Q: Why do Delivering independently AI proficiency participants rate the programme -15 pts below average — what does the D1 calibration data tell us?

The D1 calibration data does not contain AI proficiency-segmented NPS scores, so the -15 pt gap cannot be directly explained by calibration; however, verbatim evidence points to a content-depth mismatch for experienced participants.

The data brief flags that 'Delivering independently' participants rated the programme -15 pts below programme NPS (+55), but NPS broken out by AI proficiency segment is marked 'insufficient data' in the cross-cohort correlations, and no per-cohort NPS-by-proficiency table is provided. What the D1 calibration data does show is that C12 experienced a negative confidence delta (-0.11, D1=4.15 to D2-apply=4.04), the only cohort where participants left less confident than they arrived — and C12 had the largest 'Delivering independently' cohort (n=15 vs. n=7 in C11, n=4 in C13, n=7 in C14). Verbatim evidence from experienced participants consistently signals content pitched below their level: a C12 Delivering independently detractor (EPAM, NPS 5) wrote 'the program needs to be tuned for a narrower audience'; a C14 Delivering independently detractor (Deloitte, NPS 6) wrote 'too high level, too less time to just deep dive'; and a C12 Architect passive wrote that the programme is 'an excellent all-round introduction' but would benefit from 'differentiated tracks.' The total 'Delivering independently' n across all four cohorts is 33 (7+15+4+7), which is sufficient to note a pattern but too thin for statistical confidence on NPS. Key data points: · Total 'Delivering independently' participants across C11-C14: n=33 (C11: 7, C12: 15, C13: 4, C14: 7) · C12 — the cohort with the most 'Delivering independently' participants (n=15) — is the only cohort with a negative confidence delta: D1=4.15 to D2-apply=4.04 (-0.11) · 3 of 4 verbatims from 'Delivering independently' or high-experience participants are detractors or passives citing content being too introductory or insufficiently differentiated by experience level Caveats: NPS segmented by AI proficiency level is not available in the data — the cross-cohort correlations section explicitly marks this as 'insufficient data.' To answer this question with confidence, Ted would need AI proficiency × NPS crosstabs per cohort, ideally with D1 calibration scores also broken out by proficiency band.

2026-07-28 18:22 UTC

Q: What is driving the high passive rate in C13 London June 2-3, 2026 — what do the verbatims say and what would convert them to promoters?

C13's 60% passive rate is driven by 3 verbatims pointing to absent Anthropic presence, a programme pitched too high for non-technical roles, and unresolved questions — not dissatisfaction with content quality.

C13 London (June 2-3) had n=15 respondents, with 9 passives (scores 7-8) and 0 detractors, producing NPS +40. All three passive verbatims are recoverable: one Developer (b.telligent, nps=8) cited 'Anthropic wasn't there' as the gap — suggesting live Anthropic presence is a key credibility signal; one Transformation Lead (Deloitte, nps=7) noted 'more questions than when I came' but framed it positively, indicating unresolved depth rather than poor content; one Transformation Lead (Infosys, nps=7) flagged a need to 'enable someone non-technical to use Claude,' pointing to a persona-fit mismatch for lower-technical attendees. Promoters (nps=9-10) were exclusively Developers who valued hands-on time, evals, and instructor quality — suggesting the programme works well for technical participants but under-delivers for Transformation Leads, who scored NPS +0 (n=4). Converting passives to promoters likely requires: (1) Anthropic staff or engineer presence on the day, (2) clearer non-technical application pathways or a dedicated track, and (3) structured 'answers' to the questions the programme surfaces. Key data points: · C13 passive rate: 60% (9 of 15 respondents scored 7-8) · Transformation Lead NPS in C13: +0 (n=4) vs Developer NPS: +56 (n=9) · 3 of 3 C13 passive verbatims cite Anthropic absence, unresolved questions, or non-technical accessibility — none cite poor content or delivery · C13 confidence delta positive across all 3 dimensions (D1=3.93 → D2-apply=4.27, D2-client=4.33), suggesting learning landed despite passive NPS · C13 cohort is 93% engineering/PM functions (14 of 15 in Engineering or PM), yet Transformation Leads still attended — persona-mix mismatch likely Caveats: n=15 for C13 overall and n=4 for Transformation Leads — findings are directionally plausible but too small for confident segmentation conclusions. Only 3 passive verbatims are available; a post-event follow-up or exit interview with the 9 passives would sharpen conversion levers considerably.

2026-07-28 18:22 UTC

Q: Why did C12 San Francisco May 28-29, 2026 show a negative confidence delta — is this content complexity, audience composition, or expectation mismatch?

