Source: https://akholi.com/outsourcing-2026-agentic-ai-tier-1-banks/part-1-engagement-outcomes/structural-attributes-and-outcome-associations/
Author: Phil Hatch
Publisher: Akholi

Structural Attributes and Outcome Associations

The 2026 Report · Part 1

# Structural Attributes and Outcome Associations

Financial institutions, relationships, work, commercial terms, vendors, and team size

By Phil Hatch · Akholi · Version 1.8.2

Complete study · PDFPart 1 · PDFStudy overview / Part 1 / Structural Attributes and Outcome Associations

The structural analysis covers institution type, outsourcing model, work characteristics, commercial terms, vendor characteristics, and team size. Ten of 19 attributes met the study's equivalence criteria against all four core measures. Other results included supported associations and inconclusive estimates. This mixed pattern requires attribute-specific interpretation rather than a general conclusion that structural characteristics do not matter.

Each attribute's result against each core measure is stated in one of the four states of a reported result.

Table 1.2. Structural attributes against the four core study measures

| Attribute | Cancellation rate | Customer satisfaction | Engagement indicator mean | Intent to leave |
| --- | --- | --- | --- | --- |
| Type of financial institution | Equivalent to zero | Inconclusive | Equivalent to zero | Equivalent to zero |
| Asset size of the financial institution | No difference | No difference | No difference | No difference |
| Financial institution operates capability centers | Equivalent to zero | Inconclusive | Equivalent to zero | Equivalent to zero |
| Financial institution scaling agentic outsourcing | Equivalent to zero | Equivalent to zero | Equivalent to zero | Equivalent to zero |
| Financial institution adding agentic engagements | Equivalent to zero | Equivalent to zero | Equivalent to zero | Equivalent to zero |
| Financial institution shifting work to its capability centers | Associated | Associated | Associated | Associated |
| Length of the vendor's relationship | Equivalent to zero | Inconclusive | Equivalent to zero | Equivalent to zero |
| Concurrent agentic engagements with the financial institution | Equivalent to zero | Equivalent to zero | Equivalent to zero | Equivalent to zero |
| Outsourcing model | Equivalent to zero | Equivalent to zero | Equivalent to zero | Equivalent to zero |
| Class of work | Equivalent to zero | Equivalent to zero | Equivalent to zero | Equivalent to zero |
| Process family | Inconclusive | Equivalent to zero | Equivalent to zero | Inconclusive |
| Pricing model | Inconclusive | Not tested | Inconclusive | Equivalent to zero |
| Contract term | Equivalent to zero | Equivalent to zero | Equivalent to zero | Equivalent to zero |
| Annual contract value (31 engagements) | Inconclusive | Inconclusive | Inconclusive | Inconclusive |
| Vendor type | Equivalent to zero | Equivalent to zero | Equivalent to zero | Equivalent to zero |
| Vendor size | Equivalent to zero | Equivalent to zero | Equivalent to zero | Equivalent to zero |
| Vendor's agentic engagements for financial institutions | Equivalent to zero | Equivalent to zero | Equivalent to zero | Equivalent to zero |
| Current size of the delivery team | Equivalent to zero | Equivalent to zero | Equivalent to zero | Equivalent to zero |
| Peak size of the delivery team | Associated | Equivalent to zero | Associated, not among active engagements | Associated, not among active engagements |

Two attributes are exceptions, each with a limit. Engagements whose respondents agreed that the financial institution was shifting more work to its own global capability centers recorded worse outcomes on every measure, and the noncancellation associations were retained in the active-engagement analyses. The rating records the respondents' view of the financial institution's plans, and the study holds no evidence on which came first. Engagements that had reached a larger peak team had a higher cancellation rate. Two of the three associations of peak team size did not hold among active engagements, and the data do not show whether teams grew in response to problems.

Many respondents selected “I do not know” for commercial attributes: 228 of 575 for pricing model, 258 of 597 for contract term, and 243 of 597 for relationship length. All 150 engagement managers and 31 portfolio directors answered whether the institution operated capability centers, compared with 168 of 667 individual contributors (25.19%). These differences concern the information respondents supplied and may reflect role-specific access or responsibilities.

> “The playbook that exists is the old staffed-delivery playbook with the word agent inserted. Pricing was still constructed on headcount and hours.” -Portfolio Director, KPO, Corporate and back-office functions

Written accounts described preparation and pricing concerns that differ from the attributes tested statistically. Such accounts can identify possible mechanisms for further investigation, but they do not resolve inconclusive comparisons or demonstrate why outcomes differed.

> “The vendor priced the expense agent's throughput against the old FTE headcount baseline instead of actual hours saved, so the billed 'efficiency gain' stayed flat no matter how many expense reports the agent cleared on its own.” -Individual Contributor, BPO, Corporate and back-office functions

> “Pricing model still assumes FTE hours nobody is spending anymore” -Individual Contributor, CXM, Technology engineering and operations

## 1 The Financial Institution

Institution type, asset size, capability-center presence, and reported expansion plans produced a mixture of equivalent and inconclusive estimates. Reported movement of work to capability centers was associated with weaker outcomes. Each result below is interpreted according to its own measure and sample.

Information availability varied by role. “I do not know” responses exceeded numeric ratings on reported growth in agentic engagements, and 334 of 576 respondents (57.99%) selected that response for institutional asset size. The pattern does not establish whether respondents needed that information to perform their duties.

### 1.1 Institution Type and Outcome Comparisons

Indicator C031 · What type of a bank is the customer? · Full statistical analysis ›

Institution-type comparisons supported equivalence for three core measures, while the active-engagement customer-satisfaction comparison remained inconclusive. The Engagements sheet classifies the financial institution served by each engagement into 10 types, from 24 engagements in asset management to nine in private equity and private capital. The type of financial institution showed no statistically significant association with cancellation (chi-square 5.71, df = 8, p = 0.693, n = 141 engagements). It showed no statistically significant association with the engagement indicator mean (Welch F 1.03, p = 0.425, n = 141 engagements) or with intent to leave (Welch F 0.83, p = 0.583, n = 141 engagements). Each of the three results was within the study’s equivalence bounds. The test of customer satisfaction, on the four types with 10 or more active engagements, did not reach significance and was inconclusive (Welch F 2.39, p = 0.097, n = 46 engagements).

Table 1.3. The four core study measures by type of financial institution, 150 engagements

| Type of financial institution | Engagements | Cancellation rate | Customer satisfaction | Indicator mean | Intent to leave |
| --- | --- | --- | --- | --- | --- |
| Asset management | 24 | 13 of 24 (54.17%) | 37.83 | 3.85 | 7.21 |
| Wealth management and private banking | 23 | 8 of 23 (34.78%) | 44.74 | 4.09 | 6.49 |
| Investment banking (advisory and capital-markets origination) | 19 | 9 of 19 (47.37%) | 34.84 | 3.42 | 7.66 |
| Insurance (life, property and casualty, health) | 14 | 5 of 14 (35.71%) | 47.00 | 4.00 | 6.91 |
| Commercial and business banking | 13 | 4 of 13 (30.77%) | 46.54 | 4.46 | 6.50 |
| Payments and merchant acquiring (networks, acquiring, processing) | 13 | 3 of 13 (23.08%) | 49.77 | 4.20 | 6.34 |
| Credit card and payments issuing | 12 | 5 of 12 (41.67%) | 39.33 | 3.69 | 6.47 |
| Retail and consumer banking | 12 | 4 of 12 (33.33%) | 50.50 | 4.43 | 5.76 |
| Markets and trading (sales and trading, global markets) | 11 | 3 of 11 (27.27%) | 49.91 | 4.38 | 6.71 |
| Private equity and private capital | 9 | 2 of 9 | 49.11 | 4.09 | 7.11 |

Note: Cancellation rate is canceled engagements over engagements of the type. Customer satisfaction, indicator mean, and intent to leave are means of the engagement values across all engagements of the type, canceled and active. Every type rests on fewer than 30 engagements. Private equity and private capital rest on nine engagements, below the base of 10 at which the study prints a rate. Its cancellation is shown as a count.

