Credited Feedback Time, Not an Added Expectation or Training Alone: Who Absorbs the Cost of Precise Manager Feedback, and Whether It Lasts

Abstract

After all 236 managers in 27 divisions of six organizations were trained to give outcome-linked feedback, divisions were randomly assigned to one of three arrangements for 12 months. In "credited" divisions (9 divisions, n=80), managers' project work was reduced by four hours a week and feedback practice was counted in their annual evaluation. In "added-expectation" divisions (9 divisions, n=79), feedback practice was counted in the evaluation with no reduction in other work. In "training-only" divisions (9 divisions, n=77), nothing changed after the training. Sustained practice was measured by reports' fortnightly answers to whether their manager had connected a specific action of theirs to a result. At month 3, 66% of credited and 63% of added-expectation managers met the threshold; by month 12, 58% of credited managers did, against 47% of added-expectation managers (a 23% gap) and 45% of training-only managers (a 29% gap). Added-expectation managers rated their exhaustion at 4.3/7 at month 12, against 3.5/7 for credited managers (23% higher) and 3.7/7 for training-only managers. Among three randomly chosen reports per manager, (baselines 52.1 to 53.0 on a blind-scored 100-point work sample), those of credited managers gained a mean 13.4 points, against 11.0 and 10.6. In the first two quarters, credited divisions completed 81.4% of planned project deliverables, against 87.7% in added-expectation and 88.1% in training-only divisions; the gap with added-expectation divisions narrowed to 1.8 percentage points in the fourth quarter.

1. Introduction

The common assumption is that a good manager finds time for feedback, and that an organization only needs to train managers and make clear that feedback is expected. On that view, the hours a manager spends tracing a report's decisions to their results are part of the job, and nothing else in the job needs to change to make room for them.

Precise, outcome-linked feedback takes several hours a week per manager, and those hours come from somewhere. We tested what happens when the organization gives up some of the manager's other work to cover them, compared with adding the expectation alone and with training alone, and we followed managers for a year to see which arrangement kept the practice going.

2. Method (Summary)

Six organizations took part, contributing 27 divisions and all 236 managers in them. Every manager first attended the same two-day training in outcome-linked feedback: agreeing with each report on actions whose results could be traced within weeks, and holding short regular reviews tracing one recent decision to what followed. Divisions were then randomly assigned, stratified by organization, to one of three arrangements.

Credited divisions (9 divisions, n=80) reduced each manager's project work by four hours a week. Division heads reassigned tasks to other project staff, deferred them, or lowered project targets to make the reduction, and managers' annual evaluations included a section on feedback practice, drawn from their reports' pulse answers. Added-expectation divisions (9 divisions, n=79) included the same evaluation section and made no change to managers' other work. Training-only divisions (9 divisions, n=77) made no change of either kind after the training.

Every two weeks, managers' reports answered one pulse question: "In the last two weeks, did your manager connect a specific action of yours to a result?" Answers were anonymous and reported to managers only as a team percentage. A manager was counted as sustaining the practice at a given month if at least two-thirds of their reports answered yes in the two pulses closest to that month's end. Managers rated their exhaustion at months 3 and 12 (1 = not at all, 7 = completely) in a survey. For skill growth, three of each manager's reports were randomly chosen (708 in all) and completed a role-specific work sample at baseline and month 12, scored from 0 to 100 by two raters blind to condition and timepoint; their scores were averaged. Planned and completed project deliverables were taken from each division's project-tracking records.

3. Findings

Of the 236 managers, 217 were still in post at month 12 (74 credited, 72 added-expectation, and 71 training-only), and the manager results below are for them.

At month 3, 66% of credited managers and 63% of added-expectation managers met the pulse threshold, against 52% of training-only managers. At month 6, the figures were 64%, 56%, and 49%, and at month 9, 61%, 50%, and 46%. By month 12, 43 of 74 credited managers (58%) still met it, against 34 of 72 added-expectation managers (47%, a 23% gap) and 32 of 71 training-only managers (45%, a 29% gap). The added expectation raised practice early, and its effect faded steadily until, by month 12, it was close to the training-only level. Credited managers' practice also declined, but more slowly.

At month 12, added-expectation managers rated their exhaustion at 4.3/7, against 3.5/7 for credited managers (added-expectation ratings 23% higher) and 3.7/7 for training-only managers (16% higher). At month 3, all three groups had rated it between 3.4 and 3.6.

Of the 708 reports chosen for work samples, 641 completed both (219 credited, 214 added-expectation, and 208 training-only). Baseline scores were 52.1, 53.0, and 52.6. Reports of credited managers gained a mean 13.4 points over 12 months, against 11.0 for reports of added-expectation managers (2.4 points, a 22% larger gain for credited managers' reports) and 10.6 for reports of training-only managers (2.8 points, a 26% larger gain). At month 12, the three groups' mean scores were 65.5, 64.0, and 63.2.

In the first two quarters, credited divisions completed 81.4% of their planned project deliverables, against 87.7% in added-expectation divisions (a 6.3-point gap) and 88.1% in training-only divisions. The gap with added-expectation divisions narrowed to 3.5 points in the third quarter and 1.8 points in the fourth, and the gap with training-only divisions to 3.9 and 2.2 points.

4. Discussion

An added expectation is a request for the time with no source named. At month 3, added-expectation managers gave outcome-linked feedback almost as often as credited managers did, and their rate fell at each later measurement. By month 12, they were more exhausted than either other group and gave it about as often as managers who had only been trained. One reading is that they met the expectation at first by working longer, and stopped when they could not keep that up.

Credited divisions paid for the difference. Four hours a week per manager came out of planned project work, and in the first half of the year those divisions finished noticeably fewer deliverables than the divisions that had asked for feedback on top of everything else. By the fourth quarter the gap was small but had not closed, perhaps because reports who had grown more skilled needed less of their managers' time on project work, though our data cannot show that directly.

The benefit went mainly to reports, whose skill grew more. The divisions that gave up the hours were still slightly behind on deliverables in the fourth quarter, though by much less than in the first half of the year. Credited managers avoided the rise in exhaustion seen in added-expectation divisions, ending close to the training-only level, but they were not the main beneficiaries; the arrangement mattered because it kept the practice going for more of the year, though credited managers' rate also declined, from 66% to 58%. When an organization wants managers to give precise feedback, the evidence here suggests it should also decide what work the managers will stop doing to make time for it.

5. Limitations and Future Directions

Divisions knew their arrangement, and division heads in credited divisions may have supported feedback in other ways we did not record. Randomization was by division, and with nine divisions per arrangement, differences between divisions in leadership or workload may account for part of the gaps reported at the manager level. Because pulse answers fed the evaluation section in credited and added-expectation divisions but not in training-only divisions, managers in the first two arms had a reason to make the connection explicit to reports, which may have raised their pulse rates apart from the quality of their feedback; the blind-scored work sample does not share this bias. The pulse question measures reports' perception of a connection between action and result, not the accuracy of the feedback itself. Nineteen managers left before month 12, slightly more often in added-expectation divisions, and if those who left were the most exhausted, the exhaustion gap is understated. The reduction in project work was made differently in each division, so we cannot say whether reassigning tasks and lowering targets had the same effect. Deliverable counts treat all deliverables as equal in size, and we did not measure whether reassigned tasks raised other project staff's workload. All six organizations had project-tracking systems that made the four-hour reduction visible and enforceable; where workloads are less formally tracked, a credit of this kind may be harder to keep. A follow-up will examine whether the reduction can be scaled back after the first year without losing the practice.

Editorial Note

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