← Home
Partner Basecamp · Cohort 8

San Francisco

May 21-22, 2026
+43
Net Promoter Score
48
Day 1 Responses
92%
Matched Pairs
77%
Day 2 Response Rate
Programme Satisfaction
Positive NPS of +43 — promoters outweigh detractors with moderate passive presence.
+43NPS · n=37
54% Promoters35% Passives11% Detractors
95% CI: +21 → +65  ·  True NPS lies within this range with 95% confidence (n=37 respondents)
Confidence Arc
Strong end-of-Day-1 build confidence — cohort leaves Day 1 ready to apply Claude in client work (mean 4.0/5).
4.0end-of-D1 build / 5
4.1D2 Design / 5
4.2D2 Commercial / 5
4.3D2 Build / 5
Audience
48 participants across 8 organisations — Transformation Lead majority with applying in practice the most common AI experience level.
Top Organisations
Deloitte (21) PwC (8) Ascendion (4) Fractal Analytics (4) Sia (4)
Voice of Participant
What participants said — dominant theme from each NPS segment.
Promoters
Participants valued hands-on tool exploration alongside colleagues in a well-structured program matching real client needs.
Passives
Uneven experience across days and concern that technical depth exceeded attendees' preparation and developer experience levels.
Detractors
Non-technical participants struggled without live demos and guided instruction; format felt better suited for online delivery than in-person.
Survey & Data Quality
Survey & Data Quality
Day 1 responses48
Day 2 responses37 (77% of Day 1)
Matched pairs34 (92% of Day 2)
Orgs resolved48 of 48 respondents matched to named org
Day 2-only respondents3 (8.1%) — suppresses NPS by 1 points; no Day 1 data, excluded from persona & confidence analysis
Organisations
All Day 1 respondents · n=48
Deloitte21 (44%)
PwC8 (17%)
Ascendion4 (8%)
Fractal Analytics4 (8%)
Sia4 (8%)
Lovelytics3 (6%)
Praecipio3 (6%)
Capitalize1 (2%)
Function × Seniority
All Day 1 respondents · n=48
FunctionPractitionerSr PractitionerMgr / Sr MgrDirector+Partner / MDTotal
Engineering2522112
Architecture3322–10
Business Leadership1–3–59
Project / Engmt11231–17
Experience Profile
AI experience level · n=48 · % of row
By Persona
SegmentExploringApplyingDeliveringFrontiern
Transformation Lead31%42%12%15%26
Developer17%42%33%8%12
Architect—20%40%40%10
By Seniority
SegmentExploringApplyingDeliveringFrontiern
Practitioner (0-4 years in role)29%53%12%6%17
Manager or Senior Manager30%40%10%20%10
Senior practitioner (5–9 years)20%30%50%—10
Partner, Managing Director, or Executive—33%17%50%6
Director, Senior Director, or Principal——40%60%5
How would you describe your current experience with AI tools prior to Basecamp?
Learning and exploring
10 (21%)
Applying in practice
18 (38%)
Delivering independently
11 (23%)
Operating at the frontier
9 (19%)
Prior to today, how much had you worked with Claude or the Anthropic API?
Not at all
5 (10%)
A little
18 (38%)
Regularly
25 (52%)
Did the depth land for this audience?
Technical depth perception · n=48 respondents · % of row
By Persona
SegmentToo basicAbout rightToo advancedn
Transformation Lead8%73%19%26
Developer—100%—12
Architect20%70%10%10
By AI Experience Level
SegmentToo basicAbout rightToo advancedn
Applying in practice—78%22%18
Delivering independently9%82%9%11
Learning and exploring—90%10%10
Operating at the frontier33%67%—9
Overall depth distribution
Too basic
4 (8%)
About right
38 (79%)
Too advanced
6 (12%)
Did the pace work across the room?
Session pace perception · n=48 respondents · % of row
By Persona
SegmentToo slowWell pacedToo fastn
Transformation Lead8%81%12%26
Developer—92%8%12
Architect10%90%—10
By AI Experience Level
SegmentToo slowWell pacedToo fastn
Applying in practice11%78%11%18
Delivering independently—100%—11
Learning and exploring—80%20%10
Operating at the frontier11%89%—9
Overall pace distribution
Too slow — could have covered more
3 (6%)
Well paced
41 (85%)
Too fast — not enough time to apply
4 (8%)
How confident were participants to build with Claude after Day 1?