Investment banking engagements recorded the lowest customer satisfaction among active engagements (mean 56.40 points, n = 10 engagements). They also recorded the lowest engagement indicator mean (3.4, n = 19 engagements) and the highest intent to leave (7.7, n = 19 engagements). None of the three comparisons reached statistical significance.

Six written responses addressed the type of financial institution, and four were negative. Each negative response placed the difficulty in the vendor's preparation for that type of institution, not in the type itself. No written response provided additional comparative performance analysis between financial institution types.

> “The spend on training is real, I will grant that, but the courses are generic and nothing in them touches an asset manager's book.” -Portfolio Director, KPO, Trade, settlement and reconciliation

> “On the estate they are no better or worse than most asset managers we deal with at this stage. Their people answered when we raised things, and the business case never wobbled.” -Portfolio Director, CXM, Corporate and back-office functions

### 1.2 Reported Knowledge of Financial Institution Asset Size

Indicator C062 · What is the customer’s approximate total asset size, in USD? · Full statistical analysis ›

Most respondents answering the question about the financial institution's approximate total asset size did not know it. The question was answered by 576 of the 850 respondents, and 334 of the 576 (57.99%) answered “I do not know.” Those answers came from 141 of the 150 engagements. Among the 242 respondents who gave a listed size, the most frequent answer was $250 billion to under $1 trillion (90 respondents, 37.19%).

The proportion supplying an asset-size category increased with seniority: 146 of 667 individual contributors (21.89%), 64 of 150 engagement managers (42.67%), and 30 of 31 portfolio directors (96.77%). With “I do not know” retained as a response category, the answer showed no statistically significant association with cancellation (chi-square 3.96, df = 4, p = 0.411, n = 115 engagements). It showed no statistically significant association with customer satisfaction (F 1.48, df = 4 and 110, p = 0.215, n = 115 engagements), the engagement indicator mean (F 0.91, df = 4 and 110, p = 0.460, n = 115 engagements), or intent to leave (F 0.93, df = 4 and 110, p = 0.449, n = 115 engagements). The results for cancellation, the engagement indicator mean, and intent to leave were within the study’s equivalence bounds, and the customer-satisfaction result was inconclusive. Only one listed size category held 10 or more engagements, so the listed categories were not compared with one another.

### 1.3 Financial Institution Capability Centers as Reported by the Delivery Team

Indicator C063 · Does the customer operate its own global capability centers? · Full statistical analysis ›

Knowledge of whether the financial institution operated global capability centers varied substantially by respondent role. The question was answered by 596 of the 850 respondents, and 247 of the 596 (41.44%) answered: “I do not know.” The financial institution operated its own capability centers on the account of 190 of the 349 respondents who answered yes or no (54.44%).

Figure 1.16. Share answering Yes to “Does the customer operate its own global capability centers?”, by respondent role and engagement status

Scroll horizontally to view the complete figure.

Knowledge of whether the financial institution operated global capability centers varied substantially by respondent role.

Respondent role Share answering Yes

Lifecycle stage / status Share answering Yes

Exact chart data

| Series | Point index | Category | Value as stored |
| --- | --- | --- | --- |
| Share answering Yes | 0 | Portfolio director (n = 31) | 51.612900000000003 |
| Share answering Yes | 1 | Engagement manager (n = 150) | 51.333300000000001 |
| Share answering Yes | 2 | Individual contributor (n = 168) | 57.738100000000003 |

Exact chart data

| Series | Point index | Category | Value as stored |
| --- | --- | --- | --- |
| Share answering Yes | 0 | Canceled (n = 122) | 54.917999999999999 |
| Share answering Yes | 1 | Production (n = 44) | 50 |
| Share answering Yes | 2 | Transition/Cut-Over (n = 102) | 53.921599999999998 |
| Share answering Yes | 3 | Build-Out (n = 81) | 56.790100000000002 |

Note: True or Yes answers: 190 of 349 answers (54.44%).

n gives the numeric-answer base for each group. These are descriptive comparisons of reported answers.

All 150 engagement managers and 31 portfolio directors gave a yes-or-no answer, compared with 168 of 667 individual contributors (25.19%). Among the 78 engagements assigned a yes or no answer, reported capability-center presence showed no statistically significant association with cancellation (chi-square 0.60, df = 1, p = 0.439, n = 78 engagements). It showed no statistically significant association with the engagement indicator mean (Welch F 0.14, p = 0.711, n = 78 engagements) or with intent to leave (Welch F 0.02, p = 0.892, n = 78 engagements). Each of the three results was within the study’s equivalence bounds. The comparison of customer satisfaction among active engagements did not reach significance and was inconclusive (Welch F 2.15, p = 0.148, n = 53 engagements). Presence and reported transfer of work are separate constructs and should not be conflated.

### 1.4 Shift of Work to Financial Institution Capability Centers and Uniformly Weaker Outcomes

Indicator C068 · The customer is shifting more work to their global capability centers · Full statistical analysis ›

Stronger agreement that the financial institution was shifting work to its own global capability centers was associated with less favorable outcomes on every core measure. The question's scale had neither a favorable nor an unfavorable end. The mean rating was 6.1, and the median 6.0 (n = 341 respondents), and 246 respondents answered: “I do not know.”

Figure 1.17. Shift of Work to Financial Institution Capability Centers and Uniformly Weaker Outcomes: associations with the study measures

Scroll horizontally to view the complete figure.

Shift of work to financial institution capability centers and uniformly weaker outcomes: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = 0.54 [0.42, 0.65] 150 engagements p < 0.001 |
| Customer satisfaction | | r = −0.66 [−0.76, −0.53] 94 active engagements p < 0.001 |
| Indicator mean | | r = −0.76 [−0.82, −0.68] 150 engagements p < 0.001 |
| Intent to leave | | r = 0.63 [0.54, 0.71] 323 respondents p < 0.001 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | 0.41998059669999999 | 1 |
| 95% confidence interval | 1 | 0.64749817200000004 | 1 |
| Reported association | 0 | 0.54365076680000002 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.76450023950000001 | 1 |
| 95% confidence interval | 1 | -0.53422311 | 1 |
| Reported association | 0 | -0.664872503 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.81671833059999999 | 1 |
| 95% confidence interval | 1 | -0.67700894629999997 | 1 |
| Reported association | 0 | -0.75531945450000004 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | 0.54183736069999999 | 1 |
| 95% confidence interval | 1 | 0.70707551999999996 | 1 |
| Reported association | 0 | 0.6301882312 | 1 |

Note: mean rating: 6.1 (n = 341 respondents).

Survey question: “The customer is shifting more work to their global capability centers”.

Group means and numeric-answer bases: Individual contributor (n = 160): 6.53. Engagement manager (n = 150): 5.93. Portfolio director (n = 30): 4.83. Build-Out (n = 73): 5.42. Transition/Cut-Over (n = 101): 5.62. Production (n = 46): 5.13. Canceled (n = 121): 7.31.

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.

Stronger agreement was associated with worse outcomes on every measure, beginning with more cancellations (r = 0.54, 95% CI 0.42 to 0.65, n = 150 engagements). Agreement was associated with lower customer satisfaction among active engagements (r = −0.66, 95% CI −0.76 to −0.53, n = 94 engagements). It was also associated with a lower engagement indicator mean (r = −0.76, 95% CI −0.82 to −0.68, n = 150 engagements). Respondents who agreed more strongly reported higher intent to leave (r = 0.63, cluster bootstrap 95% CI 0.54 to 0.71, n = 323 respondents). The applicable robustness checks retained the associations, including active-only checks for the indicator mean and intent to leave. Cancellation itself is constant within active engagements. Agreement was strongest on canceled engagements and among individual contributors, and weakest among portfolio directors.

