AKKR Ascend · Capstone · GTM Functional Group·Fielded June 2026

AI Readiness: From Fragmented Adoption to Shared Advantage

An aggregated read of the AI Readiness Survey across 168 go-to-market respondents at 6 AKKR portfolio companies — and the case for which transformation levers move a PortCo from one adoption stage to the next.

Process A AI Knowledge Exchange Read the process doc  ↗ Process B Embedded Enablement Engine Read the process doc  ↗

The objective

Use the survey to decide where the group should focus its 1–2 deep-dive transformation processes — generic enough to apply at any AKKR PortCo — that move an organization from one AI-adoption stage to the next.

The decision in front of us

Should we focus on education (training the workforce on AI), or on best-practice sharing & an AI repository (capturing and distributing what already works)? The data answers this clearly.

Filter data widgets

Sections 01–04 update to the selected PortCo. The company comparison and strategy sections below stay portfolio-wide.

Respondents
168
Across 6 PortCos
Use AI every day
47%
74% use it most days
No training
43%
Completed no formal training in the last 12 months
Understand governance
79%
Know the guardrails
Want playbooks
65%
Rank it a top-3 lever
Build their own way
31%
No shared method
Discovery is hard
69%
Can't easily find others' work

01 — Who responded

Sample

By portfolio company

Appspace is ~39% of the sample — aggregate read is weighted toward it

167
responses
Appspace 65 39%
Kantata 31 19%
ESO 24 14%
Abrigo 22 13%
StoreForce 16 10%
VisiQuate 9 5%

Sample

By GTM role

Front-line, customer-facing roles dominate — this is a revenue-team read

Account Executives (AE)
55
Marketing
35
Sales Leadership
25
Sales Engineers (SE)
20
Sales Development Reps (SDR)
15
Revenue Operations
13

Sample

By seniority

Two-thirds are individual contributors — the people who actually run AI in the workflow

IC
114
Manager
11
Director
22
VP/SVP
13
Exec
2

02 — The paradox: heavy use, thin scaffolding

Read of the room

People aren't waiting for permission to use AI — they're waiting for structure.

Usage is high — 74% use AI daily or most days (47% every single day) and 79% say they understand the governance guardrails. Yet 43% have had no formal AI training in a year, and understanding of how AI works sits at just 3.51/5. The workforce has adopted AI faster than the organization has equipped it. That gap is not closed by awareness campaigns — it's closed by giving people vetted, reusable structure.

Behavior

How often AI is actually used

74% use it daily or most days — adoption is real and habitual

Daily
47%
Often
27%
Sometimes
16%
Rarely
8%
Never
2%

Capability

Self-rated understanding of how AI works

Mean 3.51/5 — competent users, shallow foundations

77
66
11
Low (1-2)Mid (3)High (4-5)

Capability

Confidence in customer AI conversations

Mean 3.28/5

27
51
58
15

Sentiment

Satisfaction with available AI tools

Mean 3.62/5 — tools are fine; the gap is enablement

15
48
55
35

Investment

Formal AI training in last 12 months

68 of 168 have had none

None
68
Internal only
36
External only
28
Both
26

Appetite

Time/month they'd give to AI learning

60% would commit 2+ hrs — real appetite, but capped

< 1 hr
12
1–2 hrs
51
2–4 hrs
54
4+ hrs
39

Pairs with the training gap on the left: the appetite is there, but capped. Design the program for a realistic 1–4 hr/month budget — short, hands-on, in-workflow.

Behavior

What people use AI for today

% of respondents selecting each — research, content and meeting prep lead; deep workflow automation lags

Research
86%
Content creation
74%
Meeting summaries
71%
Brainstorming
63%
Knowledge retrieval
62%
Internal documentation
53%
Client communications
51%
Data analysis
49%

03 — The real problem: fragmentation

Read of the room

Everyone is building in private. Almost no one can find what already works.

Only 16% say their team uses standardized AI approaches. 31% say most people have built their own methods, and another set has no visibility into peers at all. When a colleague builds something useful, 69% find it hard or only neutral to discover and reuse — and the top worry about self-built tools is quality and accuracy (61%). This is the signature of an organization that needs a shared, governed knowledge layer — not more individual upskilling.

Maturity

How standardized is AI use on your team?

Fewer than 1 in 6 teams operate with shared approaches

25
57
48
23
StandardizedSome standard, variesEveryone their own wayNo visibility

Friction

Can you discover & reuse a colleague's AI work?

Difficult + neutral outweigh easy almost 2 to 1

11
38
47
37
25
EasyNeutralDifficult

Friction

Biggest friction points using AI today

% selecting each — inconsistent output and lack of training top the list

Outputs are inconsistent
42%
Lack of training
42%
Don’t trust the results
35%
Too many good ideas
29%
not sure how to execute
29%
Not sure which tool to use
27%
Security or compliance concerns
21%
Duplicate tools solving the same problem
15%
Lack of approved prompts and workflows
15%

Risk

Concerns about teammates building their own AI

% selecting each — the cost of no shared standard

Quality and accuracy
61%
Lack of documentation
44%
Duplicate solutions
37%
Security risks
31%
Maintenance challenges
28%
Lack of ownership
28%
Compliance concerns
28%
Version control issues
27%

04 — The decisive signal: what would move the needle

The question, answered

Lead with best-practice sharing & a governed AI repository. Treat education as the delivery mechanism, not the headline.

Asked to rank what would most help scale AI adoption, respondents put reusable, shared assets at the top. The four highest-ranked levers — team playbooks, a central repository of approved tools, a standardized prompt library, and best-practice documentation — average 49% top-3 selection. Stand-alone training & office hours sits well below at 30%. The signal is clear: people don't primarily need to be taught AI in the abstract — they need to find, trust, and reuse what their peers have already proven works.

Greatest value in scaling AI adoption — % ranking top 3
Team-specific AI playbooks
65%
Central repository of approved tools
52%
Standardized prompt library
47%
AI best-practice documentation
32%
Approved MCP integrations
26%
Training and office hours
30%
Internal marketplace for AI solutions
12%
Governance and approval process
8%
AI champions/community of practice
23%
Tool ratings/reviews from colleagues
6%
Repository / sharing (top 4)Other infrastructureStand-alone training
When training does happen, how should it land? — % top 3
Live workshop
68%
Self-paced modules
66%
Video tutorials
52%
Peer learning
41%
Quick reference guides
44%
1:1 coaching
30%

The preferred formats are hands-on and self-serve — live workshops and self-paced modules — not lecture courses. Education works best when it's embedded in the repository: worked examples, guided practice, office hours tied to real plays.

