requirements, clarifying scope, brokering decisions between groups that were not talking to each other, and writing down answers so the team had something concrete to build from. The role she had been hired for had disappeared beneath the work she had absorbed.

The second was product ownership. The organization had three formally assigned product owners, all senior leaders with broad portfolios, for whom product ownership was one responsibility among many. Refinement sessions were scheduled and they did not attend. Backlog items arrived with unfinished acceptance criteria and no one was available to clarify them. Decisions sat unresolved for days.

The third was how the team had adapted. Fully scoped work never arrived, so they stopped waiting for it. They worked from incomplete information, built what they believed stakeholders wanted, and waited to hear what they had guessed wrong. The work shipped on time. Then portions of it were reworked once stakeholders finally reviewed it.

Figure 1
How the work got done twice
  1. Where it broke Request arrives Acceptance criteria unfinished
  2. Where it broke Product owner absent Refinement skipped, decisions wait
  3. Team guesses scope Builds on assumptions
  4. Work ships on time Counted as throughput
  5. Where it broke Stakeholders review Portions sent back
  6. Where it broke Team rebuilds Counted again
Every pass through this loop added to throughput. The rework never appeared as a cost, only as more output.

The Most Misleading Metric in the Room

When we brought the picture back, we told her plainly that throughput was the most misleading metric in her organization. It was high because the team was constantly working, and the team was constantly working because it was doing much of the work twice: once when it guessed what stakeholders needed, and again when stakeholders said what they actually wanted.

Throughput measured the visible thing. Lead time from request to accepted delivery would show the rework. The Scrum Master's calendar would show a role that no longer matched the job she was being forced to do. The team had folded the cost of working without scope into its daily routine, so no report carried it. And because nobody knew which work would come back, the team could not plan with confidence.

She was not pleased. Cost control was her responsibility, and the number she had trusted had been hiding a cost she owned.

Colleagues working through printed documents together at a table.

An Hour That Was Not Optional

The recommendation was simple. The team needed a formal weekly refinement session, backed by leadership and attended by the people with authority to make scope decisions. Attendance would not be optional. The team would bring the items that needed clarification, and product owners would arrive prepared to answer and decide. Any item stuck on missing stakeholder input would be named and tracked, with the responsible product owner accountable for clearing it. When a product owner missed a session or left a question unanswered, the cost of that delay would be made visible instead of being absorbed by the team.

She supported it and predicted the objection: senior leaders would say they did not have time. They would. But they lacked the time because no one had ever held them to the commitment the role actually requires. Once that commitment was visible, each would have to choose between giving the role its time and handing it to someone who could.

She set the expectations with the product owners in her own voice. The first session ran the following week. Two of the three product owners attended; the third sent a delegate. The team brought items that had been sitting for weeks and, for each one, laid out the unanswered questions, the assumptions they would otherwise have to make, and the cost of getting those assumptions wrong. The Scrum Master facilitated. Questions were answered, decisions were made, and twelve items moved from unrefined to ready.

Afterward the delegate stayed behind. He said he would tell the product owner he had stood in for that this was the most useful hour he had spent in a month, and that the product owner needed to be there himself next week. He was.

Figure 2
From first session to habit
  1. Week 1
    What changed First session Two of three owners; twelve items ready
  2. Week 2
    What changed Absent owner attends His delegate sent him
  3. Session 4
    Pattern holds Owners attend, items arrive prepared
  4. Two months
    Dashboard review Lead time and rework down
Attendance became consistent within four sessions. The measured change showed up two months later.

What Changed When Scope Came First

By the fourth session the pattern had settled. Product owners attended consistently. Items arrived better prepared because stakeholders knew the team would challenge incomplete scope. The team worked from refined items more often than raw requests. The Scrum Master's calendar looked different. Some project management remained, but enough requirement chasing had gone that she could spend her time coaching the team.

Two months later the team's dashboard told an honest story. Throughput dropped slightly. Lead time from request to accepted delivery dropped substantially, and so did rework. Retrospectives moved from frustration about unclear expectations to operational problems the team could solve on its own.

The executive's own conclusion was that the team had looked healthy because the metric she watched said it was. What changed the picture was not more effort or a new tool. It was a place, on a cadence the team could plan against, where the people who define scope met the people who build the work before it started.

Figure 3
What each measure said
MeasureBeforeTwo months on
ThroughputHighSlightly lower
Lead time to acceptanceUnwatchedSubstantially lower
ReworkAbsorbed by teamLower
Scrum Master's timeChasing scopeFacilitating, coaching
The number leaders watched went slightly down. The numbers that carried the real cost went down much further. Shown as the direction of change, not exact values.

Throughput counts what left the building. Lead time to accepted delivery counts what it took to get it right.

Share With Your Team

Choose who it is for. You send it yourself from your own email; Agileon never writes to your colleagues.

Not sure where to start?

Answer five questions about where work stalls. We send back a starting plan: where to begin and what to look at first.