GTM & AI — A Field Report

Research

Brief

№ 01

The question this study set out to answer.

Front matter for the field report — who the ten archetypes represent, what we wanted to learn, and the boundaries of the study. Skip ahead if you're here for the findings.

i. The Audience

Who the archetypes represent

GTM professionals at mid-market companies — VP Marketing & CMO, Head of Growth, Chief Revenue Officer, Demand Gen, Product Marketing, Sales Enablement, Content Marketing, Lifecycle/Retention Marketing, and Customer Success.

Geography: United States & Europe
Industries: B2B SaaS · HealthTech · FinTech
Revenue band: $20M – $200M ARR
Headcount: 150 – 1,500 employees
Tech stack: Salesforce, HubSpot

ii. The Goal

What we wanted to learn

How do GTM professionals determine the role of AI in achieving their goals — and where does it break down?

Strategic clarityknowing where AI belongs in the GTM motion
Process & workflowhow AI fits into existing systems and routines
Quality of outputswhen output is usable, when it isn't, and why
Trust in AIwhat's earned, what's withheld, and on what terms
Behavioral & change managementhow teams actually adopt new tools

Across the full arc: progress made, progress still to come, and what governs the difference.

iii. The Method

10 archetypes, OCEAN-based

Ten behavioral archetypes were constructed using the OCEAN framework (Openness, Conscientiousness, Extraversion, Agreeableness, Neuroticism). Each archetype was matched to a real GTM role, industry, and company profile within the audience definition above.

Each archetype was then interviewed in depth using a dynamic interview protocol — the same kind of probing follow-up that drives a 1:1 conversation, applied to a synthetic persona that represents the behavioral signature of hundreds of real people who fit that profile.

The interviews were conducted on the Synthetic Users platform, which builds each participant on an individual five-factor (OCEAN) personality model — the basis for the ten archetypes below.

→ See §04 for the ten archetypes and their OCEAN profiles.

§ 01 · The Six Findings

What everyone, eventually, agrees on.

Six themes surface across roles, geographies, and industries — together they describe AI not as a technology rollout, but as a socio-technical change effort that succeeds or fails on operational credibility.

Augmentation, never replacement — even by enthusiasts.

In customer success, growth, sales enablement, and revenue alike, the credible framing was always human-plus-machine. Even aggressive adopters refused to sell automation as a substitute for judgment.

AI was there to augment their work, not replace it. Liam · Head of Growth · FinTech

Narrow, measurable use cases beat sweeping ambition.

Lead scoring, churn prediction, ticket triage, knowledge suggestions, forecasting. The wins were unglamorous and tied to specific KPIs — Marc moved MQL→SQL conversion from 12% to 23% by scoring above a hard threshold.

Within about six weeks, our conversion rate jumped to 23%. The sales team was actually happy for once. Marc · Demand Gen Lead · B2B SaaS

Data quality is the hidden constraint behind every AI initiative.

Fragmented Salesforce/HubSpot data, duplicates, standardization debt — these consistently outweighed model sophistication as the actual blocker. Evelyn called the discovery "a bit of a wake-up call."

Without quality data, the AI's output wasn't as effective. It was a bit of a wake-up call. Evelyn · VP Marketing · B2B SaaS

Trust is task-dependent, and earned through review loops.

AI was accepted faster for summarization, triage, and first drafts — and treated cautiously for strategy, compliance, and high-stakes content. Claire could trust it to tell her what a competitor said, but not why.

The AI just processed the words. It didn't understand the real-world limitations. Claire · Product Marketing Manager · FinTech

Skeptics, often, are the ones who make the system work.

Sarah, the skeptical marketing ops manager on Marc's team, pushed for monthly model audits — which became the mechanism that fixed an enterprise bias the original rollout had baked in. Resistance was a quality-control mechanism.

I'd rather have a slightly less accurate model that I can actually troubleshoot than some cutting-edge neural network that's a complete mystery. Marc · Demand Gen Lead · B2B SaaS

Speed is not the same as efficiency.

AI accelerated first drafts and analysis, but total cycle time often grew because of verification, retraining, and exception handling. Anya's pragmatic test: if fixing the output takes longer than writing it, the AI isn't serving its purpose.

