Rate how well the quality measures both firms use capture the kinds of errors agents actually make

The mean rating was 3.6 and the median 3.0 (n = 792 respondents). Ratings of 3 or below, the distressed range, made up 51.14% of the answers. The rating was associated with all four study measures. Higher ratings came with less cancellation (r = −0.61, 95% CI −0.71 to −0.50, p < 0.001, n = 150 engagements), higher customer satisfaction among active engagements (r = 0.63, 95% CI 0.49 to 0.74, p < 0.001, n = 94 engagements), a higher engagement indicator mean computed without the indicator (r = 0.86, 95% CI 0.81 to 0.90, p < 0.001, n = 150 engagements) and lower intent to leave (r = −0.66, cluster bootstrap 95% CI −0.72 to −0.60, p < 0.001 with standard errors clustered by engagement, n = 743 respondents). After Holm correction, ratings differed by engagement status and lifecycle stage (Welch F 39.96, Holm p < 0.001, omega squared 0.378) and respondent role (Welch F 8.08, Holm p = 0.004, omega squared 0.019).

Few written responses addressed the indicator: 14, of which 11 were negative and 3 positive. Half described quality measures built to find the mistakes people make, such as transposed digits, that would not register the mistakes agents make: an out-of-date policy answer, a funding position right in total and wrong for each merchant, or a mapping that failed silently. The written-sentiment tests reached significance on cancellation and intent to leave, but the base of 14 engagements is too small to carry weight.

Ratings of 3 or below, the distressed range, accounted for 405 of the 792 answers (51.14%). The engagement mean fell in the distressed range in 66 of 150 engagements.

The primary responses numbered 13, all written in answer to C156, “Add any comments on the joint management of this engagement that the ratings in this section do not capture.” The remaining 1 responses, 7.14% of the total, were secondary and came from 1 other questions. Respondents who wrote both a primary and a secondary response numbered 0. Negative responses made up 76.92% of primary responses (10 of 13) and 1 of the 1 secondary responses.

Of the 14 responses, seven stated that the quality measures looked for the wrong errors. Checks built for transposed digits would not catch the way an agent gets a reconciliation break wrong. A policy check scored the accuracy of the answer and not whether it was out of date, the failure that actually occurred. Sampling counted breaks cleared and scored as a pass a funding position that was right in total and wrong for each merchant. A recruiting review scored the shortlist and not the interviews that had to be rebooked. In answer to C055, the question on how the vendor manipulates performance reports and dashboards, one respondent described error rates that counted only errors that raised a system exception, so mappings that were wrong but ran without error never appeared.

Against these, three positive responses said the shared error list or the tracked categories were right, and placed the weakness in the timing of review, in ownership of a shared edit, or in drift between releases.

Of the 14 responses, 11 (78.57%, 95% CI 57.14% to 100.00%) were negative and 6 (42.86%) carried the strongest negative score of −10. Positive responses numbered 3 (21.43%), and no response was neutral. The mean written score was −5.0 (95% CI −7.9 to −1.8, n = 14 responses), and the median was −5.0.

Of the 14 responses, 13 came from respondents who also rated the indicator. Among them, the written score rose with the rating (Spearman rho 0.76, 95% CI 0.25 to 0.99, p = 0.003, n = 13 responses). The written responses corroborated the rating. Of the 10 negative responses with a rating, 7 (70.00%) came from respondents whose rating was in the distressed range.

Respondents who wrote about the indicator and gave a rating numbered 13, with a mean rating of 3.6. The 779 respondents who rated the indicator without writing about it had a mean of 3.6. The difference, 0.0 (95% CI −1.4 to 1.5), included zero. Ratings in the distressed range made up 53.85% of the ratings from respondents who wrote (7 of 13) and 51.09% of the ratings from respondents who did not (398 of 779).

Every figure in the segment tables is at respondent level. Segments with fewer than 30 respondents carry no interval. Segments with fewer than 10 report counts only.

Segment families recorded at engagement level were tested on engagement means. Respondent role and vendor type were tested on respondents. Where an expected count fell below 5, the test of the distressed share used Fisher's exact test for a two-by-two table, marked exact, or a Monte Carlo test, marked simulated. After Holm correction, ratings differed by engagement status and lifecycle stage (Welch F 39.96, Holm p < 0.001, omega squared 0.378). Active engagements in Transition/Cut-Over rated highest (mean 4.6, n = 231 respondents) and canceled engagements lowest (mean 2.2, n = 300 respondents). Ratings also differed by respondent role (Welch F 8.08, Holm p = 0.004, omega squared 0.019). Portfolio directors rated highest (mean 5.0, n = 31 respondents) and individual contributors lowest (mean 3.5, n = 609 respondents). Outsourcing model, institution type, process family, class of work and vendor type showed no difference.

Written sentiment did not differ by engagement status and lifecycle stage across the 14 engagements with a written response (Kruskal-Wallis H 6.98, df 3, p = 0.072). Fewer than two groups held 10 or more responses, so the groups are reported as counts only: active engagements in Build-Out (3), active engagements in Transition/Cut-Over (2), active engagements in Production (3) and canceled engagements (6).

No test across respondent roles was run, because fewer than two roles had 30 or more written responses. Individual contributors wrote 8 and engagement managers wrote 6 responses, reported as counts only.

Cancellation rate is the share of engagements canceled by the study date, measured at engagement level (150 engagements, 56 canceled). Canceled engagements were rated after cancellation. The association is therefore a descriptive contrast and not a forecast of cancellation.

Across the study set, higher ratings came with less cancellation (r = −0.61, 95% CI −0.71 to −0.50, Holm p < 0.001, n = 150 engagements). The association held after Holm correction in fourteen of the 20 segment levels tested.