The finding has a clear limit. The rating records how respondents perceived the financial institution's plans, not a measured movement of work. The study holds no evidence on which came first. Work may have moved in-house before the engagement weakened, or after it. No written response addressed the indicator.

### 1.5 Reported Scaling of Agentic Outsourcing

Indicator C066 · The customer is scaling their agentic AI outsourcing engagements · Full statistical analysis ›

The estimated relationships between reported scaling of the financial institution's agentic AI outsourcing engagements and the core measures met the study's equivalence criteria. The question's scale had neither a favorable nor an unfavorable end. The mean rating was 5.5, and the median 5.0 (n = 364 respondents), and 256 respondents answered: “I do not know.”

Figure 1.18. Reported Scaling of Agentic Outsourcing: associations with the study measures

Scroll horizontally to view the complete figure.

Reported scaling of agentic outsourcing: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = −0.04 150 engagements p = 0.640 |
| Customer satisfaction | | r = 0.02 94 active engagements p = 0.846 |
| Indicator mean | | r = 0.06 150 engagements p = 0.476 |
| Intent to leave | | r = 0.00 352 respondents p = 0.990 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | -3.8517547499999999E-2 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 2.02952081E-2 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 5.8695412400000001E-2 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 6.1460840000000002E-4 | 1 |

Note: mean rating: 5.5 (n = 364 respondents).

Survey question: “The customer is scaling their agentic AI outsourcing engagements”.

Group means and numeric-answer bases: Individual contributor (n = 183): 5.48. Engagement manager (n = 150): 5.53. Portfolio director (n = 29): 5.55. Build-Out (n = 75): 5.29. Transition/Cut-Over (n = 115): 5.68. Production (n = 46): 5.48. Canceled (n = 128): 5.49.

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.

Estimated associations of reported scaling met the equivalence criterion for all four core measures. The correlation with cancellation was r = −0.04 (p = 0.640, n = 150 engagements). It was r = 0.02 with customer satisfaction among active engagements (p = 0.846, n = 94 engagements) and r = 0.06 with the engagement indicator mean (p = 0.476, n = 150 engagements). At respondent level, it was r = 0.00 with intent to leave (p = 0.990, n = 352 respondents). These estimates concern perceived financial institution plans, not measured changes in outsourcing volume.

### 1.6 Reported Growth in the Number of Engagements Unknown to Most Respondents

Indicator C067 · The customer is increasing the number of agentic AI outsourcing engagements · Full statistical analysis ›

More respondents did not know whether the financial institution was increasing the number of its agentic AI outsourcing engagements than supplied a rating, 345 against 247. The question's scale had neither a favorable nor an unfavorable end. The mean rating was 5.3 and the median 5.0 (n = 247 respondents).

Figure 1.19. Reported Growth in the Number of Engagements Unknown to Most Respondents: associations with the study measures

Scroll horizontally to view the complete figure.

Reported growth in the number of engagements unknown to most respondents: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = 0.04 127 engagements p = 0.635 |
| Customer satisfaction | | r = −0.07 77 active engagements p = 0.563 |
| Indicator mean | | r = −0.08 127 engagements p = 0.392 |
| Intent to leave | | r = 0.05 235 respondents p = 0.442 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 4.2494934700000001E-2 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | -6.6982933999999994E-2 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | -7.6616091299999994E-2 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 4.9468328200000002E-2 | 1 |

Note: mean rating: 5.3 (n = 247 respondents).

Survey question: “The customer is increasing the number of agentic AI outsourcing engagements”.

Group means and numeric-answer bases: Individual contributor (n = 162): 5.36. Engagement manager (n = 52): 5.38. Portfolio director (n = 31): 5.29. Build-Out (n = 59): 5.27. Transition/Cut-Over (n = 63): 5.19. Production (n = 29): 5.66. Canceled (n = 96): 5.40.

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 estimated associations of reported growth met the equivalence criterion for all four core measures. The correlation with cancellation was r = 0.04 (p = 0.635, n = 127 engagements). It was r = −0.07 with customer satisfaction among active engagements (p = 0.563, n = 77 engagements) and r = −0.08 with the engagement indicator mean (p = 0.392, n = 127 engagements). At respondent level, it was r = 0.05 with intent to leave (p = 0.442, n = 235 respondents). The high number of unknown responses limits conclusions about the financial institution's wider program.

> We asked the client's collections team for their hardship scripts in week two and got four different versions from four regions. In fairness that is not carelessness on their part. Collections practice in this bank grew up branch by branch over twenty years and every region has its own regulator conversation behind it, so of course the documentation is uneven. Their change board has been quick with us, and the account data itself is clean, which is more than I usually see at three months. What they do not have is one person who can sit with us for a full day and settle what a compliant arrangement offer looks like. That person exists somewhere. We cannot get her diary.” ENGAGEMENT MANAGER · CXM, Payments, cards and disputes

## 2 The Relationship Between the Two Firms

Relationship length and the number of concurrent agentic engagements describe different aspects of the relationship between the firms. Most estimated associations met the equivalence criteria, but the customer-satisfaction result for relationship length was inconclusive. Conflicting accounts of relationship duration also limit interpretation.

### 2.1 Relationship Length and Outcome Comparisons

Indicator C064 · How many months has your employer been serving this customer across all work, not just this engagement? · Full statistical analysis ›

The length of the vendor's relationship with the financial institution met the study's equivalence criteria for cancellation, engagement indicator mean, and intent to leave, while the active-engagement customer-satisfaction result was inconclusive. The question covered the vendor's relationship across all services and was answered in months. The mean answer was 32.6 months and the median 33.0 months, ranging from 3.0 to 67.0 months (n = 354 respondents), and 243 respondents answered: “I do not know.”

Figure 1.20. Relationship Length and Outcome Comparisons: associations with the study measures

Scroll horizontally to view the complete figure.

Relationship length and outcome comparisons: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = 0.05 150 engagements p = 0.524 |
| Customer satisfaction | | r = 0.20 [−0.00, 0.39] 94 active engagements p = 0.055 |
| Indicator mean | | r = 0.05 150 engagements p = 0.576 |
| Intent to leave | | r = −0.06 340 respondents p = 0.467 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 5.2415125E-2 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -3.8666963999999999E-3 | 1 |
| 95% confidence interval | 1 | 0.38596788580000002 | 1 |
| Reported association | 0 | 0.19890610280000001 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 4.6079230200000001E-2 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | -6.1889490200000001E-2 | 1 |

Note: Mean months the vendor has served the institution: 32.6 (n = 354 respondents).

Group means and numeric-answer bases: Individual contributor (n = 172): 33.38. Engagement manager (n = 150): 31.79. Portfolio director (n = 31): 32.06. Build-Out (n = 69): 32.13. Transition/Cut-Over (n = 101): 28.83. Production (n = 48): 37.35. Canceled (n = 136): 34.07.

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 length of the relationship showed no statistically significant association with cancellation (r = 0.05, p = 0.524, n = 150 engagements). The engagement indicator mean showed no statistically significant association with it either (r = 0.05, p = 0.576, n = 150 engagements). No statistically significant association was detected for intent to leave (r = −0.06, p = 0.467, n = 340 respondents), and each of the three results was within the study’s equivalence bounds. The association with customer satisfaction among active engagements was positive and did not reach significance (r = 0.20, 95% CI 0.00 to 0.39, p = 0.055, n = 94 engagements). Two robustness checks did not hold, and the result is inconclusive.

