Inside Finance’s AI Decision
A cross-sectional study from Pulley.
We surveyed 89 senior finance leaders across mid-market and enterprise companies on what's actually working with AI and what isn't.
Hint: The gap between “we use AI” and “AI works for us” is where most finance teams are stuck.
What you'll find inside:
The adoption gap: 70% of finance teams have deployed AI, only 10% have it deeply embedded
The 5 blockers slowing AI transformation, and which one most finance teams underestimate
Our Field Guide: A practical playbook for CFOs and next steps from the data
Key findings from the report
1. AI adoption is broad but shallow.
70% of finance teams have at least one AI tool in pilot or production. Only 10% have AI deeply embedded across multiple workflows. Deployment has arrived, but full-scale AI transformation has not.
2. Finance teams are scaling AI spend on faith, not evidence.
The function responsible for measuring everything is exempting its own AI investments from measurement rigor.
3. Internal readiness is a bigger constraint than vendor capability.
Data quality, process immaturity, and connectivity gaps are the real blockers to AI adoption.
4. Governance hasn’t kept pace with deployment.
Almost half of the respondents have no written AI policy or rely solely on informal norms. Teams are running AI in production while policies are still being drafted.
5. The market has articulated the AI product it wants.
Cross-system interoperability is what finance leaders want most. It's also what they're least prepared to deploy.
What's inside the report
What mid-market finance teams are actually doing with AI. Workflow-by-workflow adoption map. Where AI is in production, pilot, research, or nowhere.
The ROI question. Who's measuring, what they're measuring, what the numbers actually say. ROI believers vs. skeptics.
Tool sprawl and the ecosystem problem. Median tool count for early-stage versus AI-mature teams (7–9+). Why connectivity, not consolidation, is the unlock.
AI and the shape of the finance team. The skills that are gaining priority. The pullback that is and isn't happening.
The data readiness gap. Team capacity and data quality are the two dominant barriers. The constraint is organizational, not technological.
The ROI question. Who's measuring, what they're measuring, what the numbers actually say. ROI believers vs. skeptics.
Risk, governance, and trust. Hallucinated outputs in financial reporting rank as the #1 weighted risk.
What's next: The 12–18 month outlook. Spend plans, posture preferences, and the cross-system unlock. The market is ready to spend, but measurement and integration infrastructure isn't keeping pace.
The field guide: 5 plays for CFOs. Get the practical playbook and next steps pulled from the data.
Real respondent quotes
"We don’t have enough people who understand both finance and AI to evaluate what’s real versus hype."
VP of Finance, enterprise SaaS
"On a day-to-day basis, AI has enabled me to translate financial data and reporting into digestible, conveyable formats conducive to my non-finance business partners."
CFO, mid-market e-commerce
"I’ll let AI draft the variance commentary, but I’m not letting it send an email to the CEO without my review.""On a day-to-day basis, AI has enabled me to translate financial data and reporting into digestible, conveyable formats conducive to my non-finance business partners."
VP of Finance, enterprise healthcare