Respondents
65
Appspace · 39% of all responses
Use AI every day
47%
68% use it most days
No training
66%
Completed no formal training in the last 12 months
Understand governance
65%
Know the guardrails
Want playbooks
67%
Rank it a top-3 lever
Build their own way
37%
No shared method
Discovery is hard
68%
Can't easily find others' work

01 — Who responded

Sample

Appspace sample

This PortCo's slice of the survey

65
respondents · 39% of the 168 surveyed
Largest role: Account Executives (AE)

Sample

By GTM role

Front-line, customer-facing roles dominate — this is a revenue-team read

Account Executives (AE)
31
Marketing
14
Sales Leadership
8
Sales Development Reps (SDR)
6
Sales Engineers (SE)
1
Revenue Operations
1

Sample

By seniority

Two-thirds are individual contributors — the people who actually run AI in the workflow

IC
51
Manager
4
Director
6
VP/SVP
3
Exec
0

02 — The paradox: heavy use, thin scaffolding

Read of the room

People aren't waiting for permission to use AI — they're waiting for structure.

Usage is high — 68% use AI daily or most days (47% every single day) and 65% say they understand the governance guardrails. Yet 66% have had no formal AI training in a year, and understanding of how AI works sits at just 3.6/5. The workforce has adopted AI faster than the organization has equipped it. That gap is not closed by awareness campaigns — it's closed by giving people vetted, reusable structure.

Behavior

How often AI is actually used

68% use it daily or most days — adoption is real and habitual

Daily
47%
Often
22%
Sometimes
18%
Rarely
12%
Never
2%

Capability

Self-rated understanding of how AI works

Mean 3.6/5 — competent users, shallow foundations

26
29
4
Low (1-2)Mid (3)High (4-5)

Capability

Confidence in customer AI conversations

Mean 3.2/5

4
8
24
18
5

Sentiment

Satisfaction with available AI tools

Mean 3.22/5 — tools are fine; the gap is enablement

4
11
20
18
7

Investment

Formal AI training in last 12 months

39 of 65 have had none

None
39
Internal only
4
External only
13
Both
3

Appetite

Time/month they'd give to AI learning

61% would commit 2+ hrs — real appetite, but capped

< 1 hr
6
1–2 hrs
17
2–4 hrs
28
4+ hrs
8

Pairs with the training gap on the left: the appetite is there, but capped. Design the program for a realistic 1–4 hr/month budget — short, hands-on, in-workflow.

Behavior

What people use AI for today

% of respondents selecting each — research, content and meeting prep lead; deep workflow automation lags

Research
86%
Content creation
74%
Meeting summaries
72%
Client communications
60%
Brainstorming
60%
Knowledge retrieval
59%
Internal documentation
41%
Data analysis
36%

03 — The real problem: fragmentation

Read of the room

Everyone is building in private. Almost no one can find what already works.

Only 12% say their team uses standardized AI approaches. 37% say most people have built their own methods, and another set has no visibility into peers at all. When a colleague builds something useful, 68% find it hard or only neutral to discover and reuse — and the top worry about self-built tools is quality and accuracy (63%). This is the signature of an organization that needs a shared, governed knowledge layer — not more individual upskilling.

Maturity

How standardized is AI use on your team?

Fewer than 1 in 6 teams operate with shared approaches

7
20
21
9
StandardizedSome standard, variesEveryone their own wayNo visibility

Friction

Can you discover & reuse a colleague's AI work?

Difficult + neutral outweigh easy almost 2 to 1

6
13
13
15
12
EasyNeutralDifficult

Friction

Biggest friction points using AI today

% selecting each — inconsistent output and lack of training top the list

Outputs are inconsistent
45%
Lack of training
45%
Don’t trust the results
42%
Security or compliance concerns
27%
Not sure which tool to use
27%
Too many good ideas
20%
not sure how to execute
20%
Lack of approved prompts and workflows
18%
Duplicate tools solving the same problem
17%

Risk

Concerns about teammates building their own AI

% selecting each — the cost of no shared standard

Quality and accuracy
63%
Lack of documentation
53%
Lack of ownership
37%
Security risks
36%
Maintenance challenges
36%
Compliance concerns
32%
Duplicate solutions
32%
Difficulty determining which solution is approved
29%

04 — The decisive signal: what would move the needle

The question, answered

Lead with best-practice sharing & a governed AI repository. Treat education as the delivery mechanism, not the headline.

Asked to rank what would most help scale AI adoption, respondents put reusable, shared assets at the top. The four highest-ranked levers — team playbooks, a central repository of approved tools, a standardized prompt library, and best-practice documentation — average 49% top-3 selection. Stand-alone training & office hours sits well below at 25%. The signal is clear: people don't primarily need to be taught AI in the abstract — they need to find, trust, and reuse what their peers have already proven works.

Greatest value in scaling AI adoption — % ranking top 3
Central repository of approved tools
56%
Team-specific AI playbooks
67%
Standardized prompt library
40%
AI best-practice documentation
32%
Approved MCP integrations
26%
Training and office hours
25%
Governance and approval process
12%
AI champions/community of practice
25%
Internal marketplace for AI solutions
9%
Tool ratings/reviews from colleagues
9%
Repository / sharing (top 4)Other infrastructureStand-alone training
When training does happen, how should it land? — % top 3
Self-paced modules
72%
Live workshop
55%
Video tutorials
64%
Quick reference guides
53%
Peer learning
33%
1:1 coaching
22%

The preferred formats are hands-on and self-serve — live workshops and self-paced modules — not lecture courses. Education works best when it's embedded in the repository: worked examples, guided practice, office hours tied to real plays.

Respondents
31
Kantata · 18% of all responses
Use AI every day
54%
96% use it most days
No training
24%
Completed no formal training in the last 12 months
Understand governance
100%
Know the guardrails
Want playbooks
71%
Rank it a top-3 lever
Build their own way
21%
No shared method
Discovery is hard
64%
Can't easily find others' work

01 — Who responded

Sample

Kantata sample

This PortCo's slice of the survey

31
respondents · 18% of the 168 surveyed
Largest role: Sales Engineers (SE)

Sample

By GTM role

Front-line, customer-facing roles dominate — this is a revenue-team read

Sales Engineers (SE)
8
Account Executives (AE)
7
Marketing
6
Revenue Operations
5
Sales Development Reps (SDR)
3
Sales Leadership
1

Sample

By seniority

Two-thirds are individual contributors — the people who actually run AI in the workflow

IC
21
Manager
4
Director
4
VP/SVP
1
Exec
0

02 — The paradox: heavy use, thin scaffolding

Read of the room

People aren't waiting for permission to use AI — they're waiting for structure.

Usage is high — 96% use AI daily or most days (54% every single day) and 100% say they understand the governance guardrails. Yet 24% have had no formal AI training in a year, and understanding of how AI works sits at just 3.38/5. The workforce has adopted AI faster than the organization has equipped it. That gap is not closed by awareness campaigns — it's closed by giving people vetted, reusable structure.