Five written responses addressed relationship length. Three described relationships lasting 14 years, 15 years, and two decades, exceeding the maximum structured answer of 67 months. The difference remains unresolved and may reflect inconsistent units or interpretations of the relationship. The structured estimates are retained as recorded, and the longer written accounts are not treated as corroborating the same duration measure.

> “Fourteen years of relationship equity gone in one incident” -Portfolio Director, CXM, Client servicing and communications

> “Fifteen years of a dependable client relationship” -Engagement Manager, CXM, Corporate and back-office functions

> “two decades of relationship history to draw on” -Engagement Manager, KPO, Trade, settlement and reconciliation

### 2.2 Concurrent Agentic Engagements With the Same Financial Institution

Indicator C065 · How many agentic AI outsourcing engagements does your employer run concurrently with this customer today? · Full statistical analysis ›

Most respondents answering the question about concurrent agentic AI outsourcing engagements between the vendor and the financial institution did not know the number. More respondents answered “I do not know” than gave a number, 362 against 254. Among those who gave a number, the mean was 17.2 and the median 18.0, from 2.0 to 30.0 (n = 254 respondents on 129 engagements).

Figure 1.21. Concurrent Agentic Engagements With the Same Financial Institution: associations with the study measures

Scroll horizontally to view the complete figure.

Concurrent agentic engagements with the same financial institution: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = 0.10 129 engagements p = 0.267 |
| Customer satisfaction | | r = 0.03 83 active engagements p = 0.804 |
| Indicator mean | | r = 0.03 129 engagements p = 0.716 |
| Intent to leave | | r = 0.04 243 respondents p = 0.691 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 9.8422445400000003E-2 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 2.76036523E-2 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 3.2362699799999999E-2 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 3.9490746299999997E-2 | 1 |

Note: Mean concurrent agentic engagements with the institution: 17.2 (n = 254 respondents).

Group means and numeric-answer bases: Individual contributor (n = 163): 17.61. Engagement manager (n = 58): 16.36. Portfolio director (n = 31): 16.26. Build-Out (n = 56): 15.75. Transition/Cut-Over (n = 78): 17.55. Production (n = 26): 14.00. Canceled (n = 94): 18.65.

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 count of concurrent engagements was within the study’s equivalence bounds against each of the four study measures. It recorded r = 0.10 with cancellation (p = 0.267, n = 129 engagements) and r = 0.03 with customer satisfaction among active engagements (p = 0.804, n = 83 engagements). With the engagement indicator mean it recorded r = 0.03 (p = 0.716, n = 129 engagements), and with intent to leave r = 0.04 (p = 0.691, n = 243 respondents). The scale of the two firms' agentic work together did not distinguish the engagements that held from those that were canceled.

## 3 The Work Outsourced

The sample covered contact-center work, trade settlement, identity management, financial-crime screening, investor communications, and other processes. Outsourcing model and class-of-work comparisons met the study's equivalence criteria. Process-family results were mixed, including inconclusive cancellation and intent-to-leave comparisons. Delivery location is descriptive only.

Written responses also describe why respondents valued particular tasks. Their positive content must be interpreted in light of the question asking what they valued most. It is not an independent comparison of all possible sources of job satisfaction.

### 3.1 Outsourcing Model and Outcome Comparisons

Indicator C032 · What is the outsourcing type? · Full statistical analysis ›

The Engagements sheet classifies each engagement by outsourcing model, with CXM on 42 engagements, ITO on 40, BPO on 37, and KPO on 31. Every respondent's answer matched The study. No statistically significant differences by outsourcing model were detected for the four study measures, and each result was within the study’s equivalence bounds. Cancellation recorded chi-square 0.08 (p = 0.994, n = 150 engagements), and customer satisfaction among active engagements Welch F 0.78 (p = 0.510, n = 94 engagements). The engagement indicator mean recorded Welch F 0.90 (p = 0.444, n = 150 engagements), and intent to leave Welch F 0.47 (p = 0.703, n = 150 engagements).

Table 1.4. The four core study measures by outsourcing model, 150 engagements

| Outsourcing model | Engagements | Cancellation rate | Customer satisfaction | Indicator mean | Intent to leave |
| --- | --- | --- | --- | --- | --- |
| CXM | 42 | 15 of 42 (35.71%) | 44.43 | 4.24 | 6.48 |
| ITO | 40 | 15 of 40 (37.50%) | 44.20 | 4.07 | 7.09 |
| BPO | 37 | 14 of 37 (37.84%) | 45.57 | 3.71 | 6.72 |
| KPO | 31 | 12 of 31 (38.71%) | 40.58 | 4.01 | 6.81 |

Note: Cancellation rate is canceled engagements over the engagements in the row. Customer satisfaction, indicator mean, and intent to leave are means of the engagement values across all engagements in the row, canceled and active.

The models differed in vendor and process composition, while the tested outcome contrasts met the equivalence criteria. One engagement manager on a canceled engagement described a mismatch between the commercial model and staffing arrangements.

> “Training on funds and NAV work is decent, I will give them that. On agents the knowledge is thin. We were sold as an AI delivery and staffed as an ordinary BPO.” -Engagement Manager, BPO, Trade, settlement and reconciliation

### 3.2 Class of Work and Outcome Comparisons

Indicator C033 · Is the outsourcing engagement Corporate Functional, or Line of Business Operational? · Full statistical analysis ›

Table 1.5. The four core study measures by class of work, 150 engagements

| Class of work | Engagements | Cancellation rate | Customer satisfaction | Indicator mean | Intent to leave |
| --- | --- | --- | --- | --- | --- |
| LOB Operational | 76 | 30 of 76 (39.47%) | 42.54 | 3.99 | 6.85 |
| Corporate Functional | 74 | 26 of 74 (35.14%) | 45.20 | 4.05 | 6.69 |

Note: Cancellation rate is canceled engagements over the engagements in the row. Customer satisfaction, indicator mean, and intent to leave are means of the engagement values across all engagements in the row, canceled and active.

The Engagements sheet classifies each engagement as serving a line-of-business operation (76 engagements) or a corporate function (74 engagements). The respondents' answers on the class of work agreed with the study in every case. No statistically significant differences between the two classes were detected on the four study measures, and every result was within the study’s equivalence bounds. The test of cancellation gave chi-square 0.30 (p = 0.583, n = 150 engagements), and the test of customer satisfaction among active engagements gave Welch F 0.05 (p = 0.828, n = 94 engagements). The engagement indicator mean gave Welch F 0.07 (p = 0.795, n = 150 engagements), and intent to leave Welch F 0.17 (p = 0.678, n = 150 engagements).

The class followed the process served. Corporate and back-office and technology engagements were mostly corporate functions, while servicing, trade, and payments engagements were mostly line-of-business operations. Whether an engagement served the business directly or a function behind it showed no statistically significant association with its outcomes.

### 3.3 Study Use Cases

Engagement field · Use Case · Full statistical analysis ›

Each engagement's use case is assigned to one of eight process families, from 47 engagements in corporate and back-office functions to seven in financial crime and client onboarding. The process family showed no statistically significant association with customer satisfaction among active engagements (Welch F 0.16, p = 0.850, n = 63 engagements). It showed no statistically significant association with the engagement indicator mean either (Welch F 1.20, p = 0.325, n = 125 engagements), and both results were within the study’s equivalence bounds. The test of cancellation (chi-square 7.48, df = 4, p = 0.113, n = 125 engagements) did not reach significance and was inconclusive. The test of intent to leave (Welch F 1.77, p = 0.155, n = 125 engagements) was inconclusive as well.