If fixing the AI's output is going to take longer than just writing it ourselves, then the AI isn't really serving its purpose. Anya · Content Marketing Lead · HealthTech

§ 02 · Core Tensions

Adoption is shaped by the contradictions, not the technology.

The interesting story isn't whether teams want AI — most do. It's how they negotiate four structural tensions inside real work. Each one surfaced repeatedly across archetypes; the quotes below are the moments where they became visible.

Tension · 01

Efficiency vs. quality

Teams want AI to be fast, but fast outputs that need heavy verification can erase the gain entirely. The practical question is where verification cost crosses what was saved at the front end.

If a team member spends two hours prompting the AI, and then four hours meticulously verifying and heavily rewriting a draft that should have only taken two hours to refine, then we're actually losing efficiency. Anya · Content Marketing · HealthTech

The tension arises when there's this expectation to just 'run with' the AI's initial output because it was so fast. I sometimes have to push back. Claire · Product Marketing · FinTech

Tension · 02

Augmentation vs. replacement anxiety

Leaders frame AI as support. Privately, employees ask whether their role still has value. The public objection is usually about accuracy; the private one is about identity.

I think I underestimated how much people would feel threatened by it, you know? Marc · Demand Gen · B2B SaaS

In one-on-one sessions, team members often voiced concerns about their job security, wondering if AI might replace their roles. Liam · Head of Growth · FinTech

Tension · 03

Innovation vs. reliability

Growth-minded teams chase breakthroughs; regulated teams demand traceability. The HealthTech and FinTech threshold for ambiguity is dramatically lower than B2B SaaS.

For HealthTech, especially with regulations, that's a huge problem. You can't put out information that's even slightly misleading, let alone incorrect. It damages trust, and frankly, it's irresponsible. Anya · Content Marketing · HealthTech

Balancing innovation with compliance and reliability starts with setting a solid foundation. Sofia · Chief Revenue Officer · HealthTech

Tension · 04

Local optimization vs. systemic readiness

A promising use case can still fail if data infrastructure, training, or integration is weak. The model is rarely the bottleneck; the plumbing around it usually is.

Without quality data, the AI's output wasn't as effective. A bit of a wake-up call. Evelyn · VP Marketing · B2B SaaS

Our data was spread across multiple platforms, which made it tough to ensure everything was aligned for the AI tool to provide accurate insights. Robert · Director of Revenue Operations · FinTech

§ 03 · The AI Trust Modes

Trust isn't where you are. It's how you're operating right now.

A behavioral + Jobs-to-be-Done lens on the six distinct modes GTM teams operate in when working with AI. Not a ladder you climb — a set of postures you switch between depending on the task, the stakes, and the failure modes you've encountered. Most teams operate in multiple modes at once, and the same person regularly moves between them.

Trust isn't fixed — it oscillates Each archetype operates in multiple modes depending on the task. The bar shows their full mode range; the filled mark shows their primary mode. Pay attention to the gap between primary and fallback — that gap is where AI adoption actually lives or dies.

M1M2M3M4M5M6

Claire · Product Marketing

Trusts AI for raw scraping; drops to verify when strategic interpretation is at stake

Marc · Demand Gen

Pushed to M5 with autonomous lead scoring, then retreated to M2 after bias surfaced

Anya · Content Marketing

M3–M4 with templates & brand-voice examples; M2 for regulated HealthTech content

Isabella · Customer Success

M4 for ticket triage today; cautious M2 for future predictive analysis

Liam · Head of Growth

Personally at M6 — but his retention team initially sat at M1–M2, skeptical and double-checking

Evelyn · VP Marketing

M4 today; dropped to M2 during initial rollout when data quality issues surfaced

Sofia · Chief Revenue Officer

M4–M5 for daily lead scoring; M2 for any new tool because of HealthTech compliance

Robert · RevOps

M4 with Salesforce-native tools today; self-described late adopter, started at M1–M2

Oliver · Sales Enablement

M4 for rep coaching; reps drop to M2–M3 when AI timing is wrong

Daniel · Lifecycle Marketing

M4 for churn models; M2 oversight for explainability and GDPR compliance

Mode range — modes they operate in Primary operating mode Where they fell back from / had to start

★

The pattern: Every archetype operates in a range of modes, not a fixed mode. Even Liam — who personally works in Strategic mode — leads a team that started in Exploration. Even Marc — who pushed his pipeline into Autonomous mode — switched back to Assisted Use when bias surfaced. Trust isn't directional. The most important question isn't "what mode are you in" — it's "what conditions cause you to switch modes, and which mode fits the work in front of you?"