Behavior

How often AI is actually used

96% use it daily or most days — adoption is real and habitual

Daily
54%
Often
43%
Sometimes
0%
Rarely
4%
Never
0%

Capability

Self-rated understanding of how AI works

Mean 3.38/5 — competent users, shallow foundations

2
16
9
2
Low (1-2)Mid (3)High (4-5)

Capability

Confidence in customer AI conversations

Mean 3.52/5

2
2
7
15
3

Sentiment

Satisfaction with available AI tools

Mean 4/5 — tools are fine; the gap is enablement

2
4
10
11

Investment

Formal AI training in last 12 months

7 of 31 have had none

None
7
Internal only
8
External only
7
Both
7

Appetite

Time/month they'd give to AI learning

61% would commit 2+ hrs — real appetite, but capped

< 1 hr
0
1–2 hrs
11
2–4 hrs
6
4+ hrs
11

Pairs with the training gap on the left: the appetite is there, but capped. Design the program for a realistic 1–4 hr/month budget — short, hands-on, in-workflow.

Behavior

What people use AI for today

% of respondents selecting each — research, content and meeting prep lead; deep workflow automation lags

Research
96%
Content creation
79%
Brainstorming
79%
Meeting summaries
75%
Data analysis
64%
Internal documentation
61%
Knowledge retrieval
57%
Client communications
46%

03 — The real problem: fragmentation

Read of the room

Everyone is building in private. Almost no one can find what already works.

Only 21% say their team uses standardized AI approaches. 21% say most people have built their own methods, and another set has no visibility into peers at all. When a colleague builds something useful, 64% find it hard or only neutral to discover and reuse — and the top worry about self-built tools is quality and accuracy (54%). This is the signature of an organization that needs a shared, governed knowledge layer — not more individual upskilling.

Maturity

How standardized is AI use on your team?

Fewer than 1 in 6 teams operate with shared approaches

6
13
6
3
StandardizedSome standard, variesEveryone their own wayNo visibility

Friction

Can you discover & reuse a colleague's AI work?

Difficult + neutral outweigh easy almost 2 to 1

9
8
7
3
EasyNeutralDifficult

Friction

Biggest friction points using AI today

% selecting each — inconsistent output and lack of training top the list

Too many good ideas
50%
not sure how to execute
50%
Outputs are inconsistent
46%
Lack of training
39%
Difficulty sharing what works with others
29%
Don’t trust the results
29%
Too many tools available
29%
Not sure which tool to use
25%
Security or compliance concerns
18%

Risk

Concerns about teammates building their own AI

% selecting each — the cost of no shared standard

Quality and accuracy
54%
Lack of documentation
43%
Duplicate solutions
39%
Lack of ownership
32%
Maintenance challenges
29%
Difficulty determining which solution is approved
25%
No major concerns
25%
Version control issues
21%

04 — The decisive signal: what would move the needle

The question, answered

Lead with best-practice sharing & a governed AI repository. Treat education as the delivery mechanism, not the headline.

Asked to rank what would most help scale AI adoption, respondents put reusable, shared assets at the top. The four highest-ranked levers — team playbooks, a central repository of approved tools, a standardized prompt library, and best-practice documentation — average 44% top-3 selection. Stand-alone training & office hours sits well below at 39%. The signal is clear: people don't primarily need to be taught AI in the abstract — they need to find, trust, and reuse what their peers have already proven works.

Greatest value in scaling AI adoption — % ranking top 3
Team-specific AI playbooks
71%
Standardized prompt library
57%
Training and office hours
39%
AI best-practice documentation
25%
Central repository of approved tools
21%
Approved MCP integrations
32%
AI champions/community of practice
43%
Governance and approval process
7%
Internal marketplace for AI solutions
4%
Tool ratings/reviews from colleagues
0%
Repository / sharing (top 4)Other infrastructureStand-alone training
When training does happen, how should it land? — % top 3
Live workshop
79%
Video tutorials
61%
Self-paced modules
46%
Peer learning
50%
1:1 coaching
36%
Quick reference guides
29%

The preferred formats are hands-on and self-serve — live workshops and self-paced modules — not lecture courses. Education works best when it's embedded in the repository: worked examples, guided practice, office hours tied to real plays.

Respondents
24
ESO · 14% of all responses
Use AI every day
29%
71% use it most days
No training
13%
Completed no formal training in the last 12 months
Understand governance
92%
Know the guardrails
Want playbooks
65%
Rank it a top-3 lever
Build their own way
32%
No shared method
Discovery is hard
71%
Can't easily find others' work

01 — Who responded

Sample

ESO sample

This PortCo's slice of the survey

24
respondents · 14% of the 168 surveyed
Largest role: Account Executives (AE)

Sample

By GTM role

Front-line, customer-facing roles dominate — this is a revenue-team read

Account Executives (AE)
10
Revenue Operations
4
Sales Leadership
4
Marketing
3
Sales Engineers (SE)
2
Sales Development Reps (SDR)
1

Sample

By seniority

Two-thirds are individual contributors — the people who actually run AI in the workflow

IC
14
Manager
1
Director
6
VP/SVP
1
Exec
1

02 — The paradox: heavy use, thin scaffolding

Read of the room

People aren't waiting for permission to use AI — they're waiting for structure.

Usage is high — 71% use AI daily or most days (29% every single day) and 92% say they understand the governance guardrails. Yet 13% have had no formal AI training in a year, and understanding of how AI works sits at just 3.35/5. The workforce has adopted AI faster than the organization has equipped it. That gap is not closed by awareness campaigns — it's closed by giving people vetted, reusable structure.

Behavior

How often AI is actually used

71% use it daily or most days — adoption is real and habitual

Daily
29%
Often
42%
Sometimes
21%
Rarely
4%
Never
4%

Capability

Self-rated understanding of how AI works

Mean 3.35/5 — competent users, shallow foundations

2
12
8
Low (1-2)Mid (3)High (4-5)

Capability

Confidence in customer AI conversations

Mean 2.92/5

8
7
8

Sentiment

Satisfaction with available AI tools

Mean 3.67/5 — tools are fine; the gap is enablement

10
12
2

Investment

Formal AI training in last 12 months

3 of 24 have had none

None
3
Internal only
9
External only
2
Both
9

Appetite

Time/month they'd give to AI learning

48% would commit 2+ hrs — real appetite, but capped

< 1 hr
3
1–2 hrs
9
2–4 hrs
6
4+ hrs
5

Pairs with the training gap on the left: the appetite is there, but capped. Design the program for a realistic 1–4 hr/month budget — short, hands-on, in-workflow.

Behavior

What people use AI for today

% of respondents selecting each — research, content and meeting prep lead; deep workflow automation lags

Research
71%
Meeting summaries
62%
Internal documentation
58%
Content creation
54%
Data analysis
50%
Knowledge retrieval
50%
Client communications
46%
Brainstorming
46%

03 — The real problem: fragmentation

Read of the room

Everyone is building in private. Almost no one can find what already works.

