The study recorded cancellation on 56 of 150 engagements (37.33%). Comparisons by institution type, outsourcing model, and vendor type did not detect statistically significant rate differences after the applicable adjustments. This does not establish equal rates across those groups.
Among the 94 active engagements, 19 had a mean original-scale cancellation-likelihood rating of at least 7 out of 10, and another 39 had a mean of at least 5 but below 7. Combining these 58 engagements with the 56 canceled engagements yields 114 of 150 (76.00%). The combined measure describes realized cancellation and perceived risk, not a forecast of eventual cancellations.
Scroll horizontally to view the complete figure.
Most engagements were canceled or rated at moderate or high perceived risk | |||
56 Canceled of 150 engagements | 19 High perceived of 94 active | 39 Moderate perceived of 94 active | 36 Other active of 94 active |
Note: Risk bands use the engagement mean of cancellation likelihood among the 94 active engagements: high risk 7 or above, moderate risk 5 or above and below 7. Canceled or at risk: 114 of 150 engagements (76.00%).
High perceived risk means a mean cancellation-likelihood rating of 7 or above. Moderate perceived risk means a mean rating of at least 5 and below 7. These ratings express respondents’ perceptions. The strip partitions all 150 engagements. Risk categories use the 94 active engagements as their base.
Cancellation contrasts compare ratings recorded after the outcome. Active-engagement analyses avoid that particular temporal problem but remain cross-sectional and subject to confounding and selection. The cancellation-likelihood item requires an additional distinction because respondents on canceled engagements also answered it.
“If I am honest we were flying blind. We asked for custodian access in week two and still didn't have it when the plug was pulled.” -Engagement Manager, BPO, Trade, settlement and reconciliation
1 Realized Cancellation
Realized cancellation is the engagement status recorded in the study dataset. Unlike a rating of future likelihood, it identifies an event reported as already having occurred. Its associations describe differences between canceled and active engagements. The status record does not independently verify the sequence or causes of their operational difficulties.
1.1 Realized Cancellation and Its Association With Indicator Ratings and Intent to Leave
The Engagements sheet records 56 of the 150 engagements as canceled (37.33%). Canceled engagements had a markedly lower engagement indicator mean (r = −0.72, 95% CI −0.79 to −0.63, p < 0.001, n = 150 engagements). Respondents on canceled engagements reported markedly higher intent to leave (r = 0.76, cluster bootstrap 95% CI 0.70 to 0.81, p < 0.001, n = 798 respondents). The cluster bootstrap resamples whole engagements to retain the grouping of respondents within engagements. Both associations held in every robustness check computed.
No tested segment difference in cancellation rate remained statistically significant after Holm adjustment. That procedure limits false-positive conclusions within a specified family of comparisons. This includes institution type, outsourcing model, process family, class of work, and vendor type. The satisfaction distributions for canceled and active engagements did not overlap, as detailed in section 1.1 of “Customer Satisfaction, Delivery Performance, and Measured Return”. Their correlation is a full-sample descriptive contrast rather than an active-engagement estimate.
The ratings of canceled engagements were given after the cancellation, and the associations describe how those engagements were rated, not what preceded the decision. The written responses from canceled engagements describe requests that went unanswered until the engagement ended.
“I flagged forty odd vendor records with two different bank details back in March, no answer, chased it twice, still nothing when the account got pulled. There is no one on their side who is allowed to say yes to anything, so the queue just grows and we work round it.” -Individual Contributor, BPO, Corporate and back-office functions
“Before that I would put a surveillance rule change in our tracker and it would just sit. … We do not have a shared risk list either. We keep ours in jira and they keep theirs in whatever they use, and when an alert set went wrong last month it took two days to even agree it was a problem.” -Individual Contributor, ITO, Regulatory, audit, tax and legal
“The vendor's monthly report aggregated error rates across all payroll cycles run that month, diluting the visibility of the specific cycle where a paycheck error triggered the cancellation, so the dashboard never flagged the real risk building underneath.” -Individual Contributor, BPO, Corporate and back-office functions
2 Perceived Cancellation Likelihood
The active-engagement risk bands use only the 94 engagements still operating. The broader indicator analysis includes numeric answers from all 150 engagements, despite wording that restricted the question to engagements not yet canceled. The distinction is necessary when interpreting its full-sample correlations.