Mode types

Exploration mode

Stall mode

Value mode

Strategic mode

Mode 01 Exploration

Awareness — curiosity without commitment

"This looks interesting, but I'm not relying on it."

"Hype, risky, not for me" → "Credible, safe, worth trying"

Behavior: Observing, testing casually
Use: Low-stakes exploration — ideas, drafts, summaries
Trust driver: Intrigue and perceived potential
Barrier: "Is this just hype?"
Psychological shift: Legitimacy & safety — early stage

No archetype primarily operates in Exploration mode — every GTM leader interviewed had already engaged with AI on real work. But Liam's retention team started here, skeptical and worried about job security, before they could be brought along. The leader's posture is rarely the team's posture.

⚠ The default-return mode

Mode 02 Stall zone

Assisted Use — human in the loop

"I'll use it, but I don't trust it without checking everything."

"I don't see how this fits my work" → "Useful — but I need to verify everything"

Behavior: AI generates, human verifies and edits
Use: Drafting, brainstorming, first-pass outputs
Trust driver: Speed and visible utility
Barrier: Verification cost — directly tied to the Efficiency vs. Quality tension
Psychological shift: Relevance & fit — the mode teams keep returning to

Who lands here Almost everyone — at some point Claire & Anya operate here for strategic interpretation. Marc & Sarah switched back here after autonomous mode bias surfaced. Evelyn dropped here during rollout. Sofia & Daniel sit here for compliance work. Liam's team started here. Assisted Use is where everyone passes through.

I still had to go in and verify some things, add my own strategic interpretation. AI is great for speed, but it's not a human. The AI just processed the words. Claire · Product Marketing · FinTech

If fixing the AI's output is going to take longer than just writing it ourselves, then the AI isn't really serving its purpose. We just decided to scrap it. Anya · Content Marketing · HealthTech

Mode 03 Stall zone

Delegated Tasks — conditional trust

"I trust it for specific tasks under defined conditions."

"I have to check everything it does" → "I trust it for specific tasks under clear conditions"

Behavior: Partial delegation, repeatable tasks
Use: Formatting, summarization, data extraction, templated work
Trust driver: Consistency and predictability
Barrier: Edge cases and failure anxiety
Psychological shift: Competence & confidence

Delegated Tasks is more transition than destination. Claire trusts AI in this mode for raw competitive intel scraping. Anya's team operates here when feeding AI strong brand-voice templates. Oliver's reps switch here when AI timing recommendations need cross-referencing. Conditional trust without full integration.

★ Where ROI lives

Mode 04 Value zone

Integrated Workflow — embedded trust

"This is part of how we work now."

"A helpful tool I occasionally use" → "Part of how I consistently get work done"

Behavior: AI embedded in workflows and tools
Use: Multi-step processes, cross-functional workflows
Trust driver: Reliability within system context
Barrier: Integration gaps — directly tied to the Local Optimization vs. Systemic Readiness tension
Psychological shift: Immediate personal value — the mode where ROI actually shows up

Primary mode for Evelyn · Sofia · Robert · Oliver · Isabella · Daniel · Anya The integration cluster — but every one of them also operates in Assisted Use mode for new tools, compliance work, or when failure modes surface. Integrated Workflow is where they aspire to spend most of their time; it's not where they spend all of it.