Only 9% say their team uses standardized AI approaches. 32% say most people have built their own methods, and another set has no visibility into peers at all. When a colleague builds something useful, 71% find it hard or only neutral to discover and reuse — and the top worry about self-built tools is quality and accuracy (62%). This is the signature of an organization that needs a shared, governed knowledge layer — not more individual upskilling.

Maturity

How standardized is AI use on your team?

Fewer than 1 in 6 teams operate with shared approaches

2
9
7
4
StandardizedSome standard, variesEveryone their own wayNo visibility

Friction

Can you discover & reuse a colleague's AI work?

Difficult + neutral outweigh easy almost 2 to 1

7
5
5
7
EasyNeutralDifficult

Friction

Biggest friction points using AI today

% selecting each — inconsistent output and lack of training top the list

Not sure which tool to use
43%
Don’t trust the results
39%
Lack of training
39%
Too many good ideas
30%
not sure how to execute
30%
Outputs are inconsistent
26%
Security or compliance concerns
17%
Duplicate tools solving the same problem
13%
Difficulty sharing what works with others
9%

Risk

Concerns about teammates building their own AI

% selecting each — the cost of no shared standard

Quality and accuracy
62%
Security risks
46%
Compliance concerns
38%
Duplicate solutions
38%
Lack of documentation
38%
Version control issues
33%
Lack of ownership
25%
Difficulty determining which solution is approved
25%

04 — The decisive signal: what would move the needle

The question, answered

Lead with best-practice sharing & a governed AI repository. Treat education as the delivery mechanism, not the headline.

Asked to rank what would most help scale AI adoption, respondents put reusable, shared assets at the top. The four highest-ranked levers — team playbooks, a central repository of approved tools, a standardized prompt library, and best-practice documentation — average 53% top-3 selection. Stand-alone training & office hours sits well below at 39%. The signal is clear: people don't primarily need to be taught AI in the abstract — they need to find, trust, and reuse what their peers have already proven works.

Greatest value in scaling AI adoption — % ranking top 3
Central repository of approved tools
52%
Standardized prompt library
57%
Team-specific AI playbooks
65%
AI best-practice documentation
39%
Training and office hours
39%
Approved MCP integrations
17%
Governance and approval process
9%
Internal marketplace for AI solutions
13%
Tool ratings/reviews from colleagues
0%
AI champions/community of practice
9%
Repository / sharing (top 4)Other infrastructureStand-alone training
When training does happen, how should it land? — % top 3
Live workshop
91%
Self-paced modules
65%
Peer learning
43%
1:1 coaching
35%
Quick reference guides
35%
Video tutorials
30%

The preferred formats are hands-on and self-serve — live workshops and self-paced modules — not lecture courses. Education works best when it's embedded in the repository: worked examples, guided practice, office hours tied to real plays.

Respondents
22
Abrigo · 13% of all responses
Use AI every day
59%
73% use it most days
No training
23%
Completed no formal training in the last 12 months
Understand governance
91%
Know the guardrails
Want playbooks
59%
Rank it a top-3 lever
Build their own way
27%
No shared method
Discovery is hard
86%
Can't easily find others' work

01 — Who responded

Sample

Abrigo sample

This PortCo's slice of the survey

22
respondents · 13% of the 168 surveyed
Largest role: Sales Leadership

Sample

By GTM role

Front-line, customer-facing roles dominate — this is a revenue-team read

Sales Leadership
7
Sales Engineers (SE)
7
Marketing
6
Revenue Operations
2

Sample

By seniority

Two-thirds are individual contributors — the people who actually run AI in the workflow

IC
10
Manager
1
Director
3
VP/SVP
6
Exec
0

02 — The paradox: heavy use, thin scaffolding

Read of the room

People aren't waiting for permission to use AI — they're waiting for structure.

Usage is high — 73% use AI daily or most days (59% every single day) and 91% say they understand the governance guardrails. Yet 23% have had no formal AI training in a year, and understanding of how AI works sits at just 3.5/5. The workforce has adopted AI faster than the organization has equipped it. That gap is not closed by awareness campaigns — it's closed by giving people vetted, reusable structure.

Behavior

How often AI is actually used

73% use it daily or most days — adoption is real and habitual

Daily
59%
Often
14%
Sometimes
23%
Rarely
5%
Never
0%

Capability

Self-rated understanding of how AI works

Mean 3.5/5 — competent users, shallow foundations

13
7
2
Low (1-2)Mid (3)High (4-5)

Capability

Confidence in customer AI conversations

Mean 3.68/5

2
4
11
4

Sentiment

Satisfaction with available AI tools

Mean 4.23/5 — tools are fine; the gap is enablement

6
5
11

Investment

Formal AI training in last 12 months

5 of 22 have had none

None
5
Internal only
11
External only
1
Both
5

Appetite

Time/month they'd give to AI learning

57% would commit 2+ hrs — real appetite, but capped

< 1 hr
1
1–2 hrs
8
2–4 hrs
7
4+ hrs
5

Pairs with the training gap on the left: the appetite is there, but capped. Design the program for a realistic 1–4 hr/month budget — short, hands-on, in-workflow.

Behavior

What people use AI for today

% of respondents selecting each — research, content and meeting prep lead; deep workflow automation lags

Knowledge retrieval
86%
Research
82%
Content creation
82%
Brainstorming
68%
Meeting summaries
64%
Internal documentation
55%
Data analysis
50%
Client communications
45%

03 — The real problem: fragmentation

Read of the room

Everyone is building in private. Almost no one can find what already works.

Only 32% say their team uses standardized AI approaches. 27% say most people have built their own methods, and another set has no visibility into peers at all. When a colleague builds something useful, 86% find it hard or only neutral to discover and reuse — and the top worry about self-built tools is quality and accuracy (77%). This is the signature of an organization that needs a shared, governed knowledge layer — not more individual upskilling.

Maturity

How standardized is AI use on your team?

Fewer than 1 in 6 teams operate with shared approaches

7
8
6
StandardizedSome standard, variesEveryone their own wayNo visibility

Friction

Can you discover & reuse a colleague's AI work?

Difficult + neutral outweigh easy almost 2 to 1

3
12
6
EasyNeutralDifficult

Friction

Biggest friction points using AI today

% selecting each — inconsistent output and lack of training top the list

Outputs are inconsistent
55%
Lack of training
45%
Too many good ideas
41%
not sure how to execute
41%
Don’t trust the results
27%
Not sure which tool to use
23%
Security or compliance concerns
23%
Lack of approved prompts and workflows
18%
Duplicate tools solving the same problem
5%

Risk

Concerns about teammates building their own AI

% selecting each — the cost of no shared standard

Quality and accuracy
77%
Lack of documentation
45%
Duplicate solutions
36%
Security risks
32%
Version control issues
27%
Compliance concerns
27%
Lack of ownership
23%
Maintenance challenges
23%

04 — The decisive signal: what would move the needle

The question, answered

Lead with best-practice sharing & a governed AI repository. Treat education as the delivery mechanism, not the headline.