End-of-Day-1 confidence · "How confident are you in your ability to build a client solution using Claude?" · n=48 · 1–5
Mean end-of-D1 build confidence by Persona
Architect4.30/5 · n=10
Developer4.25/5 · n=12
Transformation Lead3.77/5 · n=26
Programme mean: 3.9/5 · 16 cohorts
Mean end-of-D1 build confidence by AI Experience Level
Operating at the frontier4.78/5 · n=9
Delivering independently4.55/5 · n=11
Learning and exploring3.80/5 · n=10
Applying in practice3.39/5 · n=18
Overall end-of-D1 build confidence distribution
1
0 (0%)
2
6 (12%)
3
7 (15%)
4
16 (33%)
5
19 (40%)
How relevant was today's content to your current role?
Content relevance rating · n=48 respondents · 1 = not relevant, 5 = highly relevant
Mean relevance by Persona
Architect4.60/5 · n=10
Developer4.42/5 · n=12
Transformation Lead3.77/5 · n=26
Mean relevance by AI Experience Level
Delivering independently4.45/5 · n=11
Operating at the frontier4.33/5 · n=9
Learning and exploring4.00/5 · n=10
Applying in practice3.83/5 · n=18
Overall relevance distribution
1
1 (2%)
2
1 (2%)
3
7 (15%)
4
22 (46%)
5
17 (35%)
How likely are you to recommend attending this programme to a colleague?
n=37 Day 2 respondents · 77% of Day 1
Score distribution · 0 = not at all likely · 10 = extremely likely
Cohort012345678910
Cohort 8476416
■ Detractors 0–6■ Passives 7–8■ Promoters 9–10
NPS Breakdown
54%
35%
11%
NPS +43 (n=37)
Promoters (9–10)Passives (7–8)Detractors (0–6)
Programme mean: +42 · 16 cohorts
Promoters mean: 51%
Passives mean: 39%
Detractors mean: 9%
NPS by Segment
Organisations, personas, and experience levels · ≥4 respondents shown
By Organisation
Promoters (9–10)Passives (7–8)Detractors (0–6)
Deloitte
59%
29%
12%
NPS +47 (n=17)
PwC
40%
20%
40%
NPS 0 (n=5)
By Persona
Promoters (9–10)Passives (7–8)Detractors (0–6)
Architect
88%
12%
0%
NPS +88 (n=8)
Developer
62%
38%
0%
NPS +62 (n=8)
Transformation Lead
39%
39%
22%
NPS +17 (n=18)
Programme means · 16 cohorts: Architect +48 · Developer +46 · Transformation Lead +38
By AI Experience Level
Promoters (9–10)Passives (7–8)Detractors (0–6)
Operating at the frontier
83%
17%
0%
NPS +83 (n=6)
Delivering independently
75%
25%
0%
NPS +75 (n=8)
Learning and exploring
43%
43%
14%
NPS +29 (n=7)
Applying in practice
38%
38%
23%
NPS +15 (n=13)
Programme means · 16 cohorts: Applying in practice +46 · Learning and exploring +46 · Delivering independently +39 · Operating at the frontier +32
What is the main reason for your score?
n=26 responses · organised by NPS segment
Promoters (score 9–10)· 14 responses
Participants valued hands-on tool exploration alongside colleagues in a well-structured program matching real client needs.
9“this training was excellent. my only suggestion is diving even deeper into building eval frameworks as this is going to be a critical piece of how we build credibility with our clients. i’d like to get some additional reps in with building & shipping realistic solutions, especially in a collaborative framework”
Passives (score 7–8)· 8 responses
Uneven experience across days and concern that technical depth exceeded attendees' preparation and developer experience levels.
8“Some reservations about whether they will be able to get full use out of the sessions given level of exposure to Claude Code and level of explanation/ facilitation ahead of each exercise. Would depend on their preparation or exposure before attending”
Detractors (score 0–6)· 4 responses
Non-technical participants struggled without live demos and guided instruction; format felt better suited for online delivery than in-person.
6“Was a cool experience, but felt like this could have been done online in the current format. I would have appreciated more guided live instruction. They seemed very concerned about not “lecturing” for too long, but I would have liked more. It would have been nice if we did the first couple exercises as a group before doing individual work.”