Integrating AI tools into our existing tech stack, like Salesforce and HubSpot, was a bit trickier than planned. It was important to make sure everything communicated seamlessly to avoid data silos. Evelyn · VP Marketing · B2B SaaS

I prefer leveraging enterprise Salesforce-native solutions because they tend to fit seamlessly with our existing tech stack, making the transition smoother. Oliver · Sales Enablement · B2B SaaS

I had to collaborate closely with both the tech and customer service teams to refine the algorithms… balancing the tech's capabilities with the human touch remains a daily consideration. Isabella · Customer Success · B2B SaaS

Working with the data science team, we used historical data to build models that could predict churn… personalized customer journeys and improved retention metrics. Daniel · Lifecycle Marketing · FinTech

Mode 05 Value zone

Autonomous Execution — outcome trust

"I trust it to act, not just assist."

"I need to stay in the loop" → "I trust it to execute and move work forward"

Behavior: AI executes with minimal oversight
Use: Agents, automation, decision support
Trust driver: Proven outcomes and recoverability
Barrier: Risk, irreversibility, accountability
Psychological shift: Progress proof + safety nets

Reaches here Marc · Sofia (partial) Marc pushed his pipeline workflow into Autonomous mode with lead routing — then switched back to Assisted Use when bias surfaced. Sofia's daily lead scoring runs in this mode, but new tools always begin in Assisted Use vetting first. Autonomous Execution is reachable, but rarely sustained without scaffolding.

We started routing only the leads with scores above 75 to immediate sales follow-up, and anything below that went into nurture sequences. Within about six weeks, our conversion rate jumped to 23%. Marc · Demand Gen · B2B SaaS

Some of the highest-scoring leads were still duds. The AI was good at pattern recognition, but it couldn't predict when someone's budget got slashed or priorities shifted… I think I underestimated how much people would feel threatened by it. Marc · Demand Gen · B2B SaaS

Mode 06 Strategic

Strategic Dependence — indispensable

"We couldn't operate the same way without it."

"This improves how we work" → "We rely on this to operate and compete"

Behavior: AI shapes strategy, not just execution
Use: Forecasting, decision-making, growth loops
Trust driver: Compounding value and competitive advantage
Barrier: Organizational readiness and identity shift
Psychological shift: Next-cycle commitment — "must-have" territory

Reaches here Liam · personally Liam personally operates in Strategic mode — using ML to uncover hidden behavior patterns that shape strategy itself. But his retention team initially worked in Exploration and Assisted Use modes, skeptical and double-checking every prediction. Strategic dependence at the leader level doesn't mean the team is operating there yet.

It's all about using AI to find those breakthrough growth opportunities, so I'm constantly testing and swapping vendors to see what works best. Liam · Head of Growth · FinTech

The model identified subtle behaviors that were early indicators of potential churn — patterns we hadn't considered relevant before. AI can unearth hidden insights that we'd have missed with traditional methods. Liam · Head of Growth · FinTech

§ 05 · The Arc

From foundations to useful to adaptive.

A storyline, not a timeline. Most teams aren't on a fixed schedule — they're moving through three phases. Right now, they're laying foundations and fighting manual work. Near-term, they want AI that's reliably useful — predictable, explainable, integrated. Long-term, they imagine adaptive GTM systems where the role itself evolves. Each phase carries its own themes, each anchored to a named archetype.

i. Right Now

Laying the foundations

Most archetypes describe this phase as unglamorous and necessary. The work is plumbing, not personalization.

Foundations

Data quality, governance, role-based enablement — the prerequisites that decide whether anything else works.

Daniel — auditing fragmented HubSpot/Salesforce data, building centralized pipelines
Evelyn — automated cleansing routines and cross-functional sprints to prevent data silos
Robert — clean pipelines with IT before forecasting models can be trusted

Override & oversight rules

Where AI is already in production, teams are spending real time managing the failure modes — not coasting.

Marc — manual model tuning, override variables, monthly audits to catch enterprise bias
Anya — meticulous fact-checking and prompt-tuning for HealthTech accuracy
Sofia — legal review and compliance gating for every new tool

Narrow, measurable use cases

Ticket triage, lead scoring, churn prediction — concrete jobs with concrete KPIs, already producing results.

Isabella — AI triaging tickets by urgency, refined through cross-functional workshops
Liam — ML surfacing in-app engagement drops as early churn signals
Oliver — Salesforce-native playbooks and coaching insights for reps

Claire as the daily assistant

For some, AI is already a workhorse for raw data processing — but always with human verification on top.