Asked to rank what would most help scale AI adoption, respondents put reusable, shared assets at the top. The four highest-ranked levers — team playbooks, a central repository of approved tools, a standardized prompt library, and best-practice documentation — average 42% top-3 selection. Stand-alone training & office hours sits well below at 27%. The signal is clear: people don't primarily need to be taught AI in the abstract — they need to find, trust, and reuse what their peers have already proven works.

Greatest value in scaling AI adoption — % ranking top 3
Central repository of approved tools
50%
Team-specific AI playbooks
59%
Standardized prompt library
41%
Approved MCP integrations
41%
AI best-practice documentation
18%
Training and office hours
27%
AI champions/community of practice
27%
Internal marketplace for AI solutions
14%
Tool ratings/reviews from colleagues
18%
Governance and approval process
5%
Repository / sharing (top 4)Other infrastructureStand-alone training
When training does happen, how should it land? — % top 3
Self-paced modules
76%
Live workshop
62%
Video tutorials
52%
Quick reference guides
48%
Peer learning
33%
1:1 coaching
29%

The preferred formats are hands-on and self-serve — live workshops and self-paced modules — not lecture courses. Education works best when it's embedded in the repository: worked examples, guided practice, office hours tied to real plays.

Respondents
16
StoreForce · 10% of all responses
Use AI every day
67%
87% use it most days
No training
62%
Completed no formal training in the last 12 months
Understand governance
60%
Know the guardrails
Want playbooks
53%
Rank it a top-3 lever
Build their own way
20%
No shared method
Discovery is hard
53%
Can't easily find others' work

01 — Who responded

Sample

StoreForce sample

This PortCo's slice of the survey

16
respondents · 10% of the 168 surveyed
Largest role: Marketing

Sample

By GTM role

Front-line, customer-facing roles dominate — this is a revenue-team read

Marketing
5
Account Executives (AE)
5
Sales Leadership
4
Sales Development Reps (SDR)
1
Sales Engineers (SE)
1

Sample

By seniority

Two-thirds are individual contributors — the people who actually run AI in the workflow

IC
11
Manager
1
Director
2
VP/SVP
0
Exec
1

02 — The paradox: heavy use, thin scaffolding

Read of the room

People aren't waiting for permission to use AI — they're waiting for structure.

Usage is high — 87% use AI daily or most days (67% every single day) and 60% say they understand the governance guardrails. Yet 62% have had no formal AI training in a year, and understanding of how AI works sits at just 3.69/5. The workforce has adopted AI faster than the organization has equipped it. That gap is not closed by awareness campaigns — it's closed by giving people vetted, reusable structure.

Behavior

How often AI is actually used

87% use it daily or most days — adoption is real and habitual

Daily
67%
Often
20%
Sometimes
13%
Rarely
0%
Never
0%

Capability

Self-rated understanding of how AI works

Mean 3.69/5 — competent users, shallow foundations

1
5
8
2
Low (1-2)Mid (3)High (4-5)

Capability

Confidence in customer AI conversations

Mean 3.2/5

5
5
2
3

Sentiment

Satisfaction with available AI tools

Mean 3.53/5 — tools are fine; the gap is enablement

1
1
5
5
3

Investment

Formal AI training in last 12 months

10 of 16 have had none

None
10
Internal only
0
External only
4
Both
2

Appetite

Time/month they'd give to AI learning

67% would commit 2+ hrs — real appetite, but capped

< 1 hr
1
1–2 hrs
4
2–4 hrs
5
4+ hrs
5

Pairs with the training gap on the left: the appetite is there, but capped. Design the program for a realistic 1–4 hr/month budget — short, hands-on, in-workflow.

Behavior

What people use AI for today

% of respondents selecting each — research, content and meeting prep lead; deep workflow automation lags

Research
100%
Content creation
87%
Data analysis
73%
Internal documentation
73%
Brainstorming
73%
Meeting summaries
73%
Knowledge retrieval
67%
Client communications
53%

03 — The real problem: fragmentation

Read of the room

Everyone is building in private. Almost no one can find what already works.

Only 20% say their team uses standardized AI approaches. 20% say most people have built their own methods, and another set has no visibility into peers at all. When a colleague builds something useful, 53% find it hard or only neutral to discover and reuse — and the top worry about self-built tools is quality and accuracy (36%). This is the signature of an organization that needs a shared, governed knowledge layer — not more individual upskilling.

Maturity

How standardized is AI use on your team?

Fewer than 1 in 6 teams operate with shared approaches

3
6
3
3
StandardizedSome standard, variesEveryone their own wayNo visibility

Friction

Can you discover & reuse a colleague's AI work?

Difficult + neutral outweigh easy almost 2 to 1

2
5
4
2
2
EasyNeutralDifficult

Friction

Biggest friction points using AI today

% selecting each — inconsistent output and lack of training top the list

Outputs are inconsistent
27%
Don’t trust the results
27%
Lack of training
20%
Duplicate tools solving the same problem
20%
Too many good ideas
20%
not sure how to execute
20%
Security or compliance concerns
20%
Not sure which tool to use
13%
Too many tools available
13%

Risk

Concerns about teammates building their own AI

% selecting each — the cost of no shared standard

Duplicate solutions
43%
No major concerns
43%
Quality and accuracy
36%
Security risks
29%
Lack of documentation
29%
Compliance concerns
21%
Maintenance challenges
14%
Version control issues
14%

04 — The decisive signal: what would move the needle

The question, answered

Lead with best-practice sharing & a governed AI repository. Treat education as the delivery mechanism, not the headline.

Asked to rank what would most help scale AI adoption, respondents put reusable, shared assets at the top. The four highest-ranked levers — team playbooks, a central repository of approved tools, a standardized prompt library, and best-practice documentation — average 52% top-3 selection. Stand-alone training & office hours sits well below at 20%. The signal is clear: people don't primarily need to be taught AI in the abstract — they need to find, trust, and reuse what their peers have already proven works.

Greatest value in scaling AI adoption — % ranking top 3
Central repository of approved tools
67%
Standardized prompt library
53%
Team-specific AI playbooks
53%
AI best-practice documentation
33%
Approved MCP integrations
27%
Training and office hours
20%
Internal marketplace for AI solutions
27%
Governance and approval process
0%
Tool ratings/reviews from colleagues
7%
AI champions/community of practice
13%
Repository / sharing (top 4)Other infrastructureStand-alone training
When training does happen, how should it land? — % top 3
Live workshop
60%
Self-paced modules
60%
Peer learning
60%
1:1 coaching
40%
Quick reference guides
40%
Video tutorials
40%

The preferred formats are hands-on and self-serve — live workshops and self-paced modules — not lecture courses. Education works best when it's embedded in the repository: worked examples, guided practice, office hours tied to real plays.