Table 1.6. The four core study measures by process family, 150 engagements

| Process family | Engagements | Cancellation rate | Customer satisfaction | Indicator mean | Intent to leave |
| --- | --- | --- | --- | --- | --- |
| Corporate and back-office functions | 47 | 19 of 47 (40.43%) | 41.68 | 3.92 | 6.90 |
| Technology engineering and operations | 35 | 10 of 35 (28.57%) | 49.91 | 4.33 | 6.55 |
| Client servicing and communications | 16 | 9 of 16 (56.25%) | 34.25 | 3.75 | 7.40 |
| Trade, settlement and reconciliation | 15 | 8 of 15 (53.33%) | 35.53 | 3.56 | 7.54 |
| Payments, cards and disputes | 12 | 2 of 12 (16.67%) | 56.08 | 4.36 | 5.25 |
| Credit, underwriting and suitability | 9 | 1 of 9 | 57.78 | 4.73 | 6.15 |
| Regulatory, audit, tax and legal | 9 | 4 of 9 | 37.33 | 3.81 | 7.18 |
| Financial crime and client onboarding | 7 | 3 of 7 | 37.43 | 3.42 | 6.78 |

Note: Cancellation rate is canceled engagements over the engagements in the row. Customer satisfaction, indicator mean, and intent to leave are means of the engagement values across all engagements in the row, canceled and active. Credit, underwriting and suitability, Regulatory, audit, tax and legal and Financial crime and client onboarding rest on fewer than 10 engagements, the base below which the study prints no rate. Cancellation is shown as a count.

The written responses that addressed the use case numbered 33, and 27 of the 33 responses were positive (81.82%), among the most favorable written evidence in the study. Most answered the question on what the respondent loved most about the job and named the use case or the problem it addressed. Respondents on canceled engagements valued the problem they had been asked to solve as well.

> “The fraud detection problem itself was worth solving” -Engagement Manager, ITO, Financial crime and client onboarding

> “Solving identity sprawl at this scale is real work worth doing” -Individual Contributor, ITO, Technology engineering and operations

> “Genuinely hard problem, keeps me sharp” -Individual Contributor, ITO, Technology engineering and operations

### 3.4 Delivery Location

Delivery was located in India on 92 of the 150 engagements and in the United States on 19, with the remaining 39 delivered from other locations. Delivery location showed no statistically significant association with cancellation (chi-square 0.47, df = 2, p = 0.790, n = 150 engagements). It showed no statistically significant association with customer satisfaction (F 0.34, df = 2 and 147, p = 0.711, n = 150 engagements), the engagement indicator mean (F 0.46, df = 2 and 147, p = 0.630, n = 150 engagements), or intent to leave (F 0.40, df = 2 and 147, p = 0.670, n = 150 engagements). Each of the four results was within the study’s equivalence bounds, so outcomes did not differ by delivery location.

## 4 Commercial Terms

No statistically significant association was detected in the reported tests of pricing model, contract term, or annual value, but several estimates were inconclusive and customer satisfaction could not be tested for pricing model. Written pricing accounts address perceived incentives rather than providing an alternative statistical test of those associations.

### 4.1 Pricing on Human Effort for Work Performed by Agents

Indicator C044 · What is the pricing model for this engagement? · Full statistical analysis ›

Of 575 respondents answering the pricing question, 228 selected “I do not know.” Among the 347 naming a model, 140 reported full-time-equivalent staffing as the pricing basis (40.35%). Tested differences did not reach statistical significance for cancellation (chi-square 2.97, simulated p = 0.413, n = 71 engagements), engagement indicator mean (Welch F 1.31, p = 0.290, n = 71), or intent to leave (Welch F 0.79, p = 0.511, n = 71). Only the last met the equivalence criterion. Customer satisfaction was not tested.

Table 1.7. The four core study measures by pricing model, 107 engagements

| Pricing model | Engagements | Cancellation rate | Customer satisfaction | Indicator mean | Intent to leave |
| --- | --- | --- | --- | --- | --- |
| FTE-based | 35 | 11 of 35 (31.43%) | 47.80 | 4.07 | 6.50 |
| Fixed price | 13 | 5 of 13 (38.46%) | 44.00 | 4.20 | 6.63 |
| Hybrid | 13 | 4 of 13 (30.77%) | 43.77 | 3.75 | 6.59 |
| Consumption based | 10 | 6 of 10 (60.00%) | 29.00 | 3.38 | 7.71 |
| Transaction or outcome based | 8 | 6 of 8 | 25.00 | 3.79 | 8.44 |
| I do not know | 28 | 12 of 28 (42.86%) | 41.68 | 3.85 | 6.82 |

Note: Cancellation rate is canceled engagements over the engagements in the row. Customer satisfaction, indicator mean, and intent to leave are means of the engagement values across all engagements in the row, canceled and active. Each engagement carries the pricing answer assigned to it from its respondents’ answers. Transaction or outcome based rests on fewer than 10 engagements, the base below which the study prints no rate. Cancellation is shown as a count.

All 33 written pricing responses were negative. Twenty-one described pricing or billing based on headcount, seats, or hours while agents performed some tasks. At least three respondents described a perceived incentive conflict in which automation reduced billable human effort without providing a corresponding commercial benefit to the vendor. These accounts concern particular arrangements. They do not establish a general effect of staffing-based pricing.

> “As long as this is billed on heads, every hour an agent saves is an hour I am asked to explain, and on that basis there is no incentive anywhere in my chain to let the thing work properly.” -Engagement Manager, BPO, Trade, settlement and reconciliation

> “We are still sold and staffed as a headcount deal, forty seats on a monitoring queue, while the deck that won the work promises the agents do the first pass. So when the agents do less than that, the only lever I have is more analysts, which I cannot get, and when they do more, we are billing for seats nobody needs.” -Engagement Manager, BPO, Financial crime and client onboarding

> “The vendor billed the engagement on a per-FTE-equivalent basis calculated from pre-agent staffing levels, so the invoice looked justified on paper even though the agent had already absorbed most of the manual expense-processing work months earlier.” -Individual Contributor, BPO, Corporate and back-office functions

> “Merchant onboarding volume went up nearly double after client launched new acquiring product, and our agents absorbed it without adding headcount. Good news, except that our invoice to client is still built on seats. So we did more work and billed less, and i had to explain to my own director why revenue on account went flat in a month where we handled forty percent more cases.” -Engagement Manager, CXM, Payments, cards and disputes

> “We must stop pricing this as if it were bodies on a floor.” -Portfolio Director, ITO, Technology engineering and operations

### 4.2 Contract Term and Outcome Associations

Indicator C045 · What is the contract term of this engagement, in months? · Full statistical analysis ›

Reported contract terms averaged 30.9 months, with a median of 36 months, or three years, and a range of 12–60 months (n = 339 respondents on 150 engagements). Another 258 respondents selected “I do not know.”

Figure 1.22. Contract Term and Outcome Associations: associations with the study measures

Scroll horizontally to view the complete figure.

Contract term and outcome associations: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = −0.10 150 engagements p = 0.223 |
| Customer satisfaction | | r = 0.03 94 active engagements p = 0.744 |
| Indicator mean | | r = 0.08 150 engagements p = 0.302 |
| Intent to leave | | r = −0.02 327 respondents p = 0.781 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | -0.1000561789 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 3.4158749000000002E-2 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 8.4837898699999997E-2 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | -2.3098861799999999E-2 | 1 |

Note: Mean contract term in months: 30.9 (n = 339 respondents).

Group means and numeric-answer bases: Individual contributor (n = 156): 31.38. Engagement manager (n = 150): 30.48. Portfolio director (n = 31): 30.58. Build-Out (n = 73): 32.55. Transition/Cut-Over (n = 95): 32.08. Production (n = 43): 29.30. Canceled (n = 128): 29.62.

