CME Respondent Count

The Engagements sheet records, for each engagement, the number of its respondents who were customer management executives. The count was zero on 148 of the 150 engagements and one on two, a mean of 0.01 (n = 150 engagements). All of its variation therefore lies in two engagements.

The count was not associated with any of the four study measures: cancellation (r = 0.03, 95% CI −0.13 to 0.19, p = 0.712, n = 150 engagements), customer satisfaction among active engagements (r = 0.00, 95% CI −0.20 to 0.20, p = 0.993, n = 94 engagements), the engagement indicator mean (r = 0.02, 95% CI −0.14 to 0.18, p = 0.779, n = 150 engagements) and intent to leave (r = 0.00, p = 0.978 with standard errors clustered by engagement, n = 798 respondents). Each association was formally equivalent to zero. With the variation confined to two engagements, these results carry far less information than their sample sizes suggest.

The count is a whole number recorded once for each engagement, and no distressed range applies. The table groups values in bands of one tenth, so only the lowest band, holding the value zero, and the highest, holding the value one, contain engagements. No engagement recorded more than one customer management executive among its respondents.

Every figure in the segment tables is at engagement level. Segments with fewer than 30 engagements carry no interval, and segments with fewer than 10 report counts only. The two engagements with a customer management executive lay in the ITO and BPO outsourcing models, and both were LOB Operational engagements. One was active in Transition/Cut-Over and the other canceled. Every other segment level recorded zero on all its engagements.

No difference between segments was tested. In every segment family, at least one level recorded zero on all its engagements, so that level has no spread and Welch's test is undefined. The other tests in the table were not computed for the same reason.

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, so the association is a descriptive contrast and not a forecast of cancellation.

Across the study set, the count was not associated with cancellation (r = 0.03, 95% CI −0.13 to 0.19, Holm p = 1.000, n = 150 engagements). The odds ratio of 1.69 carried a 95% CI from 0.10 to 27.58, the width that two engagements allow. An association could be estimated only in segment levels holding an engagement with a customer management executive, and it held after Holm correction in none of the seven 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, the count was not associated with customer satisfaction (r = 0.00, 95% CI −0.20 to 0.20, Holm p = 1.000, n = 94 engagements). Only one active engagement recorded a customer management executive, so the estimate rests on a single engagement. The association held after Holm correction in none of the four 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 count is not a rated indicator and does not enter the mean.

The count was not associated with the engagement indicator mean (r = 0.02, 95% CI −0.14 to 0.18, Holm p = 1.000, n = 150 engagements). The association held after Holm correction in none of the nine 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, each carrying the count recorded for the respondent's engagement.

The count was not associated with intent to leave (r = 0.00, p = 0.978 with standard errors clustered by engagement, n = 798 respondents). Every respondent on an engagement carries the same count, so the clustered test governs. The association held after Holm correction in none of the nine segment levels tested. Those segment tests treated respondents as independent.

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, none of the six pairs differed. Every association was near zero.

The count is recorded once for each engagement, so every respondent on an engagement shares its value. Every test of the count used engagements except the respondent-level form of intent to leave. That form used standard errors clustered by engagement.

All four associations were equivalent to zero at the bound shown. The tests and the power figures treat the count as a measure that varies across the engagements. It varies on two engagements only, so the equivalence records how little the count varies and is weak evidence about any association with the measures.

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

Eleven of the 21 checks held, and one failed. A check holds when its test falls on the same side of 0.05 as the primary test. With the most influential engagement, ENG-027, removed, intent to leave reached r = 0.09 (95% CI 0.02 to 0.16, p < 0.001), so the check failed. Removing one of the two engagements that carry all the variation changes the estimate, and the failure shows how little the primary result rests on.

The same check could not be computed for customer satisfaction, because a single active engagement carries all the variation. The bootstrap of engagements was not computed for any measure, because too few of its resamples were valid, as expected when two engagements carry the variation.

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.

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 count departed from the study's typical pattern on all four measures. The typical indicator is a rating with a strong association with each measure, and the study medians ran from 0.60 to 0.84. The count is not a rating, its associations were near zero, and the departure is not a finding about customer management executives.

Customer management executives were almost absent from the respondents. The count was one on two engagements and zero on the remaining 148. Its associations with cancellation, customer satisfaction, the engagement indicator mean and intent to leave lay near zero and were formally equivalent to zero, but the robustness checks show that a single engagement moves the estimate. The count supports no statement about how the presence of a customer management executive among the respondents relates to the outcome of an engagement.

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. Canceled engagements were described after cancellation.

The count records respondents and not the staffing of the engagement. An engagement may have had a customer management executive who did not respond.

Two engagements carry all the variation in the count. The intervals, equivalence tests and power figures are computed as though the count varied across the engagements, and they overstate the information available.

The segment tests of intent to leave treated respondents as independent. Every respondent on an engagement carries the same count, so those tests overstate the information in the respondent answers.