Respondents
9
VisiQuate · 5% of all responses
Use AI every day
22%
44% use it most days
No training
50%
Completed no formal training in the last 12 months
Understand governance
78%
Know the guardrails
Want playbooks
67%
Rank it a top-3 lever
Build their own way
50%
No shared method
Discovery is hard
67%
Can't easily find others' work

01 — Who responded

Sample

VisiQuate sample

This PortCo's slice of the survey

9
respondents · 5% of the 168 surveyed
Largest role: Sales Development Reps (SDR)

Sample

By GTM role

Front-line, customer-facing roles dominate — this is a revenue-team read

Sales Development Reps (SDR)
4
Account Executives (AE)
2
Sales Leadership
1
Marketing
1
Sales Engineers (SE)
1

Sample

By seniority

Two-thirds are individual contributors — the people who actually run AI in the workflow

IC
6
Manager
0
Director
1
VP/SVP
2
Exec
0

02 — The paradox: heavy use, thin scaffolding

Read of the room

People aren't waiting for permission to use AI — they're waiting for structure.

Usage is high — 44% use AI daily or most days (22% every single day) and 78% say they understand the governance guardrails. Yet 50% have had no formal AI training in a year, and understanding of how AI works sits at just 3.56/5. The workforce has adopted AI faster than the organization has equipped it. That gap is not closed by awareness campaigns — it's closed by giving people vetted, reusable structure.

Behavior

How often AI is actually used

44% use it daily or most days — adoption is real and habitual

Daily
22%
Often
22%
Sometimes
22%
Rarely
33%
Never
0%

Capability

Self-rated understanding of how AI works

Mean 3.56/5 — competent users, shallow foundations

4
5
Low (1-2)Mid (3)High (4-5)

Capability

Confidence in customer AI conversations

Mean 3.22/5

2
3
4

Sentiment

Satisfaction with available AI tools

Mean 3.67/5 — tools are fine; the gap is enablement

1
2
5
1

Investment

Formal AI training in last 12 months

4 of 9 have had none

None
4
Internal only
3
External only
1
Both
0

Appetite

Time/month they'd give to AI learning

78% would commit 2+ hrs — real appetite, but capped

< 1 hr
0
1–2 hrs
2
2–4 hrs
2
4+ hrs
5

Pairs with the training gap on the left: the appetite is there, but capped. Design the program for a realistic 1–4 hr/month budget — short, hands-on, in-workflow.

Behavior

What people use AI for today

% of respondents selecting each — research, content and meeting prep lead; deep workflow automation lags

Content creation
88%
Research
88%
Meeting summaries
88%
Knowledge retrieval
62%
Internal documentation
50%
Brainstorming
50%
Data analysis
50%
Client communications
25%

03 — The real problem: fragmentation

Read of the room

Everyone is building in private. Almost no one can find what already works.

Only 0% say their team uses standardized AI approaches. 50% say most people have built their own methods, and another set has no visibility into peers at all. When a colleague builds something useful, 67% find it hard or only neutral to discover and reuse — and the top worry about self-built tools is quality and accuracy (75%). This is the signature of an organization that needs a shared, governed knowledge layer — not more individual upskilling.

Maturity

How standardized is AI use on your team?

Fewer than 1 in 6 teams operate with shared approaches

1
4
3
StandardizedSome standard, variesEveryone their own wayNo visibility

Friction

Can you discover & reuse a colleague's AI work?

Difficult + neutral outweigh easy almost 2 to 1

2
1
5
1
EasyNeutralDifficult

Friction

Biggest friction points using AI today

% selecting each — inconsistent output and lack of training top the list

Lack of training
67%
Don’t trust the results
44%
Not sure which tool to use
33%
Outputs are inconsistent
33%
Hard to find reusable tools
22%
Duplicate tools solving the same problem
22%
Too many tools available
11%
Too many good ideas
11%
not sure how to execute
11%

Risk

Concerns about teammates building their own AI

% selecting each — the cost of no shared standard

Quality and accuracy
75%
Duplicate solutions
50%
Maintenance challenges
38%
Difficulty determining which solution is approved
38%
No major concerns
25%
Lack of documentation
25%
Version control issues
25%
Security risks
12%

04 — The decisive signal: what would move the needle

The question, answered

Lead with best-practice sharing & a governed AI repository. Treat education as the delivery mechanism, not the headline.

Asked to rank what would most help scale AI adoption, respondents put reusable, shared assets at the top. The four highest-ranked levers — team playbooks, a central repository of approved tools, a standardized prompt library, and best-practice documentation — average 61% top-3 selection. Stand-alone training & office hours sits well below at 33%. The signal is clear: people don't primarily need to be taught AI in the abstract — they need to find, trust, and reuse what their peers have already proven works.

Greatest value in scaling AI adoption — % ranking top 3
Central repository of approved tools
89%
Team-specific AI playbooks
67%
AI best-practice documentation
56%
Standardized prompt library
33%
Training and office hours
33%
Approved MCP integrations
0%
Internal marketplace for AI solutions
22%
Governance and approval process
0%
AI champions/community of practice
0%
Tool ratings/reviews from colleagues
0%
Repository / sharing (top 4)Other infrastructureStand-alone training
When training does happen, how should it land? — % top 3
Live workshop
78%
Self-paced modules
67%
Peer learning
44%
Quick reference guides
56%
1:1 coaching
33%
Video tutorials
22%

The preferred formats are hands-on and self-serve — live workshops and self-paced modules — not lecture courses. Education works best when it's embedded in the repository: worked examples, guided practice, office hours tied to real plays.

05 — Differences between the companies

Read of the room

Same destination, different starting lines.

The top three value levers — playbooks, a central repository, and a prompt library — are the same at every single company, which is why a generic transformation process is the right bet. But the starting points differ sharply. Appspace and StoreForce combine low governance clarity (~56-60%) with high untrained populations (60-62%) — they need the enablement engine first. Kantata, ESO and Abrigo have governance largely understood (90%+) and lower untrained rates — they're ready for the knowledge exchange now. VisiQuate is the most fragmented (50% build their own way, 0% standardized) despite small size — a sharing layer would pay off fastest there.

Benchmark

Readiness scorecard by PortCo

Green = ahead · Amber = mixed · Red = needs attention. Sorted by sample size.