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.

Each association of the contract term with the four study measures was within the study’s equivalence bounds. Its correlation with cancellation was r = −0.10 (p = 0.223, n = 150 engagements), and with customer satisfaction among active engagements r = 0.03 (p = 0.744, n = 94 engagements). With the engagement indicator mean it was r = 0.08 (p = 0.302, n = 150 engagements), and with intent to leave r = −0.02 (p = 0.781, n = 327 respondents). A longer contract term showed no statistically significant association with less cancellation.

### 4.3 Contract Value Unknown to Most of the Delivery Team

Indicator C046 · What is the annual contract value of this engagement, in USD? · Full statistical analysis ›

Most respondents answering the question about annual engagement contract value did not know it, and knowledge increased with seniority. The question was answered by 601 of the 850 respondents, and 353 of the 601 (58.74%) answered “I do not know.” Among the 248 who gave a listed value, 106 (42.74%) answered $1 million to under $5 million and 102 (41.13%) $5 million to under $10 million. The share who gave a value was greater in each more senior role, from 152 of the 667 individual contributors (22.79%) to 30 of the 31 portfolio directors (96.77%).

At engagement level, no statistically significant association was detected for contract value, and most results were inconclusive on a base of 31 engagements. Respondent-level comparisons retained “I do not know” as a category and differed for cancellation (F = 3.74, p = 0.006, n = 601) and the respondent indicator mean (F = 3.06, p = 0.018, n = 601). Respondents selecting the largest contract-value category were most often on canceled engagements, but both extreme categories contained fewer than 30 respondents. The results require caution and do not establish a relationship between contract value and performance.

> What would actually help is having one person on the team from day one who has run an internal service desk. We have people who know the AI side well, that part is fine. What we don't have is anyone who can tell a desk owner what a realistic deflection number looks like, so I end up guessing. The other piece is pricing. We are still selling this shaped like a staffing deal, and it gets awkward when the agent takes the volume and the invoice looks the same.” ENGAGEMENT MANAGER · CXM, Client servicing and communications

## 5 The Vendor

Three survey questions described the firm that delivered each engagement. They asked for the type of company, the number of its employees worldwide, and the number of agentic AI engagements it runs for financial institutions. No statistically significant cancellation differences were detected for these attributes. Each was tested against cancellation, customer satisfaction among active engagements, the engagement indicator mean, and intent to leave, and all 12 results were within the study’s equivalence bounds. The results do not establish identical outcomes across vendor types. They showed no statistically significant association with the size of the vendor or with the breadth of its agentic work for financial institutions.

For vendor size, 260 of 798 respondents answering the question selected “I do not know” (32.58%). For the vendor's number of financial-institution agentic engagements, 328 of all 850 respondents supplied a number (38.59%). These figures describe response coverage, not verified ignorance of the vendor's business.

### 5.1 Vendor Type and Outcome Comparisons

Indicator C076 · What type of company is your employer? · Full statistical analysis ›

IT services majors accounted for the largest reported share of vendors, followed by BPO specialists. An IT services major was named in 316 of the 793 answers (39.85%), and a BPO specialist in 198 (24.97%). A consulting firm was named in 117 of the 793 answers (14.75%), a boutique or specialist provider in 87 (10.97%), and a pure-play agentic AI firm in 75 (9.46%). The respondents on each engagement gave the same answer, and the question classifies the 150 engagements as well. IT services majors delivered 59 of the 150 engagements (39.33%) and pure-play agentic AI firms 14 of the 150 (9.33%).

Table 1.8. The four core study measures by vendor type, 150 engagements

| Vendor type | Engagements | Cancellation rate | Customer satisfaction | Indicator mean | Intent to leave |
| --- | --- | --- | --- | --- | --- |
| IT services major | 59 | 19 of 59 (32.20%) | 46.92 | 4.07 | 6.62 |
| BPO specialist | 37 | 17 of 37 (45.95%) | 37.22 | 3.68 | 7.11 |
| Consulting firm | 23 | 8 of 23 (34.78%) | 44.04 | 4.31 | 6.59 |
| Boutique or specialist provider | 17 | 7 of 17 (41.18%) | 44.41 | 4.04 | 6.99 |
| Pure-play agentic AI firm | 14 | 5 of 14 (35.71%) | 47.50 | 4.17 | 6.57 |

Note: Cancellation rate is canceled engagements over the engagements in the row. Customer satisfaction, indicator mean, and intent to leave are means of the engagement values across all engagements in the row, canceled and active. Each engagement carries the vendor type assigned to it from its respondents’ answers.

The engagement-level comparisons did not detect statistically significant differences by vendor type, and each met the equivalence criterion: cancellation chi-square 2.02 (p = 0.731, n = 150 engagements), active-engagement customer satisfaction Welch F 0.76 (p = 0.524, n = 85), engagement indicator mean Welch F 0.74 (p = 0.573, n = 150), and engagement-level intent to leave Welch F 0.31 (p = 0.872, n = 150). These tests differ from respondent-level vendor-type comparisons used in individual indicator analyses. Vendor types differed in their outsourcing models and process families without a detected outcome difference under the reported tests.

### 5.2 Vendor Size and Outcome Associations

Indicator C077 · Approximately how many employees does your employer have worldwide? · Full statistical analysis ›

Respondents described vendors with widely differing worldwide headcounts, but their estimates were not independently verified. The mean was approximately 110,955 employees, the median 90,854, and the range 321–340,114 (n = 538 respondents on 150 engagements). Another 260 respondents selected “I do not know,” including at least one respondent on 127 of 150 engagements (84.67%). The estimates were not independently verified against employer records.

Figure 1.23. Vendor Size and Outcome Associations: associations with the study measures

Scroll horizontally to view the complete figure.

Vendor size and outcome associations: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = −0.06 [−0.22, 0.10] 150 engagements p = 0.471 |
| Customer satisfaction | | r = −0.07 [−0.27, 0.13] 94 active engagements p = 0.500 |
| Indicator mean | | r = −0.01 [−0.17, 0.15] 150 engagements p = 0.901 |
| Intent to leave | | r = −0.01 [−0.17, 0.15] 508 respondents p = 0.901 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.21755596129999999 | 1 |
| 95% confidence interval | 1 | 0.1018664442 | 1 |
| Reported association | 0 | -5.9364219000000003E-2 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.26919367199999999 | 1 |
| 95% confidence interval | 1 | 0.13411295500000001 | 1 |
| Reported association | 0 | -7.0417503800000003E-2 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.17018756839999999 | 1 |
| 95% confidence interval | 1 | 0.1503029655 | 1 |
| Reported association | 0 | -1.02043612E-2 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.16888130160000001 | 1 |
| 95% confidence interval | 1 | 0.1492568222 | 1 |
| Reported association | 0 | -1.02986021E-2 | 1 |

Note: Mean vendor employees worldwide: 110,955 (n = 538 respondents).

Group means and numeric-answer bases: Individual contributor (n = 356): 111,081.72. Engagement manager (n = 150): 108,780.67. Portfolio director (n = 31): 122,090.48. Build-Out (n = 121): 105,430.31. Transition/Cut-Over (n = 152): 98,560.89. Production (n = 70): 161,431.73. Canceled (n = 195): 105,924.09.

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.

Vendor size recorded results within the study’s equivalence bounds on all four study measures. Against cancellation, it recorded r = −0.06 (95% CI −0.22 to 0.10, p = 0.471, n = 150 engagements). Against customer satisfaction among active engagements, it recorded r = −0.07 (95% CI −0.27 to 0.13, p = 0.500, n = 94 engagements). Against the engagement indicator mean, it recorded r = −0.01 (95% CI −0.17 to 0.15, p = 0.901, n = 150 engagements). Against intent to leave it recorded r = −0.01 (cluster bootstrap 95% CI −0.17 to 0.15, p = 0.901, n = 508 respondents). The typed sizes differed markedly by vendor type, from the IT services majors to the pure-play agentic AI firms, consistent with the vendor types respondents named.

