The Moment I Knew FP&A Was About to Change
I built my FP&A team from one person to six inside a cybersecurity company. No playbook, no big budget, just the usual grind — monthly close, variance analysis, board decks, repeat. Then six months ago, I watched an AI agent do in four minutes what used to take one of my analysts two full days. That's when I knew the game had changed, and I started writing this newsletter.
This is not about hype. This is about what's actually happening on the ground — and why the gap between early movers and everyone else is widening faster than most people realize.
First, Let's Kill the Buzzwords
"AI" has been slapped on every finance software pitch deck for the last three years. Most of it was nonsense. So let me be precise.
An AI agent is not a chatbot. It's not a fancy Excel macro. It's an autonomous system that can:
- Reason about what you're trying to accomplish — not just follow a script
- Plan a multi-step approach to get there
- Execute by pulling data from your ERP, CRM, HRIS, and planning tools
- Self-correct when something doesn't look right
Think of RPA as a very fast intern who follows instructions literally. If the spreadsheet layout changes, the bot breaks. An AI agent understands the intent behind the task and adapts.
AI Agent: "Find this month's EMEA revenue, compare to forecast, explain why it's off, and flag if the variance is material."
One breaks when you move a column. The other figures it out.
Where Real FP&A Teams Are Using This Right Now
This isn't theoretical. Here's what's happening across the FP&A landscape today.
Variance Analysis: From 15 Hours to 2
A mid-market SaaS company ($80M ARR, FP&A team of four) was spending 15 analyst-hours every month building variance analysis. Pulling actuals from NetSuite. Comparing to forecast. Drilling into department-level detail. Writing commentary for the CFO.
They configured an AI agent to run on close day +2. It connects to the ERP via API, pulls actuals, identifies variances above $50K or 10% materiality, drills into transaction detail, categorizes each driver (timing, volume, rate, one-time), and drafts two-sentence explanations.
15 analyst-hours. Report delivered close day +5. Team buried in spreadsheets during the most critical week of the month.
2 hours total — 1 hour for the agent, 1 hour for the lead to review and add strategic context. Report delivered close day +3. Team spends the freed-up 13 hours on forward-looking analysis.
The agent didn't eliminate FP&A judgment. It eliminated the manual work that was preventing the team from applying their judgment.
Rolling Forecasts: Weekly Updates Without Adding Headcount
Pigment and Planful are both building agent-like capabilities into their platforms — systems that monitor incoming actuals, update driver assumptions automatically, and flag forecast drift before your monthly review meeting. Anaplan has announced AI-powered scenario generation that builds cases from natural language prompts.
The implication is massive: teams that currently do monthly forecast updates can shift to weekly cadence without adding a single person. One Series C fintech company reported that their AI-assisted rolling forecast caught a $2.3M revenue shortfall three weeks earlier than their traditional process would have.
Three weeks. In a business moving that fast, that's the difference between a proactive board conversation and a crisis.
Scenario Modeling: From "Give Me Two Days" to "Give Me Twenty Minutes"
Here's a conversation that plays out in every FP&A team:
CFO: "What happens if we freeze hiring, cut marketing 20%, and lose our second-largest customer?"
FP&A lead: "I'll have something for you by Thursday."
With an AI agent plugged into your planning model, that answer comes back in minutes — not as a rough guess, but as a structured comparison across scenarios with cash impact, runway implications, and key risk thresholds clearly laid out.
The Real Business Partner: Building, Not Just Reporting
I want to share something from my own team because I think it illustrates where the FP&A function is heading.
My team is working hand-in-hand with Navan — the travel and expense platform — to build personalized travel dashboards. Not the generic, one-size-fits-all view that most tools ship with. Each user gets a customized dashboard tailored to what they actually need. Different views for the road warrior VP, the occasional traveler, the department head tracking team spend.
FP&A isn't sitting on the sidelines waiting for a report. We're in the room with the vendor, shaping the tool, defining the data architecture, making sure the output actually drives decisions.
If your FP&A team's entire output is a monthly deck and a quarterly forecast, you're leaving enormous value on the table.
This is what "business partner" actually looks like. Not the buzzword on a job description. The real thing — where finance has a seat at the design table because they earned it.
The Urgency Is Real. The Window Is Closing.
Let me lay out what I'm seeing across the FP&A community, and I'll let you draw your own conclusions.
The gap is widening. Early-adopting FP&A teams are compounding their advantage. They're faster, more accurate, and spending more time on strategic work. Every month that gap grows, the harder it becomes to catch up.
The role itself is being redefined. Companies that have deployed AI agents in finance are rethinking what they need from their FP&A teams. They're hiring fewer people, but expecting more strategic impact per person. The teams that can't deliver strategic value — because they're buried in manual work — are the ones getting restructured.
Your competitors are already experimenting. A recent Gartner survey found that over 60% of finance leaders plan to invest in AI agents within the next 18 months. The question isn't whether this is coming to your industry. It's whether you'll be leading the adoption or reacting to it.
I want to be clear: this is an opportunity, not a threat. But opportunities have expiration dates.
Your Move: Five Things to Do This Month
You don't need a six-month AI strategy. You need to start building muscle memory. Here's what I'd do if I were starting today.
1. Get your hands dirty with AI tools. This week. Not a webinar. Not a demo. Open Claude or GPT-4, feed it a real variance analysis, and see what it produces. You need personal intuition for what these tools can and can't do — and you won't get that from a slide deck.
2. Map where your team's time actually goes. Track it for one month. I guarantee you'll find that 40-60% of your team's hours go to work that doesn't require their expertise — data gathering, reconciliation, formatting. That's your automation target list.
3. Fix your data access problem. AI agents are only as good as the data they can reach. If your source of truth is someone's desktop Excel file, or if pulling actuals requires three emails and a VPN, that's your bottleneck — not the AI.
4. Pick one deliverable and pilot an AI-assisted workflow. Not the board deck. Something lower-stakes. Monthly variance analysis is a great starting point. Run the AI output alongside your normal process for one cycle. Compare. Learn.
5. Redefine what "good" looks like for your team. If an agent can produce a variance report in ten minutes, the team's value is no longer in producing the report. Start shifting performance expectations toward insight quality, decision influence, and speed of strategic response.
Data security matters. Sending financial data to external AI services raises real concerns. Work with IT and legal to set guardrails.
Change management is hard. Some team members will see this as a threat. Lead with honesty about what's changing and genuine investment in helping people grow.
The tech is still maturing. The gap between a demo and production-grade finance workflow is real. Start with low-risk use cases and expand from there.
None of these are reasons to wait. All of them are reasons to be thoughtful about how you move.
The Bottom Line
The FP&A function hasn't seen a shift this significant since the spreadsheet. The leaders who engage now — critically, practically, and with clear eyes about both the potential and the risks — will have an advantage that compounds every quarter.
In the next issues, we'll go deeper: specific tools, implementation playbooks, real case studies, and the skills FP&A professionals need to stay ahead. This newsletter exists because I believe the best FP&A leaders are builders — and builders need honest, practical information, not marketing fluff.
The function is changing. The only question is whether you're shaping that change or reacting to it.