PortCo n Usage /5 Underst. /5 Cust. conf /5 Gov. clarity No training Build own way Hard to discover
Appspace 65 4 3.6 3.2 60% 60% 37% 46%
Kantata 31 4.46 3.38 3.52 90% 23% 21% 36%
ESO 24 3.88 3.35 2.92 92% 12% 32% 50%
Abrigo 22 4.27 3.5 3.68 91% 23% 27% 32%
StoreForce 16 4.53 3.69 3.2 56% 62% 20% 27%
VisiQuate 9 3.33 3.56 3.22 78% 44% 50% 11%

06 — A generic maturity model

Framework

Four stages of GTM AI maturity

Every PortCo sits somewhere on this line. The transformation processes below are the bridges between stages.

Stage 1
Ad Hoc
Individuals experiment privately. No shared tools, no governance clarity, training is rare.
Stage 2
Emerging
Usage is widespread but uneven. Some guardrails understood; everyone still builds their own way.
Stage 3
Standardized
Shared playbooks, an approved repository and a prompt library are in use. Work is discoverable and reusable.
Stage 4
Scaled
AI is woven into the operating model. Continuous capture, measured outcomes, compounding advantage.
Where the surveyed PortCos sit today
Appspacecurrently at stage1→2
Kantatacurrently at stage2→3
ESOcurrently at stage2→3
Abrigocurrently at stage2→3
StoreForcecurrently at stage1→2
VisiQuatecurrently at stage2
The two bridges

Process A AI Knowledge Exchange — the sharing & repository engine. Bridges Stage 2 → 3 and sustains 3 → 4.

Process B Embedded Enablement Engine — practical, role-based capability building. Bridges Stage 1 → 2 and feeds the exchange.

07 — Two deep-dive transformation processes

Process A — Sharing / RepositoryBridges Stage 2 → 3 · sustains 3 → 4

The AI Knowledge Exchange

Why this, grounded in the data
  • Playbooks, a central repository and a prompt library are the #1, #2 and #3 ranked levers — at every company.
  • 31% build their own way and 69% can't easily discover peers' work — the exact gap a curated exchange closes.
  • The top fear about self-built tools is quality & accuracy (61%); a vetting + approval lane converts that risk into trust.
  • Usage is already high (47% daily) — there is a rich body of proven plays to harvest right now; the value is in capturing and redistributing it.
The deep-dive process
1
Harvest
Run structured 'show your best play' sessions per role to surface prompts, workflows, agents and tools already producing wins.
2
Vet & approve
A cross-functional review lane scores each asset for accuracy, security and compliance, then tags it Approved, with an owner.
3
Publish to one home
A single searchable repository — playbooks by role, prompt library, approved-tool catalog with peer ratings — replaces scattered docs.
4
Activate the community
AI champions per function curate, run office hours, and route new plays into the harvest queue so the exchange keeps filling itself.
5
Measure & prune
Track reuse, time saved and win-rate lift; retire stale assets quarterly so the repository stays trusted, not bloated.
Process B — Education / EnablementBridges Stage 1 → 2 · feeds the exchange

The Embedded Enablement Engine

Why this, grounded in the data
  • 43% have had no training in 12 months and understanding of how AI works is only 3.51/5 — a real floor to raise.
  • The #1 barrier to building AI knowledge is time, and the top friction is inconsistent output + 'lack of training' — so enablement must be short, practical and in-workflow.
  • Preferred formats are hands-on workshops and self-paced modules, delivered via an LMS and embedded in onboarding — not lecture series.
  • Respondents will give 1–4 hrs/month realistically; the program must fit that budget and tie directly to the repository's approved plays.
The deep-dive process
1
Baseline by role
Use this survey as the recurring diagnostic; set a target stage and capability bar per role (AE, SE, SDR, Marketing, RevOps).
2
Build role-based paths
Short, hands-on modules in the LMS, each anchored to an approved play from the exchange — learn the concept by running the real workflow.
3
Embed, don't bolt on
Put core modules into onboarding and Sales Kick-Off; reinforce with quarterly live workshops and standing office hours.
4
Certify & credential
Lightweight role badges (e.g. 'AI-Ready AE') signal capability and create healthy pull; managers reinforce in 1:1s.
5
Close the loop
Graduates contribute their own plays back into the Knowledge Exchange — education and sharing become one compounding flywheel.

Synthesis

The Flywheel: Learn, then Contribute Back

Why these are one system, not two separate projects

The two processes are not sequential — they compound. The Knowledge Exchange supplies the vetted plays the Enablement Engine teaches; people apply them in real work, and graduates publish new and better plays back into the Exchange. Every turn raises the floor, makes the next turn cheaper, and makes the advantage harder for competitors to copy.

LEARN Approved plays become role-based training CONTRIBUTE BACK Graduates publish new & better plays PROCESS A AI Knowledge Exchange Vetted, reusable plays live here — the shared source of truth PROCESS B Embedded Enablement Engine People build capability on real plays — then put them to work COMPOUNDING ADVANTAGE

08 — Individual PortCo recommendations

Two high-level value-creation opportunities per company, sequenced to its current stage. These are the headline bullets; the full process docs accompany this analysis.

Appspacen=65 · stage 1→2
  1. Stand up the Embedded Enablement Engine first — 60% are untrained and governance clarity is only 60%; a role-based LMS path + governance refresher closes the biggest gap in the largest team.
  2. Appoint AI champions across the 55-strong AE / SE front line to seed the prompt library with proven prospecting and demo plays, converting heavy individual usage into shared assets.
Kantatan=31 · stage 2→3
  1. Launch the AI Knowledge Exchange now — governance is understood (90%) and the team explicitly ranks 'AI champions / community of practice' top-3; formalize the community and central repository.
  2. Operationalize the stated ambition to 'lift the whole org' by having advanced users publish playbooks that pull lagging teams from Stage 2 to 3.
ESOn=24 · stage 2→3
  1. Prioritize customer-conversation enablement — customer confidence is the lowest in the group (2.92/5); build a workshop + battlecard track on discussing AI capabilities with prospects.
  2. Pair the repository rollout with a discoverability fix; 50% find peers' work hard to reuse, so a searchable approved-tool catalog is high-leverage here.
Abrigon=22 · stage 2→3
  1. Convert high tool satisfaction (4.23) and strong governance (91%) into a Stage-3 standardization push — codify the methods of the 32% already standardized into team playbooks.
  2. Use the large 'neutral' discoverability share (55%) as a quick win: a single repository home will move fence-sitters to 'easy' fast.
StoreForcen=16 · stage 1→2
  1. Address the training gap urgently — 62% untrained against the highest usage frequency (4.53/5) means heavy unguided use; deploy core LMS modules + a governance baseline.
  2. Leverage the already-strong discoverability (47% easy) by formalizing it into an approved-tool catalog before fragmentation sets in.
VisiQuaten=9 · stage 2
  1. Deploy the Knowledge Exchange as the top priority — 50% build their own way and 0% are standardized; a shared repository + prompt library will cut the most fragmentation per dollar.
  2. Build an MCP / systems-integration play (a recurring 'magic wand' ask here) into the approved-tool catalog to connect siloed systems for the GTM team.

