The CFO's 90-day AI implementation roadmap

68% of CFOs say they've been slow to adopt AI because they don't know where to start.

The tools feel powerful. The potential is vast. Yet the path forward feels ambiguous or risky.

"There's definitely no playbook," says Jessica Pillow from OpenAI. "We're all sitting around wondering how we have this AI, and it's not going anywhere. What we're doing at our company is treating it like a product: constantly testing, iterating and making small, incremental movements."

The finance leaders seeing results aren't waiting for the perfect roadmap. They're starting with focused experiments, building momentum and scaling what works.

Here's a practical 30-60-90 day plan to embed AI into your finance function, based on what leading CFOs are actually doing.

Why most AI initiatives fail before they start

The irony of AI adoption in finance is this: teams are too busy to implement the tools that would make them less busy.

"The biggest challenge is time and focus," says Ido Peled, Finance Data & Technology Lead at Adyen. "Finance professionals are busy with closing and reporting. To innovate, we must create protected time for experimentation."

Monthly and quarterly close cycles absorb most spare capacity. Teams end up relying on manual processes even when automation exists. Without structured time, AI remains a "side project" that never launches.

The solution isn't finding more time. It's treating AI implementation as a core project with defined milestones, protected resources and clear accountability.

Days 1-30: Immediate actions

The first 30 days should focus narrowly on scoping and early tests. The goal is not to "do AI everywhere" but to create momentum and lay foundations for scale.

Identify one high-friction workflow for AI testing

Select a finance process that is manual, repetitive and consistently frustrating for your team. This could be spend categorisation, variance analysis, reconciliations or management reporting.

Choose a workflow with clear inputs and outputs so impact can be observed quickly. A narrow, well-defined use case reduces risk and makes results obvious to stakeholders.

What leading teams are testing:

• OpenAI started with contract analysis and ASC 606 memo drafting

• Spendesk focused on real-time spend reconciliation

• Adyen began with knowledge retrieval for business partnering

The key is specificity. "Improve reporting" is too vague. "Automate monthly variance commentary for the CEO pack" is concrete and measurable.

Audit your existing tech stack before adding new tools

Before purchasing new AI solutions, assess what capabilities your ERP, FP&A tools, BI platforms and productivity software already provide.

Many organisations underuse embedded AI features they already pay for. Microsoft Copilot agents, for example, are available to Microsoft 365 users but rarely deployed.

Two ready-built agents finance teams can use today:

• Researcher Agent: Acts as your personal chief of staff. Prompt it: "Help me plan my week ahead. What are my key meetings, deadlines and priorities?" It compiles your schedule, flags important emails and pulls documents you need.

• Analyst Agent: Handles repetitive analytical tasks. Upload your budget file and actuals, then ask: "Compare these files and show me all variances greater than 10%. Highlight which departments are over budget and create a visualisation." Tasks that take analysts hours now take 5-10 minutes.

"The main challenge isn't finding AI tools," says Dan Zhang, CFO at ClickUp. "It's having too many. We replaced five separate AI note-taking tools with a native one built in ClickUp."

Begin measuring AI impact beyond time saved

Time savings are easy to quantify but rarely tell the full story. Start tracking additional metrics such as speed of decisions, forecast accuracy, error reduction and stakeholder satisfaction.

Establishing these measures early lets you build a more compelling business case for future investment.

Track three types of value:

• One-to-ten automation: Time savings from automating repetitive tasks

• Zero-to-one unlock: New capabilities previously impossible

• C-to-A quality boost: Better insights and decision-making

"If we only measure AI by time saved, we miss its real value – capabilities and quality," warns Zhang.

Related reading: AI ROI in finance: How leading CFOs measure success

Days 31-90: Medium-term goals

With early lessons in hand, the next phase is about structure. Over the following 60 days, focus on building capability, accountability and guardrails around AI use.

Launch a 90-day automate-upskill-govern plan

Run a focused 90-day programme that balances automation with people development and risk management.

This plan should clearly define:

• Which processes will be automated

• What skills the team needs to develop

• How AI use will be governed

At Zapier, 98% of employees use AI tools. CFO Ryan Roccon explains: "Transformation requires mandates, not nudges. AI usage is not optional or additional – it's now part of the job."

They measure ROI through objective metrics (cycle times, pull requests merged, issues resolved) and subjective surveys. "It begins with clean data collection and segmentation," Ryan explains. "We pinpoint where AI adds the most value: in code writing, refinement or QA."

Establish AI champions within your finance team

Identify a small group of finance professionals who are curious, credible and close to the work. Empower them to test tools, share learnings and act as a bridge between finance, IT and data teams.

These champions accelerate adoption by translating AI capabilities into practical, finance-specific use cases.

ClickUp's company communications include weekly AI highlights, monthly awards and quarterly hackathons. According to CFO Dan Zhang, this shows teams that "AI is not optional; AI is required."