End-of-Programme Confidence
n=37 respondents · 1 = not confident, 5 = very confident
How confident are you in your ability to design an evaluation for an AI solution?
12345Mean
—3%11%57%30%4.14
Programme mean: 4.1/5 · 16 cohorts
Leaving today, how comfortable are you advising clients on AI when questions arise?
12345Mean
——16%49%35%4.19
Programme mean: 4.2/5 · 16 cohorts
How confident are you in your ability to lead a conversation about Anthropic and Claude with a client?
12345Mean
——14%41%46%4.32
Programme mean: 4.2/5 · 16 cohorts
Confidence Trajectory by Persona
D1→D2 delta · apply-AI confidence · matched pairs only · n=34
Positive delta = confidence grew · negative = dropped
Architect (n=8)
+0.50
Developer (n=8)
+0.12
Transformation Lead (n=18)
+0.33
What one takeaway will you share with a colleague or client?
n=18 responses
Participants recognized Claude's architectural relevance to enterprise software design, beyond current market positioning.
“Anthropic positioning for their product is more aligned to software architecture than they think.”
Participants gained confidence selling AI solutions to clients by learning eval frameworks to rigorously measure accuracy and performance.
“Lots of times we hear that clients don't know how they can trust AI solutions or clearly evaluate how they're doing. I now have a framework (evals) to robustly test AI solutions that I can bring confidently to clients and colleagues. That was super helpful.”
Participants valued peer exchange on practical challenges, though the data provided lacks explicit networking quotes.
“We should all be using Claude or a similar AI.”
Participants learned that rigorous model testing for consistency matters more than adopting the newest tools.
“While I struggled with the labs(and my 'outer), it did show the importance of these components in our agent building. I do like the fact Anthropic stresses that developers need to test the models decide which is best for consistent behavior and response, rather than what's new and shiny.”
Participants found labs valuable despite difficulty, recognizing their role in demonstrating real agent-building fundamentals.
“While I struggled with the labs(and my 'outer), it did show the importance of these components in our agent building.”
What will you build for yourself or apply at work in the next 30 days?
n=27 responses
Participants plan to build custom agents and evaluation systems to automate repetitive tasks into scalable, reusable routines.
“Plan to build custom agents to automate many time sucking tasks into repeatable routines”
Participants intend to apply Claude's code and tool capabilities to accelerate administrative work like PowerPoint and email generation.
“Some playwright tool managed by Claude to accelerate my admin work”
Participants see immediate applications for agentic AI in production environments, from monitoring loads to cloud deployment workflows.
“Agentic ai application and deploy in cloud”
Participants have gained enterprise access to foundational tools and are eager to experiment with agent SDKs despite unclear use cases.
“Cowork workflows. I just got access to it in my enterprise environment, super excited to try it. I also will be developing with the agent SDK, but don't know specifically what I want to build yet.”
Participants are building systematic evaluation frameworks and performance corpora to measure AI project outcomes across their organizations.
“Im going to build a process for us to evaluate all of our ai projects.”
Partner Basecamp Survey Report · Cohort 8 · Generated by analyze.py v2.5
NPS Reason
n=26 responses · grouped by NPS segment
Promotersscore 9–10 · 14 responses
1Well matched against the product capabilities and how to think about the larger set of questions for clients.10
2Interactive, engaging10
3Loved the dedicated time to learn about, and get hands onto the tools. Great to sit with colleagues, competitors, and learn about anthropic tools and application10
4Extremely insightful, great content and activities.10
5Amazing Anthropic Team and phenomenal content and execution! This team deserves a round of applause and thanks 🙏🏽 for their excellent delivery!10
6This is the right level of content for anyone at any level.10
7Facilitatators10
8Extremely cool event, really good to be in the room to see where everyone's at and get to focus on claude building skills for two whole days10
9I thought this was really cool, from a networking perspective to also feeling like Anthropic truly values our partnership. I was able to learn a ton about the entire Anthropic suite of software, models, etc Anthropic has.10
10The mix of hands-on exercises coupled with the practical slides / sessions felt like a good mix of learn/do. The inventory management exercise at the beginning of day 1 felt like it ran a bit too long, but for those who are newer it was a good shakedown.10
11Was an amazing challenge. I learned a lot & was able to skill up quickly from having little Claude code experience.10
12I really enjoyed the content!9
13Depth of training materials9
14this training was excellent. my only suggestion is diving even deeper into building eval frameworks as this is going to be a critical piece of how we build credibility with our clients. i’d like to get some additional reps in with building & shipping realistic solutions, especially in a collaborative framework9