Claire — AI for competitive intel summarization, but manual deep-dive on pricing and reviews to catch market subtext

ii. Near-Term

Make AI reliably useful

The middle phase is about converting current scaffolding into dependable production capability. Less heroics, more reliability.

Confidence in prediction

Churn models, lead scoring, forecasting that teams can act on without endless tuning.

Marc — confident pipeline prediction for both volume and quality
Sofia — proactive churn prediction to drive retention plays
Daniel — segmentation refinement, smarter touchpoint timing, tighter messaging

Less manual work

Cut reporting, sorting, triage, and first-draft burden so people can focus on customers.

Marc — AI flagging anomalies before they become firefighting
Anya — granular prompts and standardized editing to kill over-verification
Evelyn — feedback loops that reduce data entry errors at the source

Explainability & integration

Models that fit Salesforce / HubSpot cleanly and produce signals humans can audit.

Robert — advanced predictive analytics + robust governance and monitoring
Liam — team data literacy so reps can integrate AI insight with traditional metrics
Sofia — automated forecasting that anticipates market trends

Personalization, refined

Deeper tailoring of customer interactions — not yet "true personalization," but a real step toward it.

Isabella — personalized customer insights, real-time self-service suggestions, predictive analysis
Sofia — deeper personalization in client interactions
Oliver — reps applying AI to refine pitches and adjust follow-up timing

iii. Long-Term

From assistive tools to adaptive systems

The far horizon is more uncertain — and more transformative. AI stops being a feature and becomes infrastructure for how GTM works.

Autonomous optimization

Campaign tuning, content adaptation, lead routing happening in the background.

Marc — AI handling all operational tasks so he can focus entirely on strategy
Sofia — AI chatbots for support, AI-driven personalized content marketing
Robert — agile RevOps framework with team freed from routine admin

Earlier buyer insight

Spotting intent and risk before traditional signals show. Proactive over reactive.

Marc — predicting buyer intent before the buyer realizes it themselves
Liam — ML uncovering breakthrough growth patterns traditional analysis misses
Daniel — predictive models expanding into upsell and cross-sell, not just retention

True personalization at scale

Individualized journeys — but only if data foundations and trust safeguards mature alongside the ambition.

Anya — storytelling efficiency at scale across all content types
Evelyn — clean data continuously powering accurate AI personalization
Daniel — explainability and GDPR scaling alongside model sophistication

Role evolution

GTM teams becoming more analytical, supervisory, cross-functional. Less doing, more orchestrating.

Isabella — team becoming more strategic and proactive, actively shaping AI strategy
Marc — demand gen evolving into "AI prompt engineering"
Evelyn — team shifting fully to strategic initiatives, troubleshooting becomes background

§ 06 · For Operators

Five plays that actually worked.

Not vendor promises. Not best practices. The five moves participants made that changed how AI adoption played out for their teams — each one anchored to a named GTM leader who lived through it.

Start narrow & measurable

Lead scoring, churn prediction, ticket triage, draft generation. Define success in business terms — conversion rate, retention, response time — before picking the tool.

By implementing AI, we were able to automate this process, ensuring that our sales team could concentrate on leads that had the highest potential for conversion. Sofia · Chief Revenue Officer · HealthTech

We wanted to identify customers at risk of leaving before it actually happened… seeing our efforts translated into more personalized customer journeys and improved retention metrics was rewarding. Daniel · Lifecycle Marketing · FinTech

ii.

Invest in data readiness first

Every durable success rested on cleaner data, better integration, and clearer governance. It's part of the AI program, not a prerequisite to defer.

We quickly realized that the effectiveness of AI heavily depended on clean and well-organized data. Without reliable data, any AI-driven personalization would have been hit-or-miss. Evelyn · VP Marketing · B2B SaaS

Our data was spread across multiple platforms, which made it tough to ensure everything was aligned for the AI tool to provide accurate insights. Robert · Director of Revenue Operations · FinTech

iii.

Design review & override rules

Explicit human-in-the-loop guidance for when AI output ships, when it needs review, and who owns exceptions. Critical in regulated and customer-facing work.

