Rate the depth of your employer's knowledge of agentic AI

The mean rating was 3.6 and the median 3.0 (n = 807 respondents). Ratings of 3 or below, the distressed range, made up 51.55% of the answers. The rating was associated with all four study measures. Higher ratings came with less cancellation (r = −0.57, 95% CI −0.67 to −0.45, p < 0.001, n = 150 engagements), higher customer satisfaction among active engagements (r = 0.68, 95% CI 0.56 to 0.78, p < 0.001, n = 94 engagements), a higher engagement indicator mean computed without the indicator (r = 0.86, 95% CI 0.81 to 0.90, p < 0.001, n = 150 engagements) and lower intent to leave (r = −0.59, cluster bootstrap 95% CI −0.66 to −0.51, p < 0.001 with standard errors clustered by engagement, n = 758 respondents). After Holm correction, ratings differed by engagement status and lifecycle stage (Welch F 27.21, Holm p < 0.001, omega squared 0.314) and respondent role (Welch F 9.55, Holm p = 0.001, omega squared 0.020).

Few written responses addressed the indicator: 26, all from C135, the comment field on the employer's capability. Eighteen were negative and eight positive. The negative responses placed the shortfall in agentic knowledge above the delivery team, with teams learning the agent side on live work after brief training. The positive responses held that the team knew the technology and placed the gap elsewhere. Written sentiment was not associated with the study measures, though the base is small.

Ratings of 3 or below, the distressed range, accounted for 416 of the 807 answers (51.55%). The engagement mean fell in the distressed range in 63 of 150 engagements.

The primary responses numbered 26, all written in answer to C135, “Add any comments on your employer that the ratings in this section do not capture.” No response from any other question addressed the indicator, so every response was primary.

Ten responses said the people above them, or the firm's executives, did not know agentic AI to the depth the work needed. A new graduate wrote of knowing more about AI than anyone on the executive team. Another respondent asked that the leads be trained first, because they were asking a two-year employee how the tool worked. A third described executives who spoke in the language of the firm's marketing.

Seven responses described learning the agent side on live work: on shift, on live tickets or from free material found online when no one could be asked. When a batch of alerts went wrong late at night, no one in one respondent's firm could explain what the agent had done. Five responses described agent training of two days of slides, a few videos or a recorded webinar that did not match the tool the respondent used.

The eight positive responses held that the team or the firm knew the technology. One said the knowledge sat with perhaps a dozen individuals. Each placed the gap elsewhere: in leadership that had not run agents live, in internal processes run on headcount, in domain knowledge, in a platform built for other work, or in the written route from prototype to governed production.

Of the 26 responses, 18 (69.23%, 95% CI 51.85% to 84.62%) were negative and 8 (30.77%) carried the strongest negative score of −10. Positive responses numbered 8 (30.77%), and no response was neutral. The mean written score was −3.5 (95% CI −5.6 to −1.3, n = 26 responses), and the median was −5.0.

Of the 26 responses, 26 came from respondents who also rated the indicator. Among them, the written score rose with the rating (Spearman rho 0.66, 95% CI 0.37 to 0.83, p < 0.001, n = 26 responses). The written responses corroborated the rating. Of the 18 negative responses with a rating, 13 (72.22%) came from respondents whose rating was in the distressed range.

Respondents who wrote about the indicator and gave a rating numbered 26, with a mean rating of 3.7. The 781 respondents who rated the indicator without writing about it had a mean of 3.6. The difference, 0.2 (95% CI −0.7 to 1.1), included zero. Ratings in the distressed range made up 53.85% of the ratings from respondents who wrote (14 of 26) and 51.47% of the ratings from respondents who did not (402 of 781).

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

Segment families recorded at engagement level were tested on engagement means. Respondent role and vendor type were tested on respondents. Where an expected count fell below 5, the test of the distressed share used Fisher's exact test for a two-by-two table, marked exact, or a Monte Carlo test, marked simulated. After Holm correction, ratings differed by engagement status and lifecycle stage (Welch F 27.21, Holm p < 0.001, omega squared 0.314). Active engagements in Production rated highest (mean 4.6, n = 100 respondents) and canceled engagements lowest (mean 2.4, n = 308 respondents). Ratings also differed by respondent role (Welch F 9.55, Holm p = 0.001, omega squared 0.020). Portfolio directors rated highest (mean 4.7, n = 30 respondents) and individual contributors lowest (mean 3.4, n = 625 respondents). Outsourcing model, institution type, process family, class of work and vendor type showed no difference.