### 5.3 Breadth of Vendor Agentic Work for Financial Institutions

Indicator C079 · How many agentic AI engagements does your employer run for banks today? · Full statistical analysis ›

Fewer than half the respondents supplied a number for the vendor's agentic AI engagements serving financial institutions. A number was typed by 328 of the 850 respondents (38.59%), 242 (28.47%) answered “I do not know,” and 280 (32.94%) left the question blank. Among those who gave a number, the mean was 24.6 engagements and the median 25.0, from 5.0 to 40.0 (n = 328 respondents on 150 engagements).

Figure 1.24. Breadth of Vendor Agentic Work for Financial Institutions: associations with the study measures

Scroll horizontally to view the complete figure.

Breadth of vendor agentic work for financial institutions: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = 0.02 [−0.14, 0.18] 150 engagements p = 0.833 |
| Customer satisfaction | | r = 0.09 [−0.11, 0.29] 94 active engagements p = 0.390 |
| Indicator mean | | r = 0.03 [−0.13, 0.19] 150 engagements p = 0.743 |
| Intent to leave | | r = −0.05 [−0.21, 0.10] 312 respondents p = 0.501 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.1433114824 | 1 |
| 95% confidence interval | 1 | 0.17711757910000001 | 1 |
| Reported association | 0 | 1.7348492100000001E-2 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.1149742683 | 1 |
| 95% confidence interval | 1 | 0.28712962720000001 | 1 |
| Reported association | 0 | 8.9733094799999996E-2 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.13382830640000001 | 1 |
| 95% confidence interval | 1 | 0.18646686979999999 | 1 |
| Reported association | 0 | 2.7012572799999999E-2 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.20605104420000001 | 1 |
| 95% confidence interval | 1 | 9.8557204499999995E-2 | 1 |
| Reported association | 0 | -5.1997802000000003E-2 | 1 |

Note: Mean agentic AI engagements the vendor runs for financial institutions: 24.6 (n = 328 respondents).

Group means and numeric-answer bases: Individual contributor (n = 145): 24.17. Engagement manager (n = 150): 24.39. Portfolio director (n = 31): 27.65. Build-Out (n = 72): 24.43. Transition/Cut-Over (n = 98): 25.20. Production (n = 40): 24.00. Canceled (n = 118): 24.42.

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 breadth of the vendor's agentic work showed no statistically significant association with the four study measures, each result within the study’s equivalence bounds. The correlation with cancellation was r = 0.02 (95% CI −0.14 to 0.18, p = 0.833, n = 150 engagements). With customer satisfaction among active engagements, it was r = 0.09 (95% CI −0.11 to 0.29, p = 0.390, n = 94 engagements). With the engagement indicator mean, it was r = 0.03 (95% CI −0.13 to 0.19, p = 0.743, n = 150 engagements). With intent to leave, it was r = −0.05 (cluster bootstrap 95% CI −0.21 to 0.10, p = 0.501, n = 312 respondents).

The count records the breadth of a vendor's agentic work for financial institutions. It does not record whether the vendor had delivered the same service before. An engagement manager described a vendor that had sold a service it had never delivered.

> “…this is not about the engineers... they were good, they were simply sent in without a pipeline of our own to deploy from. We sold release automation to a global bank and we do not have a working reference of it anywhere.” -Engagement Manager, ITO, Technology engineering and operations

## 6 Size of the Delivery Team

The survey distinguished current, peak, and planned team size. Current headcount had estimated associations within the equivalence bounds for all four measures. Peak headcount was higher on canceled engagements. Planned headcount was associated with engagement status and, among active engagements, with customer satisfaction. These measures do not establish whether staffing changes preceded or followed performance changes.

Written accounts addressed staffing sufficiency rather than headcount alone. That distinction is retained below. Part 4 examines its relationship to workforce experience.

### 6.1 Current Team Size and Outcomes

Indicator C040 · How many people are on the vendor team for this engagement today? · Full statistical analysis ›

Reported current delivery-team headcounts agreed with the engagement-level data on every engagement. The mean was 8.4 people and the median 8.0, from 3.0 to 16.0 (n = 150 engagements). Respondents were also asked how many people were on the delivery team today, and every answer equaled the headcount recorded for the engagement (n = 789 respondents on 150 engagements).

Figure 1.25. Current Team Size and Outcomes: associations with the study measures

Scroll horizontally to view the complete figure.

Current team size and outcomes: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = 0.12 [−0.05, 0.27] 150 engagements p = 0.159 |
| Customer satisfaction | | r = 0.02 [−0.18, 0.22] 94 active engagements p = 0.851 |
| Indicator mean | | r = −0.10 [−0.26, 0.06] 150 engagements p = 0.226 |
| Intent to leave | | r = 0.06 742 respondents p = 0.442 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -4.5439933699999997E-2 | 1 |
| 95% confidence interval | 1 | 0.27090381400000002 | 1 |
| Reported association | 0 | 0.1156638886 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.1836853279 | 1 |
| 95% confidence interval | 1 | 0.22139834680000001 | 1 |
| Reported association | 0 | 1.9663466899999999E-2 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.25563332579999998 | 1 |
| 95% confidence interval | 1 | 6.1800544899999997E-2 | 1 |
| Reported association | 0 | -9.9445918600000002E-2 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 6.3266528099999997E-2 | 1 |

Note: Mean delivery team headcount today: 8.8 (n = 789 respondents).

Group means and numeric-answer bases: Individual contributor (n = 617): 8.93. Engagement manager (n = 150): 8.35. Portfolio director (n = 22): 8.86. Build-Out (n = 157): 8.49. Transition/Cut-Over (n = 232): 9.11. Production (n = 96): 7.71. Canceled (n = 304): 9.12.

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.

Against the four study measures, every association of the headcount was within the study’s equivalence bounds. Its correlation with cancellation was r = 0.12 (95% CI −0.05 to 0.27, p = 0.159, n = 150 engagements). With customer satisfaction among active engagements, it was r = 0.02 (95% CI −0.18 to 0.22, p = 0.851, n = 94 engagements). With the engagement indicator mean, it was r = −0.10 (95% CI −0.26 to 0.06, p = 0.226, n = 150 engagements). With intent to leave, it was r = 0.09 (cluster bootstrap 95% CI −0.07 to 0.24, p = 0.289, n = 798 respondents).

The headcount followed the type of engagement. After Holm correction, it differed by process family (Welch F 4.28, Holm p = 0.022, n = 150 engagements) and by class of work (Welch F 12.32, Holm p = 0.003, n = 150 engagements). Corporate and back-office engagements and corporate functional engagements had the smaller teams.

The written responses addressed whether the team was large enough for its tasks. There were 33 written responses on the size of the team, from 33 respondents on 29 engagements, and 23 of the 33 (69.70%) were negative. The most common statement, made in 17 of the 33, was that the team was too small for its tasks. Six described members who had left, had been moved, or were looking to leave.

> “If you only read one line of this: we don't have the people. I was given three analysts for a diligence workstream that needs six, and two of them had never seen a data room before this.” -Engagement Manager, KPO, Corporate and back-office functions

> “…my team is only four people and two already looking for other job.” -Individual Contributor, CXM, Corporate and back-office functions

> “My lead is asking for more people every week, nobody give him...” -Individual Contributor, KPO, Corporate and back-office functions

> “Talent shortage means we're always short-staffed” -Individual Contributor, BPO, Corporate and back-office functions

The voluntary accounts concern perceived sufficiency for assigned tasks. A headcount measure does not directly test that judgment, so its null associations do not establish that staffing was adequate.