09 — What to expect: ROI outlook

Read of the room

This pays for itself on efficiency alone — the prize is revenue, and it compounds across the portfolio.

For a value-creation audience the hero isn't time saved, it's revenue. GTM teams exist to win deals, and customer-AI confidence sits at just 3.28/5 today — so consistent, vetted plays have real room to move win rate. Framed as a share of pipeline, a balanced +2 points of win rate is ~2% of new-business pipeline in incremental ARR (e.g. $1.0M/yr on a $50.0M pipeline). Underneath it, the efficiency floor (~$3.8K per GTM user) makes the program self-funding regardless. Scaled across the portfolio, the floor alone runs into the millions before a single point of win rate is counted.

Revenue lift
2% of pipeline
+2 pts win rate (15%→17%) · e.g. $1.0M/yr per $50.0M pipeline
Self-funding floor
$3.8K/user/yr
efficiency alone — pays for the program many times over
Portfolio scale
$3.8M/yr
efficiency floor per 1,000 GTM seats — before revenue
AI-spend takeout
~75%
fewer tokens on standardizable tasks — a gross-margin lever

The prize

Revenue impact = a share of new-business pipeline

Pipeline-agnostic by design: pick your own pipeline row. Baseline win rate ~15%.

Each +1 point of win rate (e.g. 15% → 16%) converts roughly 1% of new-business pipeline into incremental ARR. A balanced +2 points — a ~13% relative lift in close rate from consistent, vetted customer-AI plays — is 2% of pipeline.

New-biz pipeline / yr+1 pt · 1%+2 pts · 2%+3 pts · 3%
$10.0M $100K $200K $300K
$25.0M $250K $500K $750K
$50.0M $500K $1.0M $1.5M
$100.0M $1.0M $2.0M $3.0M

No pipeline figure is asserted — the grid is a sensitivity range. Find your PortCo's pipeline row and read across. Reclaimed selling time and ~30% faster ramp add further capacity on top, not modeled here.

The floor

Self-funding economics — per GTM user

Efficiency alone covers the program. Built bottom-up so it's trivially defensible.

Time reclaimed · 1 hr/week × 48 × $60/hr$3K
Waste avoided · dedup tools/builds + less rework$1K
AI-spend takeout · fewer tokens per task$0K
Floor / GTM user / yr$3.8K
Floor at scale (efficiency only — revenue is on top)
6 surveyed PortCos — the 167 survey respondents alone$635K
Per 1,000 GTM seats$3.8M
Per 5,000 GTM seats (portfolio-scale)$19.0M

The floor, itemized

Efficiency value by lever — per 100 GTM users

The self-funding floor broken out. Every lever ties back to a specific survey signal, not assumed in a vacuum.

Value leverSurvey signal it draws on ConservativeExpectedUpside
Reuse over rebuild
Vetted playbooks & prompt library replace per-person trial-and-error.
#1 barrier is time; ~62% can't easily find peers' work $72K $144K $252K
Faster onboarding
New hires reach productive AI use sooner via embedded modules.
40% untrained today; preferred delivery is onboarding + LMS $25K $60K $110K
Tool & build de-duplication
One approved-tool catalog ends redundant licenses & parallel builds.
36% see duplicate solutions; 15% duplicate tools $15K $40K $90K
Quality & rework
Approved, rated assets cut inconsistent output and re-dos.
42% cite inconsistent outputs; 35% don't trust results $20K $55K $120K
Efficiency floor — per 100 GTM users / yr $132K $299K $572K

This is the floor only — it deliberately excludes the revenue lever above (win-rate + capacity), which is larger but more assumption-sensitive. Together they make the case both ways: self-funding on efficiency, transformative on revenue.

To turn these into your numbers, confirm three inputs
1 · GTM headcount per PortCo and across the portfolio (model currently floors at the 167 survey respondents — actual GTM is larger).   2 · Annual new-business pipeline per PortCo — read it off the grid above; win rate is modeled as a share of pipeline (baseline ~15%), so no pipeline figure is hard-coded.   3 · Fully-loaded labor rate (placeholder: $60/hr).

Balanced posture: the efficiency floor is grounded in survey signals; the revenue lever is expressed as a % of pipeline so nothing financial is invented. Win-rate lift is the largest and most assumption-sensitive lever — shown as a 1–3 point range, not a point estimate.

10 — Token economics: standardization as a cost lever

Read of the room

Standardization and enablement aren't just adoption levers — they're a direct way to conserve tokens (and spend).

Most token waste in a GTM org isn't from too much usage — it's from unguided usage: the same prompt re-attempted 4–5 times, context re-pasted every session, the same workflow rebuilt and re-run by five teams, and defensive re-runs because nobody trusts the first output (35% say exactly that). Every one of those is removed by a vetted prompt library, an approved repository, and shared playbooks. The enablement engine compounds it by teaching people to prompt and model-select well the first time.

Fewer attempts
4→1
ad-hoc trial-and-error vs. a vetted prompt
Token reduction
~75%
on a common task class, once standardized
Tokens saved / task class
15M/mo
at 1,000 runs/mo · illustrative
Re-runs avoided
↓ defensive retries
35% don't trust results today and re-run

Mechanisms

How the engines conserve tokens

Five concrete paths from fragmentation to fewer, cheaper calls

Kill trial-and-error
An ad-hoc prompt often takes 3–5 tries to get a usable answer; a vetted, library prompt lands it in one. That's the single biggest token sink, removed.
Standardized prompt library
Stop re-pasting context
An approved repository / RAG layer means source docs and context aren't re-stuffed into the model every session.
Central repository
End duplicate runs
Standardization stops N teams independently running the same expensive agent workflow in parallel.
Team playbooks + approved tools
Right-size the model
Best-practice docs steer users to the right tool/model for the job — no frontier-model spend on cheap tasks.
Best-practice documentation
Cut defensive retries
Vetted, peer-rated outputs reduce the re-runs people do today because they don't trust the first result.
Tool ratings + approval lane

Illustration

One common task class, org-wide

A worked example of the trial-and-error sink

Runs of one task / month1,000
Ad-hoc: 4 attempts × 5K tok20M tok/mo
Vetted prompt: 1 attempt × 5K tok5M tok/mo
Saved on this task class alone15M tok/mo (~75%)

Illustrative figures for a single task class; multiply across the dozens of repeatable tasks a GTM team runs. The point is directional: the savings scale with how repeatable and standardizable the work is — which is most of it.

AKKR Ascend Capstone · GTM Functional Group · AI Readiness Survey (168 responses) Generated from live survey data · 6 portfolio companies