The biggest change we made was adding what I called 'override variables.' So if a lead hit certain criteria, it would get a boost regardless of company size. That helped balance out some of the enterprise bias. Marc · Demand Gen · B2B SaaS

If I look at an AI draft and my gut tells me that correcting all the issues is going to take 70% or more of the time it would take a human writer to just do it from scratch, then we scrap it. Anya · Content Marketing · HealthTech

iv.

Treat adoption as change management

Pilots, role-based training, regular check-ins, visible quick wins. Bring skeptics in early — they improve the system more than they slow it.

I literally brought Sarah [the biggest skeptic] into the vendor demos. Not because I needed her permission, but because I learned that if she pokes holes in it upfront, it's way better than having her undermine it after we've already committed. Marc · Demand Gen · B2B SaaS

By involving them in refining the process, they became more open to working with the AI and trusted it as an effective tool. This collaboration also strengthened our team's morale and cohesiveness. Isabella · Customer Success · B2B SaaS

Optimize for explainability

Where decisions need to be defended cross-functionally or legally, an interpretable model beats a more accurate black box every time.

That black box thing? Never again. I spent too many nights trying to explain to my CEO why a 'high-quality' lead didn't convert, and all I could say was 'well, the algorithm thought it was good.' I'd rather have a slightly less accurate model I can actually troubleshoot. Marc · Demand Gen · B2B SaaS

Making sure stakeholders — and even customers — understand and trust AI-driven decisions can be tricky. We need to keep AI transparent and accessible to various audiences. Daniel · Lifecycle Marketing · FinTech

§ 07 · Work With Us

Making a high-stakes GTM decision with incomplete customer insight?

Match the method to the stakes of the decision — not the other way around. We run B2B Customer Intelligence Sprints in three modes, each producing a fundamentally different kind of evidence.

Choose based on your situation

Need fast direction → Synthetic Need real buyer validation → AI-Moderated Need executive-level certainty → 1:1 Expert

Different unit of analysis → answers different questions.

Synthetic User Research interviews validated behavioral archetypes — each interview reflects the distribution of hundreds of real buyers who fit that profile.

AI-Moderated Interviews talk to real buyers at scale with consistent moderation. 1:1 Expert Interviews deliver one human's perspective, deeply probed.

Method · 01Fastest

Synthetic User Research

Simulated interviews with validated behavioral archetypes. Each interview reflects hundreds of real buyers. This is the method behind this report, run on the Synthetic Users platform.

What 1 interview =: An archetype · population-level signal
Best for: Directional decisions — buyer clarity, persona shape, messaging tests, pre-investment alignment
Not for: Final decision validation or compliance-sensitive sign-off
Timeline: 7–10 days

Discovery call & research design
Single or multi-industry focus
4–8 OCEAN-based persona interviews
Full transcripts & findings report
GTM Motion refinement (optional)

Get directional insight in 10 days $3,500 – $7,500 depending on scope

Method · 02Balanced

AI-Moderated Interviews

Real buyers, structured discovery, AI moderation. Each interview = one real buyer — at scale, with consistent methodology across the panel.

What 1 interview =: One real buyer · sample-level validation
Best for: Validated decisions — confirming hypotheses with real customers, broader sample sizes, repeatable studies
Not for: Deeply nuanced executive positioning or causal "why" questions
Timeline: 2–3 weeks typical

Custom research design & recruitment criteria
AI-moderated 1:1 interviews at scale
Full transcripts & thematic synthesis
Findings report with verbatim quotes
Strategic readout session

Validate with real buyers at scale Custom scoped

Method · 03Deepest

1:1 Expert Interviews

Human-led, deeply probed conversations with senior buyers. One interview = one person, fully explored — irreplaceable when nuance and follow-up probing matter most.

What 1 interview =: One real buyer · individual-level depth
Best for: Irreversible / high-stakes decisions — positioning, regulated buyers, M&A diligence, executive nuance
Not for: Early-stage hypothesis generation or quick directional clarity
Timeline: 4–6 weeks typical

Recruitment of vetted senior buyers
Expert-led 1:1 interviews with deep probing
Verbatim transcripts & coded analysis
Strategic findings & positioning recommendations
Executive presentation & workshop

De-risk your highest-stakes decision Custom scoped