### 6.2 Team Size at Peak

Indicator C041 · What was the peak vendor team headcount on this engagement to date? · Full statistical analysis ›

Mean peak delivery-team headcount exceeded mean current headcount at engagement level. The mean answer was 10.5 people, and the median 10.0, from 3.0 to 18.0 (n = 787 respondents on 150 engagements), and no respondent answered: “I do not know.” At engagement level, the mean peak team was 10.0 people against a mean current headcount of 8.4 (n = 150 engagements).

Figure 1.26. Team Size at Peak: associations with the study measures

Scroll horizontally to view the complete figure.

Team size at peak: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = 0.24 [0.09, 0.39] 150 engagements p = 0.003 |
| Customer satisfaction | | r = −0.10 [−0.30, 0.10] 94 active engagements p = 0.322 |
| Indicator mean | | r = −0.25 [−0.39, −0.09] 150 engagements p = 0.002 |
| Intent to leave | | r = 0.23 [0.09, 0.37] 739 respondents p = 0.002 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | 8.7825803899999999E-2 | 1 |
| 95% confidence interval | 1 | 0.38962926980000001 | 1 |
| Reported association | 0 | 0.24464394280000001 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.29963091069999997 | 1 |
| 95% confidence interval | 1 | 0.1014557744 | 1 |
| Reported association | 0 | -0.1032843719 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | -0.39199207279999998 | 1 |
| 95% confidence interval | 1 | -9.0592346099999999E-2 | 1 |
| Reported association | 0 | -0.24726397829999999 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | 8.5160238999999999E-2 | 1 |
| 95% confidence interval | 1 | 0.36921877949999998 | 1 |
| Reported association | 0 | 0.2315374808 | 1 |

Note: Mean peak delivery team headcount: 10.5 (n = 787 respondents).

Group means and numeric-answer bases: Individual contributor (n = 611): 10.66. Engagement manager (n = 150): 10.01. Portfolio director (n = 25): 10.88. Build-Out (n = 167): 10.02. Transition/Cut-Over (n = 230): 10.39. Production (n = 91): 8.81. Canceled (n = 299): 11.46.

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.

Engagements with a larger peak team had a higher cancellation rate (r = 0.24, 95% CI 0.09 to 0.39, p = 0.003, n = 150 engagements). They also had a lower engagement indicator mean (r = −0.25, 95% CI −0.39 to −0.09, p = 0.002, n = 150 engagements). Respondents on those engagements reported higher intent to leave (r = 0.23, cluster bootstrap 95% CI 0.09 to 0.37, p = 0.002 with standard errors clustered by engagement, n = 739 respondents). The peak team showed no statistically significant association with customer satisfaction among active engagements, and that association was within the study’s equivalence bounds (r = −0.10, 95% CI −0.30 to 0.10, p = 0.322, n = 94 engagements). Canceled engagements reported the largest peak teams.

Two of the three associations did not hold among active engagements alone. Within the 94 active engagements, the association with the engagement indicator mean was r = −0.15 (p = 0.136, n = 94 engagements). Among respondents on those engagements, the association with intent to leave was r = 0.09 (p = 0.358). Those two associations reflect the difference between canceled and active engagements more than a gradient within active work.

The data record the largest size the team reached. They do not record the course of the team before or after that point, or whether it grew in response to problems on the engagement. One individual contributor described a team reduced in a single day, partway through a task.

> “in March they moved two people off my team to another account in one day, no handover. And we were mid way through tuning the screening rules. My lead found out from the rota, not from anyone above him. After that backlog went to nine days and the same managers sat in review asking why disposition quality dropped... nobody joined those two things.” -Individual Contributor, ITO, Financial crime and client onboarding

### 6.3 Planned Team Size as a Marker of Continuation

Indicator C042 · Per the current plan, how many people will be on the vendor team 12 months from now? · Full statistical analysis ›

Knowledge of the planned delivery-team headcount 12 months ahead was limited across the engagements. A number was given by 339 of the 850 respondents (39.88%). The question was answered by 589 respondents, and 250 of the 589 (42.44%) answered: “I do not know.” On 131 of the 150 engagements (87.33%), at least one respondent did not know the plan for the team. Among the respondents who gave a number, the mean was 5.6 people and the median 6.0 (n = 339 respondents).

Figure 1.27. Planned Team Size as a Marker of Continuation: associations with the study measures

Scroll horizontally to view the complete figure.

Planned team size as a marker of continuation: associations with the study measures

Scroll horizontally to view the complete comparison.

| Comparison | Correlation r | Estimate and analysis base |
| --- | --- | --- |
| Cancellation | | r = −0.86 150 engagements p < 0.001 |
| Customer satisfaction | | r = 0.47 [0.30, 0.62] 94 active engagements p < 0.001 |
| Indicator mean | | r = 0.73 150 engagements p < 0.001 |
| Intent to leave | | r = −0.78 323 respondents p < 0.001 |
| | −1 0 +1 | |

Exact comparison data

Cancellation

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | -0.85541198419999998 | 1 |

Customer satisfaction

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| 95% confidence interval | 0 | 0.29993939250000001 | 1 |
| 95% confidence interval | 1 | 0.61714036189999999 | 1 |
| Reported association | 0 | 0.47376436560000001 | 1 |

Indicator mean

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | 0.73491451340000002 | 1 |

Intent to leave

Exact chart data

| Series | Point index | X | Y |
| --- | --- | --- | --- |
| Reported association | 0 | -0.77922184409999995 | 1 |

Note: Mean planned delivery team headcount in 12 months: 5.6 (n = 339 respondents).

Group means and numeric-answer bases: Individual contributor (n = 157): 5.46. Engagement manager (n = 150): 5.71. Portfolio director (n = 31): 6.29. Build-Out (n = 63): 9.70. Transition/Cut-Over (n = 100): 9.50. Production (n = 41): 8.46. Canceled (n = 135): 0.00.

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.

All 135 numeric answers from respondents on canceled engagements reported a planned headcount of zero. The resulting cancellation association (r = −0.86, p < 0.001, n = 150 engagements) largely reflects reported engagement status. The full-sample associations with engagement indicator mean (r = 0.73, n = 150) and intent to leave (r = −0.78, n = 323 respondents) require the same qualification. They do not establish that staffing plans caused continuation or cancellation.

Among active engagements, where the plan is a forecast, the 204 respondents gave a number typed from 2.0 to 19.0 people. A larger planned team was associated with higher customer satisfaction (r = 0.47, 95% CI 0.30 to 0.62, p < 0.001, n = 94 engagements). It was associated with a higher engagement indicator mean (r = 0.35, 95% CI 0.16 to 0.52, p < 0.001, n = 94 engagements). Among respondents on the active engagements, it was associated with lower intent to leave (r = −0.35, 95% CI −0.47 to −0.22, p < 0.001). The association is descriptive. The data do not show whether the plan for the team followed the state of the engagement or anticipated it.

The structural results provide limited support for attributing outcomes to engagement type, vendor scale, or current headcount. Several estimates remain inconclusive, and associations involving peak and planned staffing have unresolved temporal direction. Part 2 examines the reported delivery conditions that covaried with outcomes, without treating those associations as causal explanations.

> ai made this job easier for me... i came in a year ago and i was scared i would be answering the same laptop and leave balance questions all day. Now i get the odd ones, the ones where somebody is upset about their shift or a policy nobody understands, and i actually have time to read the case properly. My team is smaller than it would have been” INDIVIDUAL CONTRIBUTOR · CXM, Corporate and back-office functions

← The Engagement Indicator Mean and the Interdependence of the Four Measures