Make AI use central to performance conversations and team objectives.

Create governance frameworks for AI tool adoption

Put clear rules in place for data usage, model validation, access controls and human oversight.

Well-defined guardrails build trust with auditors, regulators and senior leadership while still letting teams innovate with confidence. Governance should enable safe experimentation, not shut it down.

Photoroom's principle is "freedom with accountability". Julien Lafouge explains: "If a tool costs less than $100 and saves at least one hour, buy it and test it – no approval needed."

This combination of autonomy and oversight breeds innovation without chaos.

Months 4-12: Long-term vision

Once pilots prove their value, the focus shifts to scale and sustainability. Over the following year, AI should become embedded at the core of finance operations.

Build a governed finance data core

AI performance depends on data quality, consistency and accessibility. Create a trusted finance data foundation with clear ownership, standard definitions and strong controls.

This data core becomes the backbone for scalable AI use across reporting, planning and decision-making.

Adyen built a unified data hub with established governance that powers AI-driven operations. Their knowledge retrieval system now provides instant answers globally without waiting for the finance team.

"Nine out of ten finance leaders think their data is an absolute hot mess," says Phil Sharp, Interim CEO & CMO at Subscript. "Everyone whispers it like they're the only one facing this problem – but I hear it five times a day. The real question becomes: how do you operate when data will always be somewhat messy?"

Start small. Pick one data domain (spend, revenue, headcount) and get it right. Then expand.

Redesign roles around AI capabilities

As AI absorbs more routine work, finance roles should evolve toward judgement, insight and business partnership.

Update role definitions, performance metrics and career paths to reflect this shift. Make AI literacy a baseline expectation, not a specialist skill.

At Vitalize Health (via NetSuite), an accountant automated himself out of his reconciliation role and moved into strategic FP&A. "He feels more fulfilled, and the company benefits from his new focus," says Rebeca Bichachi, Product Marketing Director at NetSuite.

The professionals who advance fastest will be those who proactively automate their own tasks to free themselves for higher-value strategic work.

Scale successful pilots across the organisation

Identify the AI use cases that have delivered measurable value and standardise them across teams, regions or business units.

Integrate these capabilities into core processes, rather than leaving them as optional tools. At this stage, AI moves from innovation to infrastructure.

OpenAI uses a value-effort matrix to focus on the right initiatives:

• Low value/low effort: Repetitive controls

• High value/low effort: Reporting and disclosure drafting

• High value/high effort: Long-term automation linked to LLMs and ERP systems

"The goal is to move quickly on impactful wins while laying the technical foundation for deeper integration over time," says Yubo She, Head of Technical Accounting at OpenAI.

Best practices that accelerate success

Beyond the timeline, certain practices separate teams that thrive from those that stall.

Pair finance with engineering and data teams

The most successful implementations come from cross-functional collaboration, rather than finance working in isolation.

Examples include rotating controllers into technical teams, running joint workshops with engineers and building internal "AI councils". At ClickUp and Zapier, this happens through quarterly hackathons and mandatory usage expectations across departments.

Add AI skills to your hiring requirements

As AI becomes a core skill for finance teams, you need to both train and hire for it.

Zapier spotlights AI literacy in all hiring. Candidates are evaluated on how they use AI to enhance workflows, not just awareness of LLMs.

ClickUp has candidates complete take-home tasks with AI allowed. They're then asked to explain their prompting and iteration process.

ARIA Finance Director Mike Tsang looks for behaviours rather than specific AI experience. "People who can approach things from first principles, who have that creative mindset, who've worked with data, who can understand it and who can break down and approach, tactically and strategically, why we would do something."

Soft skills like curiosity and rigour are timeless. But AI amplifies their importance.

Communicate ROI to different stakeholders

Tailor your messaging based on the audience:

• For employees: Celebrate quick wins. "What took 30 minutes now takes 30 seconds."

• For leaders: Connect AI to OKRs. "AI reduces close time by 3 days, enabling faster strategic decisions."

• For executives and boards: Show how AI reduces costs and unlocks capabilities across revenue functions.

"Start bottom-up to spark adoption," says Zapier's Ryan Roccon, "but pair it with a top-down edict to drive real transformation."

The bottom line

The competitive advantage doesn't go to teams with the perfect plan. It goes to those who start with one workflow, track three metrics, learn what works and scale what delivers value.

"Finance leaders increasingly value problem-solving and storytelling skills," says NetSuite's Rebeca Bichachi. "Only 23% of CFOs now rank deep accounting knowledge as the top hiring priority."

The shift is already happening. AI is moving finance from operational to strategic, from reporting the past to shaping the future.

The question is whether your team will lead that transformation or react to it.

Start with 30 days. Pick one workflow. Build from there.

Related reading: 7 AI tools finance teams actually use in 2026