C12's -0.11 confidence delta is primarily an expectation mismatch and audience composition problem, not content complexity — three verbatim detractors explicitly cite calibration failure.

C12's apply-confidence dropped from D1=4.15 to D2=4.04, the only negative delta across all four cohorts. The detractor verbatims are diagnostic: a KPMG Partner (Transformation Lead) wrote 'More technical than business oriented — had a lot of struggles with setting up environment than learning'; a KPMG Practitioner said 'the audience level was also not exactly understood for me — there should have been expectation setting'; and an EPAM Director stated 'the program needs to be tuned for narrower audience.' This points directly to expectation mismatch and audience composition as the root causes. C12's function split explains much of this: Engineers were 31 of 67 Day 1 attendees — the heaviest technical skew of any cohort — yet Transformation Leads (NPS +40) and the higher frontier/independent proficiency cohort (13 frontier + 15 delivering independently = 28 of 67) likely found content either too basic technically or too technical for their business role. Deloitte scored NPS +83 vs KPMG's +17 within the same cohort, suggesting org-level expectation-setting varied significantly. Content complexity is unlikely the cause: design and client confidence both held flat (D2-design=4.16, D2-client=4.14), meaning participants did not feel confused — they felt miscalibrated to where they started. Key data points: · Apply-confidence delta: D1=4.15 → D2=4.04, a -0.11 drop — the only negative delta across C11–C14 · KPMG NPS +17 vs Deloitte NPS +83 within C12 (n=18 each) — same content, sharply different outcomes by org · C12 had the heaviest engineer load: 31/67 Day 1 attendees in Engineering function vs 15 in Project/Engagement and 7 in Business Leadership · 13 participants at 'Operating at the frontier' and 15 'Delivering independently' — 28 high-proficiency participants who may have found content insufficiently advanced · 3 of 4 detractor verbatims explicitly cite audience calibration or expectation failure, not content quality Caveats: Confidence delta is based on matched respondents only (n=40 matched of 67 Day 1); segment-level confidence breakdowns by persona or proficiency are not available in this data, which would sharpen whether the delta is concentrated in Transformation Leads or frontier-proficiency participants specifically.

2026-07-28 18:22 UTC

Q: What do promoters and passives say about the programme, and what specifically separates them?

Promoters cite engagement, interaction, and instructor quality; passives consistently flag pace and depth mismatch — specifically that the programme is too broad or too slow for their experience level.

Across all four cohorts, promoter verbatims cluster around three themes: interactive/hands-on format ('Interactive and fun', 'Informative, educational and hands-on', 'Really engaging, great instructors'), strong instructor quality ('The instructors are amazing', 'Friendly, supportive team'), and content quality ('great content and challenges', 'Evals, able to exchange ideas'). Passive verbatims tell a different story: experienced participants consistently felt the programme moved too slowly or covered too much ground without sufficient depth — 'Programme covered a lot of ground in a span of 2 days' (C11, NPS 8), 'excellent all-round introduction… I could promptly summarize learning points… the only thing that'd give them a reason to attend would be differentiated tracks' (C12 KPMG Architect, NPS 8), and 'a bit slow for people who have been using Claude a lot' (C14 Deloitte Transformation Lead, NPS 8). A secondary passive theme is Anthropic presence: 'Anthropic wasn't there' (C13, NPS 8) and 'hearing more from Anthropic engineers… would be better' (C12, NPS 7). The clearest separator is experience-level fit: promoters skew toward 'Applying in practice' proficiency, while passives and detractors more frequently hold 'Operating at the frontier' or 'Delivering independently' profiles — though cross-cohort correlation data is marked insufficient to confirm this statistically. Key data points: · Promoter verbatims across all 4 cohorts consistently reference interactivity, instructor quality, and hands-on content as the reasons for high scores · 4 of 12 passive verbatims explicitly cite pace/depth mismatch — programme too broad, too slow, or insufficiently differentiated for experienced participants · 2 of 12 passive verbatims cite lack of Anthropic engineer presence as a gap (C12 NPS 7, C13 NPS 8) · C13 passive rate is 60% (n=15, flag caution on small n) — highest across all cohorts; C12 passive rate is 34% (n=50) · Passive verbatims most frequently come from 'Operating at the frontier' or 'Applying in practice' profiles, not 'Learning and exploring' Caveats: Verbatim sample is small (approximately 12 passives across 4 cohorts) — patterns are directionally consistent but not statistically robust. Cross-cohort correlation data is marked insufficient, so experience-level as the separator cannot be confirmed quantitatively with this dataset; a verbatim-to-proficiency crosstab with full n would sharpen this answer.

2026-07-28 18:23 UTC