2.1 Delivery Team Estimates of Cancellation Likelihood
The cancellation-likelihood item received 328 numeric responses across 150 engagements and 269 “I do not know” responses. Its favorable-direction mean and median were both 5.0. Higher values indicate lower perceived likelihood of cancellation. These pooled summaries include responses from canceled engagements and should not be read as prospective ratings for active engagements alone.
Scroll horizontally to view the complete figure.
Scroll horizontally to view the complete comparison.
| Comparison | Correlation r | Estimate and analysis base |
|---|---|---|
Cancellation | r = −0.54 [−0.65, −0.42] | |
Customer satisfaction | r = 0.66 [0.52, 0.76] | |
Indicator mean | r = 0.69 [0.59, 0.76] | |
Intent to leave | r = −0.64 [−0.72, −0.55] | |
−10+1 |
Exact comparison data
Exact chart data
| Series | Point index | X | Y |
|---|---|---|---|
| 95% confidence interval | 0 | -0.64547125809999994 | 1 |
| 95% confidence interval | 1 | -0.41710830069999999 | 1 |
| Reported association | 0 | -0.54119301519999996 | 1 |
Exact chart data
| Series | Point index | X | Y |
|---|---|---|---|
| 95% confidence interval | 0 | 0.52380435030000005 | 1 |
| 95% confidence interval | 1 | 0.75842137720000002 | 1 |
| Reported association | 0 | 0.65672229569999996 | 1 |
Exact chart data
| Series | Point index | X | Y |
|---|---|---|---|
| 95% confidence interval | 0 | 0.59355624910000004 | 1 |
| 95% confidence interval | 1 | 0.7642893596 | 1 |
| Reported association | 0 | 0.68833996580000001 | 1 |
Exact chart data
| Series | Point index | X | Y |
|---|---|---|---|
| 95% confidence interval | 0 | -0.71731943639999995 | 1 |
| 95% confidence interval | 1 | -0.54673442750000001 | 1 |
| Reported association | 0 | -0.64220136770000003 | 1 |
Note: Survey question: “If the engagement has not been canceled, how likely is the customer to cancel the engagement?”.
Distressed ratings: 76 of 328 answers (23.17%).
Group means and numeric-answer bases: Individual contributor (n = 145): 4.72. Engagement manager (n = 150): 5.06. Portfolio director (n = 31): 6.26. Build-Out (n = 76): 5.59. Transition/Cut-Over (n = 91): 5.91. Production (n = 40): 5.92. Canceled (n = 121): 3.70.
r is Pearson’s correlation coefficient, measuring signed linear association from −1 to +1. Whiskers show 95% confidence intervals where reported. Intent-to-leave intervals, where shown, use a cluster bootstrap. The bootstrap resamples engagements to preserve the grouping of respondent answers. Associations are observational.
The indicator is converted to the favorable direction as 11 minus the original rating before these comparisons.
The favorable-direction rating was associated with lower realized cancellation (r = −0.54, 95% CI −0.65 to −0.42, n = 150 engagements), higher customer satisfaction among active engagements (r = 0.66, 95% CI 0.52 to 0.76, n = 94), and a higher engagement indicator mean (r = 0.69, 95% CI 0.59 to 0.76, n = 150). At the respondent level, it was associated with lower original-scale intent to leave (r = −0.64, cluster bootstrap 95% CI −0.72 to −0.55, n = 316). These results passed the applicable checks in the 22-check protocol. The cancellation contrast includes retrospective responses and cannot establish predictive accuracy. Distressed ratings came from 76 of the 328 answers (23.17%), and the engagement mean was distressed on 26 of the 150 engagements.
Individual contributors judged cancellation more likely than portfolio directors, at a mean of 4.7 (n = 145 respondents) against 6.3 (n = 31 respondents). Nine written responses addressed the likelihood of cancellation. Seven, all from canceled engagements, described an end the respondent had expected.
“…this whole engagement already feel like it is ending badly so one more issue doesnt change much” -Individual Contributor, KPO, Regulatory, audit, tax and legal
“i dont know if anyone above me even knows, happens almost every day, i just override it and keep moving because nobody will fix this before it gets canceled” -Individual Contributor, CXM, Client servicing and communications