![Generic Blog Visual Calculator](data:image/svg+xml,%3Csvg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 750 393"%3E%3C/svg%3E)
Summit 2025 Recap #2: Measuring AI ROI and Driving Transformation — Insights from Zapier and ClickUp CFOs
Welcome to the CFO Connect Summit 2025 Recap Series — your go-to source for key insights and actionable takeaways from 2025’s most impactful finance leadership sessions.
On October 07, 2025, nearly 500 finance leaders joined the CFO Connect Summit to explore one of the biggest challenges in modern finance: how to measure AI ROI and drive transformation without creating another SaaS sprawl.
The session, “Measuring AI ROI and Driving Transformation,” featured Dan Zhang, CFO at ClickUp, and Ryan Roccon, CFO at Zapier — two leaders building finance teams and AI products in parallel. Both offered a candid view of what actually works when scaling AI across an organization.
The discussion revealed a clear reality: the challenge isn’t finding AI tools, but managing too many. ClickUp cut costs by consolidating five AI note-takers into one native solution, while Zapier built governance around platform use to balance experimentation with control. Both CFOs shared practical frameworks to measure ROI beyond efficiency — from Dan’s three-bucket model (automation, capability unlocks, and quality improvements) to Ryan’s outcome-based pricing lens, which ties AI value directly to measurable results.
A key insight: real transformation requires mandates, not memos. At Zapier, nearly 98% of employees now use AI thanks to a mix of bottom-up experimentation and top-down expectations. Both CFOs stressed the importance of aligning communication — celebrating speed gains with teams, linking impact to OKRs for leadership, and demonstrating long-term leverage to boards.
Moderated by Lucile Cornet, Partner at Eight Roads, this session gave finance leaders a concrete roadmap to measure AI’s true impact and turn automation experiments into sustainable business transformation.
Speaker Credentials
Dan Zhang, CFO at ClickUp – Oversees finance, accounting, legal, and HR at the all-in-one work management platform. Formerly at Amazon and Expedia, Dan brings a decade of experience scaling SaaS finance operations and building AI-powered collaboration tools.
Ryan Roccon, CFO at Zapier – Leads finance, data, business operations, and pricing at the AI orchestration platform connecting 8,000+ apps. Having built Zapier’s finance systems from scratch, Ryan shared lessons from scaling automation to enterprise AI agents.
Key Session Themes & Takeaways
Theme 1: The Three-Bucket Framework for AI ROI
Dan introduced a clear framework for categorizing AI investments:
One-to-ten automation – Time savings from automating repetitive tasks.
Zero-to-one unlock – New capabilities previously impossible (e.g., reaching every customer at scale).
C-to-A quality boost – Better insights and decision-making, not just faster outputs.
Key Insight: “If we only measure AI by time saved, we’ll miss its real value — capabilities and quality,” said Dan. Traditional ROI models often fail to capture AI’s transformative potential.
Practical Application: Classify each AI use case by these three buckets. Use time-savings metrics for automation, but assess strategic or quality gains for the latter two.
Theme 2: Strategic Alignment Over Shiny Demos
Both CFOs stressed that AI ROI depends on alignment with company priorities, not hype. Dan described the CFO’s role as “holding the map” — ensuring every AI project drives growth, efficiency, or operational excellence.
He compared two tools: one automating sales plans (efficiency) and one using AI to engage the long-tail of small customers (capability unlock). The second, though harder to quantify, drives far greater impact. Ryan added that many AI vendors now price on outcomes — such as resolution rates — making ROI clearer for cost-saving tools but more nuanced for strategic ones.
Framework: Before approving AI budgets, ask: “Is this a Holy Grail, Invisible Platform, Copilot, or Sandbox?”
Holy Grail: high autonomy and attribution — long-term investment.
Copilot: low autonomy, high attribution — short contracts, multi-vendor testing.
Theme 3: Consolidation Over Proliferation
Dan noted that ClickUp once paid for five AI note-takers before building one internally — a classic case of tool overload. The lesson: check whether platform vendors already offer similar capabilities before adding point solutions.
Ryan echoed this with Zapier’s principle of model independence — ensuring flexibility to work with OpenAI, Google, or Claude, switching based on performance or cost.
Practical Application: Before buying new AI tools, verify existing vendor roadmaps. Account for hidden costs like onboarding, compliance, and integration. As Dan put it, “There’s no such thing as turnkey when deploying to hundreds of thousands of users.”
Theme 4: Organizational Mandate and Cultural Transformation
Both CFOs agreed: AI transformation requires mandates, not nudges. At Zapier, nearly 98% of employees use AI — achieved through bottom-up experimentation combined with top-down expectations.
Communication differs by audience:
Employees: celebrate quick wins (“What took 30 minutes now takes 30 seconds”).
Leaders: connect progress to OKRs.
Boards: show multi-year operating leverage — how AI reduces costs across revenue functions.
Dan shared that ClickUp’s cadence includes weekly AI highlights, monthly awards, and quarterly hackathons — reinforcing that “AI is not optional; AI is required.”
Hiring Lens: Both companies now assess candidates’ AI fluency. At ClickUp, case studies include permission to use AI, then debriefing prompts and model choices. “Your first question might fail — we care how you iterate,” said Dan.