Passivesscore 7–8 · 8 responses
1Very intensive with a lot to learn8
2Some reservations about whether they will be able to get full use out of the sessions given level of exposure to Claude Code and level of explanation/ facilitation ahead of each exercise. Would depend on their preparation or exposure before attending8
3Great course, but very few of my peers are experienced developers and a lot of this material is too technical.8
4Intro to a lotta cool new stuff.7
5I liked it :)7
6Evaluation and process7
7Day 1 was not the same experience as Day 2 in terms of admin setup7
8I do wish this course was more formated in a ‘I do, we do, you do’ format. I wish more examples or demos were done7
Detractorsscore 0–6 · 4 responses
1I think I was a little behind in almost all of the sessions, would be helpful to get a live demo in all the sessions to understand how to get working because I felt a little lost sometimes6
2Felt like course wasn't beginner friendly. Was a lot to take in coming from a non technical background.6
3Good intro to Claude code but I wish there was some more building rather than just running notebooks, etc although they were good to help learn the process6
4Was a cool experience, but felt like this could have been done online in the current format. I would have appreciated more guided live instruction. They seemed very concerned about not “lecturing” for too long, but I would have liked more. It would have been nice if we did the first couple exercises as a group before doing individual work.6
Most Valuable
What one takeaway will you share with a colleague or client? · n=18 responses
1Anthropic positioning for their product is more aligned to software architecture than they think.
2Usuage of playwright mcp server
3We should all be using Claude or a similar AI.
4Everything i have learned
5This training raises the bar for all AI-based trainings!
6That it was awesome and how to build an eval
7That is a world full of opportunities ahead
8Process of using app
9Prompt caching
10Eval methods
11Knowledge
12How important it is to stay up on new features and releases, which is almost a full-time job with the Claude ecosystem!
13eval frameworks are critical in defining the accuracy of AI solutions
14That Anthropic has just about everything you need to go from 0-100 when it comes to integrating AI the right way.
15Lots of times we hear that clients don't know how they can trust AI solutions or clearly evaluate how they're doing. I now have a framework (evals) to robustly test AI solutions that I can bring confidently to clients and colleagues. That was super helpful.
16Importance of evals
17potential is unlimited!
18Come for the food… While I struggled with the labs(and my ‘outer), it did show the importance of these components in our agent building. I do like the fact Anthropic stresses that developers need to test the models decide which is best for consistent behavior and response, rather than what’s new and shiny. Overall a good session that even non-technical users would benefit from.
30-Day Intentions
What will you build for yourself or apply at work? · n=27 responses
1I will be enhancing my ways context progressive disclosure system with more evaluation agents.i also plan on enhancing our intelligence gateway product to build a corpus of performance evaluations too.
2Agentic ai application and deploy in cloud
3A bunch if dashboards
4Application for prod monitoring loads
5Call summ automation
6Continue working with Claude code capabilities
7Apply Claude with PowerPoint creation and emails.
8Plan to build custom agents to automate many time sucking tasks into repeatable routines
9Agents to automate processes and workflows. Features and bug fixes for internal tools
10More than I can capture in words 😎💪🏾!
11I will advance my personal workflows and agents.
12Im going to build a process for us to evaluate all of our ai projects.
13Some playwright tool managed by Claude to accelerate my admin work
14Health care payer apps
15A personal agent to manage day to day repeatable tasks
16Agent team
17Update some of the hacked ways I’ve built older agents using some of these practices
18Use AI for customer problems
19Probably will further productionize the RFP response but I build our hackathon
20training materials for my team to learn how to build production-ready AI systems
21Some custom agents for numerous tasks. I am also a part time student so I’d love to figure out how I can integrate Claude into my school.
22More Claude code usage for technical aspects of the cloud modernization effort.
23Cowork workflows. I just got access to it in my enterprise environment, super excited to try it. I also will be developing with the agent SDK, but don't know specifically what I want to build yet.
24I might expand on the RFP agent to be usable for real clients.
25Build agents
26Automating manual excel analyses for clients & log analyses
27Actively working with other teams on building agents. And while we do work in Claude for much of our work we ultimately have to support our customers with their approved models