Customer satisfaction is the percentage recorded against each engagement. The primary estimate uses the 94 active engagements. A descriptive contrast uses all 150 engagements.

Across the active engagements, higher ratings came with higher customer satisfaction (r = 0.63, 95% CI 0.49 to 0.74, Holm p < 0.001, n = 94 engagements). The association held after Holm correction in thirteen of the 16 segment levels tested.

The engagement indicator mean is the mean of an engagement's ratings across the rated indicators, measured at engagement level (150 engagements). The indicator is one of the rated indicators. Each comparison therefore uses the mean recomputed without it.

Higher ratings came with a higher engagement indicator mean (r = 0.86, 95% CI 0.81 to 0.90, Holm p < 0.001, n = 150 engagements). The association held after Holm correction in twenty-four of the 24 segment levels tested.

Intent to leave is the answer to C019, “How likely are you to leave your current job in the next 12 months?”, rated 1 to 10. It is measured at respondent level on up to 798 respondents, with the engagement-level form reported alongside.

Respondents who gave higher ratings reported lower intent to leave (r = −0.66, 95% CI −0.70 to −0.62, Holm p < 0.001, n = 743 respondents). The association held after Holm correction in thirty-seven of the 37 segment levels tested.

Two comparisons set the engagements with at least one negative response against all other engagements. Of the 139 other engagements, 136 had no written response addressing the indicator. These two comparisons therefore do not isolate written sentiment. Of the 11 engagements with a negative response, 6 (54.55%) were canceled, against 50 of the 139 other engagements (35.97%). The difference did not reach significance (Fisher exact p = 0.331). The comparison of customer satisfaction was not computed, because one of the two groups held too few active engagements.

Two correlations test written sentiment directly. Across the 14 engagements with a written response, a more negative engagement mean written score came with cancellation (point-biserial r −0.64, p = 0.014). Across 14 respondents, a more negative respondent mean written score came with higher intent to leave (Spearman rho −0.63, p = 0.016). Of the three comparisons, two reached significance at 0.05.

The four associations were compared in pairs on their engagement-level forms, each oriented so that a higher value is the better outcome, across the 150 engagements with all four measures. After Holm correction, five of the six pairs differed. The association with the engagement indicator mean was the strongest and the association with cancellation the weakest. The associations with customer satisfaction and intent to leave did not differ (Holm p = 0.851).

Answers clustered within engagements (ICC 0.66). Clustering reduced the 792 respondent answers to the equivalent of about 207 independent answers. Respondent-level intervals therefore used the cluster bootstrap. Respondent-level tests of intent to leave used standard errors that allow for clustering by engagement.

No association was equivalent to zero. Each of the four differed from zero. The detection floor shown is the study floor at 150 engagements. Intent to leave was tested on its effective sample size.

Each change was computed before rounding and may differ by 0.01 from the difference between the two printed values.

All 22 checks held. A check holds when its test falls on the same side of 0.05 as the primary test. Where the primary test falls below 0.05, the check must also keep the sign of the primary correlation. The thin-base check removes the engagements in any level of outsourcing model or process family with fewer than 30 engagements. Institution type is not used for that check, because every one of its levels has fewer than 30 engagements.

Each of the four measures was computed from data the indicator does not contain. Each result is therefore a finding and not an arithmetic consequence. The indicator is one of the rated indicators, and the engagement indicator mean was recomputed without it. The correction changed r by 0.00, from 0.86 to 0.86.

Each deviation was computed before rounding and may differ by 0.01 from the difference between the two printed values.

The comparison covered 150 engagements. The indicator followed the study's typical pattern on all four measures, and no deviation held after Holm correction.

Among the 71 Part 2 indicators with 30 or more written responses, the indicator's share of negative responses, 78.57%, ranked 43rd from the highest, against a median of 80.82%. Its mean written score, −5.0, ranked 30th from the most negative, against a median of −4.7.

The indicator most often present in the same responses was C136, “Rate how effectively the customer and vendor jointly manage this engagement now that agents perform part of the work.”, in 5 of the 14 responses (35.71%). C150, “Rate how reliably the joint quality assurance loop catches agent errors before they reach live operations.” (4 responses), C144, “Rate how well the joint change control process keeps pace with the speed of agent behavior changes.” (3 responses), C157, “Rate the factual reliability of the AI tools' outputs on this engagement before any human review.” (3 responses) and C138, “Rate how clearly ownership of each workflow step is assigned when agents, vendor staff, and customer staff work in the same process.” (2 responses) followed.

The most frequent theme was “Detection only by failure”, in 3 responses (21.43%). The second was “Raised it, nothing came back”, in 2 responses.

Just over half of all ratings of how well the quality measures used by both firms capture the errors agents make fell in the distressed range (405 of 792 answers, 51.14%). The rating was associated with all four study measures, and every association held under each of the 22 robustness checks. Canceled engagements rated far below the three active stages (mean 2.2, n = 300 respondents).

The written evidence is small. It describes quality measures carried over from manual work that passed plausible agent errors: correct in total and wrong in detail, accurate but out of date, or wrong without raising an exception.

The evidence is the reported experience of the vendor's delivery teams, collected in a single cross-section. The associations reported here are not causal.

Tests by respondent role and by vendor type treated respondents as independent. The section on clustering reports how far clustering within engagements reduced the information in respondent-level figures.

Customer management executives gave 2 answers, a base of anecdote strength.

Shares of written sentiment are shares of responses and not of the study set of respondents. The written score takes five fixed values, so its mean and median are coarse. Responses clustered within engagements (14 responses on 14 engagements). Intervals for written sentiment therefore used a bootstrap by engagement.

The written responses number fewer than 30, so their shares and the written-sentiment tests rest on a small base.