Phased Disengagement, Not an Abrupt Cut or Sustained Engagement: Stepping Back from a Difficult Teammate and Short-Term Team Throughput

Abstract

In 188 teams across ten organizations where one member reported a persistently difficult relationship with a teammate, we compared three ways that member handled it. Members who reduced discretionary contact over six weeks or more and handed over at least one shared task or channel on record ("phased," n=71) saw their teams hold 97% of prior-quarter throughput in the first eight weeks, against 80% when contact was cut to the task minimum within two weeks with no handover ("abrupt," n=64; a 21% gap) and 95% when the member kept full engagement ("sustained," n=53). Abrupt teams also logged more rework (6.2 items per 100 completed, versus 4.9 for phased teams; a 27% gap). By months four to six, phased teams stood at 103% of baseline, abrupt teams at 98%, and sustained teams at 96%. Phased and abrupt members reported similar exhaustion at six months (3.6/7 and 3.5/7), both lower than sustained members (4.4/7), but phased members carried more strain during the transition (4.6/7 at week four, versus 4.1/7 for abrupt members). Among 142 teammates on the other side who took part, those whose colleague had disengaged abruptly felt more shut out (4.5/7) than those whose colleague had phased out (3.5/7; a 29% gap).

1. Introduction

The common advice on a coworker you cannot work with runs in two directions. Keep trying, because the team needs you both. Or cut your losses, because your own wellbeing comes first. Each assumes the choice is about whether to step back.

Our earlier work in this unit looked at disagreement inside teams. It found that how people disagreed mattered more than whether they did. We expected something similar here. Stepping back from a teammate can protect the person who does it. It can also leave gaps in shared work, and the speed of the step may decide how large those gaps are.

We compared team throughput, rework, and the experience of both people involved under three approaches.

2. Method (Summary)

Teams of 5–12 members in ten organizations (software, insurance claims, municipal services, and industrial maintenance) ran a quarterly survey that asked members whether they had a persistently difficult working relationship with a teammate. Where a member said yes and agreed to take part, the team enrolled; one initiating member per team was followed. Of 229 enrolled teams, 41 whose approach fit none of the three definitions below (most often a gradual reduction with no recorded handover, n=14) were excluded, leaving 188.

Approach was classified from task-tracker and email records over the following eight weeks, confirmed by the initiating member. "Phased" meant discretionary contact (meetings, messages, and joint tasks outside assigned work) fell gradually over six or more weeks and at least one shared task, review duty, or communication channel was formally handed to another teammate on record (n=71). "Abrupt" meant contact dropped to assigned work within two weeks with no recorded handover (n=64). "Sustained" meant no reduction (n=53).

Team throughput was work items completed per week, expressed as a percentage of the same team's average in the preceding quarter, from tracker records. Rework was items reopened within 30 days per 100 completed. Initiating members self-reported strain at week four and exhaustion at six months (1–7). The teammate on the other side of each relationship was invited, without being told why, to rate at week eight how far they felt shut out of the team's work (1–7, higher meaning more shut out); 142 took part (55 phased, 47 abrupt, 40 sustained).

3. Findings

Over weeks one to eight, phased teams held 97% of baseline throughput, abrupt teams 80% (a 21% gap), and sustained teams 95%. Rework ran at 4.9 items per 100 completed in phased teams, 6.2 in abrupt teams (a 27% gap), and 5.0 in sustained teams.

By months four to six, phased teams stood at 103% of baseline, abrupt teams at 98%, and sustained teams at 96%.

At six months, initiating members reported exhaustion of 3.6/7 (phased), 3.5/7 (abrupt), and 4.4/7 (sustained). At week four, phased members reported strain of 4.6/7, against 4.1/7 for abrupt members (a 12% gap) and 4.5/7 for sustained members.

Teammates on the other side rated feeling shut out at 4.5/7 after an abrupt disengagement, 3.5/7 after a phased one (a 29% gap), and 3.2/7 where engagement was sustained.

4. Discussion

Phased and abrupt members ended up about equally exhausted at six months, and less so than sustained members. Their teams did not end up in the same place. An abrupt step back took a fifth off team output for two months and raised rework, and the teams had not fully caught up by month six. A phased step back cost the team little in the first eight weeks, and over months four to six those teams ran ahead of their own baseline.

The handover is the likely reason. In abrupt cases, work that had run through the two teammates stopped without anyone else taking it on. In phased cases, someone did. The design cannot separate the slow pace from the handover, because every phased member did both; the excluded teams that reduced contact gradually without a handover are the comparison that could, and fourteen were too few for a reliable comparison.

The phased members paid for this. They spent six weeks or more in contact they had already decided to reduce, and at week four they reported more strain than abrupt members. That cost bought something for others: a team that kept producing, and a teammate who felt less pushed out. The shut-out ratings after phased disengagement were close to those where nothing changed.

Sustained engagement kept output steady in the first eight weeks and kept the other teammate included, though by months four to six those teams ran slightly below both other groups, and the member bore the cost. At six months they were the most exhausted of the three groups.

5. Limitations and Future Directions

Members chose their own approach, and those who phased out may have had more cooperative teammates or more slack in their teams to absorb a handover. Throughput as a share of each team's own baseline controls for team size and type of work but not for seasonal changes in workload. Strain and exhaustion were self-reported by members who knew their approach. Excluding the 41 teams whose approach fit none of the three definitions likely sharpened the contrast between groups. Approach was classified over the same eight weeks in which throughput was measured, so early declines in team output may have prompted some abrupt cuts rather than followed from them. The 142 participating teammates were a self-selected 76% of the 188 invited. A follow-up will randomly offer a structured handover template to half of the members who report a difficult teammate.

Editorial Note

This report underwent CfAHC's Rapid Review Protocol, including AI-assisted literature cross-checking and a two-reviewer human sign-off.