Written sentiment did not differ by engagement status and lifecycle stage across the 24 engagements with a written response (Kruskal-Wallis H 0.34, df 2, p = 0.842). Fewer than two groups held 10 or more responses, so the groups are reported as counts only: active engagements in Build-Out (7), active engagements in Transition/Cut-Over (10), active engagements in Production (1) and canceled engagements (8).

No test across respondent roles was run, because fewer than two roles had 30 or more written responses. Individual contributors wrote 18, engagement managers wrote 6 and portfolio directors wrote 2 responses, reported as counts only.

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

Across the study set, higher ratings came with less cancellation (r = −0.57, 95% CI −0.67 to −0.45, Holm p < 0.001, n = 150 engagements). The association held after Holm correction in ten of the 20 segment levels tested.

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

Across the active engagements, higher ratings came with higher customer satisfaction (r = 0.68, 95% CI 0.56 to 0.78, Holm p < 0.001, n = 94 engagements). The association held after Holm correction in fourteen of the 16 segment levels tested.

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

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

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

Respondents who gave higher ratings reported lower intent to leave (r = −0.59, 95% CI −0.63 to −0.54, Holm p < 0.001, n = 758 respondents). The association held after Holm correction in thirty-six of the 37 segment levels tested.

Two comparisons set the engagements with at least one negative response against all other engagements. Of the 132 other engagements, 125 had no written response addressing the indicator. These two comparisons therefore do not isolate written sentiment. Of the 18 engagements with a negative response, 7 (38.89%) were canceled, against 49 of the 132 other engagements (37.12%). The difference did not reach significance (Fisher exact p = 1.000). Customer satisfaction among active engagements did not differ between the two groups (mean 58.91% on 11 engagements against 64.78% on 83, Welch t −2.04, p = 0.059).

Two correlations test written sentiment directly. Across the 25 engagements with a written response, the engagement mean written score was not associated with cancellation (point-biserial r −0.18, p = 0.392). Across 26 respondents, the respondent mean written score was not associated with intent to leave (Spearman rho −0.20, p = 0.339). Of the four comparisons, none reached significance at 0.05.

The four associations were compared in pairs on their engagement-level forms, each oriented so that a higher value is the better outcome, across the 150 engagements with all four measures. After Holm correction, four of the six pairs differed. The association with the engagement indicator mean was the strongest and the association with cancellation the weakest. The associations with cancellation and intent to leave did not differ (Holm p = 0.074). The associations with customer satisfaction and intent to leave did not differ (Holm p = 0.137).

Answers clustered within engagements (ICC 0.60). Clustering reduced the 807 respondent answers to the equivalent of about 221 independent answers. Respondent-level intervals therefore used the cluster bootstrap. Respondent-level tests of intent to leave used standard errors that allow for clustering by engagement.

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

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

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

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

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

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

Among the 71 Part 2 indicators with 30 or more written responses, the indicator's share of negative responses, 69.23%, ranked 61st from the highest, against a median of 80.82%. Its mean written score, −3.5, ranked 55th from the most negative, against a median of −4.7.

The indicator most often present in the same responses was C123, “Rate the adequacy of your employer's investment in training its workforce.”, in 9 of the 26 responses (34.62%). C132, “Rate the depth of your employer's experience running agentic AI systems in live production, beyond pilots.” (7 responses), C125, “Rate how well your employer's training programs build financial industry knowledge.” (5 responses), C126, “Rate the maturity of your employer's playbook for taking agentic use cases from concept to governed production.” (5 responses) and C082, “My manager knows agentic AI well.” (4 responses) followed.

The most frequent theme was “Practice invented on the live engagement”, in 5 responses (19.23%). The second was “Comparison to a prior engagement”, in 3 responses.

About half of all ratings of the depth of the employer's knowledge of agentic AI fell in the distressed range (416 of 807 answers, 51.55%). The rating was associated with all four study measures, and every association held under each of the 22 robustness checks. Canceled engagements rated far below the three active stages (mean 2.4, n = 308 respondents).

The written evidence is small. It places agentic knowledge in individual delivery staff and in a few people across the firm, with little of it in the management line that directs the work. Respondents described acquiring the knowledge themselves on live engagements.

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

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

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

Shares of written sentiment are shares of responses and not of the study set of respondents. The written score takes five fixed values, so its mean and median are coarse. Responses clustered within engagements (26 responses on 25 engagements). Intervals for written sentiment therefore used a bootstrap by engagement.

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