Practical Implications for Finance Leaders
Set Department-Level AI Mandates: Assign every function clear AI transformation goals and appoint an AI lead to track progress across teams.
Apply the Four-Quadrant Framework: Use Dan’s autonomy-attribution model to define contract terms and budgets by AI category.
Create Lightweight A/B Tests: Compare metrics — like close rates or cycle times — between AI users and non-users to prove value early.
Tailor Communication Cadences: Share wins weekly with teams, OKR progress monthly with leaders, and efficiency gains quarterly with the board.
Showcase Internal Wins: Highlight employee-built automations to boost morale and inspire further innovation.
Challenges & Pitfalls to Watch
Vendor Lock-In: Avoid long contracts for fast-evolving copilot tools.
Premature ROI Tracking: Early AI gains may be intangible — allow time for scale.
Shadow AI Risks: Implement secure, enterprise-approved AI tools to protect data.
Automation ≠ Transformation: Tools amplify clarity, not chaos — prioritize strategic alignment.
Board Miscommunication: Frame AI ROI in terms of multi-year leverage, not short-term wins.
Key Quotes & Sound-Bites
“The challenge isn’t finding AI tools — it’s having too many.” — Dan Zhang, CFO, ClickUp
“Transformation doesn’t happen through helpful nudges but mandates. At Zapier, 98% of staff use AI because it’s expected.” — Ryan Roccon, CFO, Zapier
“If we only measure AI by time saved, we miss the real value — capabilities and quality.” — Dan Zhang, ClickUp
“AI metrics now include outcome-based pricing — resolution rates and AI-routed leads — making ROI easier to calculate.” — Ryan Roccon, Zapier
“If your house has no design theme, the best furniture won’t fix it. Same with AI — clarity comes before tools.” — Dan Zhang, ClickUp
Real-World Applications Shared
ClickUp’s Finance Knowledge Agent: Combines competitive insights with internal data to democratize investor-level intelligence across the finance team.
Zapier’s Virtual Mailbox Automation: AI now processes scanned mail from 10 entities, classifies content, and posts summaries to Slack — fully automated.
ClickUp’s Native AI Note-Taker: Replaced five third-party tools, reducing cost and now resold to customers.
Zapier’s Engineering Analytics: Benchmarked AI copilot usage by developer productivity, pinpointing where AI adds the most value.
ClickUp’s Accounting Automation: Automating accruals cut close time from five days to two — a 60% gain driven by employee initiative.
Career and Hiring Implications
Ryan shared that Zapier embeds AI literacy into all hiring. Candidates are evaluated on how they use AI to enhance workflows, not just awareness. At ClickUp, candidates complete take-home tasks with AI allowed — then explain their prompting and iteration process.
Both CFOs predict finance teams will evolve into “obelisk-like” structures — leaner, with humans managing AI agents rather than junior staff.
Future Cost Mix: Today, 80% of finance costs are payroll; within three years, Dan expects a growing share to come from AI tooling and model usage. Skills in Demand: Workflow automation, prompt iteration, and no-code prototyping — especially using core platforms like Zapier, ClickUp, and Airtable.
Frequently Asked Questions
1. How should CFOs measure AI ROI when benefits aren’t immediate? Use Dan’s three-bucket model: time-savings for automation, capability metrics for unlocks, and decision-quality for improvements. Avoid rigid ROI tracking too early — focus on long-term leverage.
2. How can leaders avoid SaaS sprawl while encouraging experimentation? Apply the four-quadrant framework: commit long-term to “Holy Grail” tools, cap budgets for invisible platforms, test multiple copilots, and reserve small shared funds for sandboxes.
3. What structure best drives AI transformation? Adopt dual ownership — every department gets an AI goal, while a transformation officer tracks cross-team progress. Combine bottom-up wins with top-down accountability.
4. How should CFOs communicate AI progress? Use tailored storytelling: celebrate quick wins with teams, tie metrics to OKRs for leadership, and present multi-year leverage metrics to boards.
5. How will AI reshape finance careers? Teams will become leaner, with professionals orchestrating AI agents. Finance roles will shift toward analysis, creativity, and strategic advising — freeing humans for higher-value work.
Recommended Resources & Further Learning
CFO Connect Community: Watch full session recordings and connect with peers navigating AI transformation.
Vendor AI Roadmaps: Engage existing SaaS providers (Coupa, NetSuite, Pigment, etc.) as partners in AI integration.
Low-Code AI Platforms: Experiment with Zapier, ClickUp, Airtable, or Google AI Studio.
Peer Learning Networks: Prioritize finance-specific AI communities over generic courses.
Outcome-Based Pricing Models: Study vendors aligning fees with customer success.
Conclusion
The message from this panel is clear: AI in finance has moved from experimentation to execution.
Success depends on disciplined implementation — aligning investments with strategy, consolidating platforms, and building AI skills into team culture and hiring.
Both CFOs painted an optimistic future: finance teams freed from manual work, empowered to analyze, plan, and connect. Leaders who adopt structured frameworks, foster experimentation, and communicate AI progress transparently will drive the next wave of transformation.
Ready to apply these insights? Join the CFO Connect community to access the full recording, connect with peers, and explore upcoming sessions on AI in finance.
