Executive Summary
Most FP&A teams rely on a single forecasting method and wonder why their numbers keep missing. The fix is not working harder. It is thinking differently about method, model design, and how finance shows up in decisions.
In this episode of The Diary of a CFO, I spoke with Christian Wattig, founder of InsideFPA.com and director of the Wharton FP&A Certificate Program, about what separates forecasts that executives ignore from ones they actually use.
You will learn the three forecasting methods worth mastering, what makes a model trustworthy, and how FP&A teams shift from explaining variances to shaping decisions.
You will discover practical tools to improve forecast accuracy, simplify your models, and build the business trust that makes strategic impact possible.
Why Most FP&A Forecasts Miss the Mark
Finance professionals at Procter and Gamble, Unilever, Squarespace, Google, Lowe’s, and Merck all wrestle with the same challenges. Christian has trained FP&A teams at all of them, and the pattern is consistent. The specific numbers differ. The problems do not.
The most common one: teams use a single forecasting method and try to improve results by doing more of the same. The forecast becomes a ritual rather than a decision tool. Actuals come in. Variances get explained. The deck gets updated. Then the cycle repeats.
What is missing is the connection between the forecast and what the business is actually doing. A revenue projection that does not reflect whether a new product is launching, a sales team is expanding, or a market is being entered is not a forecast. It is a trend line with a confidence problem.
The deeper issue is that most FP&A teams stop at the what and the why. Christian pushes further: the forecast should answer the so what. What does this variance actually tell us about how we grow more sustainably? What is the concrete opportunity or risk? That is the shift from scorekeeper to business partner.
Choosing the Right Forecasting Methods for Your Business
The single biggest improvement most FP&A teams can make is to stop relying on one method and run several in parallel. Each method answers a different question. Together they create a more complete and more defensible forecast.
Christian recommends three as the foundation:
Driver-based forecasting
This is his preferred starting point for revenue. Instead of projecting last year’s number forward, you identify the five to ten things the company is actually doing that have an impact on revenue. New hires. Marketing investment. Product launches. Market expansion. Each gets an estimated impact.
The forecast then stays relevant throughout the year. Every month, you learn something new about at least one driver and can update accordingly. It also forces a conversation with the people responsible for each driver, which is the foundation of real business partnering.
Zero-based planning
Most useful for costs. The question is simple: if we were starting from scratch, which of these expenses would we actually choose? For most companies, a full zero-based rebuild is impractical. The modified version works well: group expenses by purpose, review each group, and ask whether the spend is earning its place. Christian used this for a North America marketing budget of $1.2 billion at Unilever. The efficiencies it surfaced would never have appeared through trend-based planning.
Time series analysis
The oldest method and still one of the most valuable. Look at how actuals have behaved historically and let that pattern inform the future. On its own, it anchors the forecast too heavily in the past. Combined with driver-based forecasting, it provides a realism check, connecting the plan to history and flagging when a projection is more optimistic than the data supports.
What Makes a ‘Good Enough’ Forecast Model
A better model is not a more complex model. It is one that can be updated quickly, audited easily, and trusted by the people using it.
Christian teaches a practical test. Put your thumb on the screen. If a formula is longer than your thumb, it needs to be simplified. The fix is not to remove logic. It is to break the formula into multiple rows, each calculating one part of the answer, then adding them up at the bottom. The result:
Every formula fits on one screen
Pressing F2 in Excel shows the full logic without scrolling
Last-minute changes take minutes, not hours
Mistakes have fewer places to hide
“The rule of thumb in modeling: put your thumb on the screen. If the formula is longer than your thumb, you’re not there yet. Break it into multiple steps.” — Christian Wattig
The same discipline applies to how AI fits into modeling. When a team member asks how to write a particular formula, Christian’s response is not to explain it. He asks them to open an AI tool, build the prompt, then evaluate what comes back. The skill he is developing is not formula memorization. It is asking precise questions and knowing whether the answer is right.
That is also what makes a good forecast model. You do not need perfection. You need a model fast enough to update, simple enough to explain in two minutes, and honest enough that the numbers can be defended in a room full of skeptical executives.
Turning Forecasts into Strategic Decisions (Not Just Slides)
The goal of FP&A is not to produce a number. It is to help leaders make better decisions. A team that consistently achieves that becomes the team leaders go to before making a big call, not after. Getting there requires two things Christian comes back to throughout the episode.
Business acumen. You need to understand how the company makes money at a granular level. Not just the P&L. The strategies, the tactics, the metrics that measure whether those tactics are working. A finance team that cannot speak the language of the people it supports will always be received as a finance opinion rather than a business insight.
Psychological safety. Christian references Google’s Project Aristotle, a large internal study of what made their best teams perform. The biggest predictor was not the right mix of personalities or shared interests. It was psychological safety: an environment where people feel safe to flag mistakes and speak up when something is not working. For FP&A leaders, this is not a soft concern. If your team is afraid to flag a forecast that looks wrong, the numbers that reach the board will be wrong. The environment you create is a data quality issue.
“The goal should be that leaders across the company come to you before making a big decision, for your advice and your analysis. That’s what we want, because then we have real impact.” — Christian Wattig
Christian also addresses the transition from individual contributor to manager, which is where many FP&A professionals stall. You get promoted for technical skills, but the role changes and interpersonal skills become the primary lever. He covers this in depth in the full episode, which you can listen to here.
Listen to the full episode with Christian Wattig on The Diary of a CFO podcast. And if you want to know how ready you are for the CFO seat, start with the free CFO Readiness Assessment.
🎧 Also worth listening to: The Art of Becoming a Strategic CFO with Dr. Tamer Alsayed explores how strategic finance leaders bridge vision and execution, including how to make forecasting a tool for real business decisions rather than a compliance exercise.
FAQ
Why does using multiple forecasting methods improve accuracy?
Each method captures a different dimension of the business. Driver-based forecasting connects the forecast to strategy and keeps it current throughout the year. Zero-based planning surfaces cost inefficiencies that trend-based models miss. Time series analysis provides a historical reality check. Together they create a more complete and defensible picture than any single method can produce.
How do you know when a financial model is too complex?
Apply the thumb rule. If a formula is longer than your thumb on the screen, it needs to be broken into smaller steps. A good model has short formulas, clearly separated inputs and outputs, and can be updated and audited quickly. Complexity is not rigor. It is usually where errors hide.
What is driver-based forecasting and how do you build one?
Driver-based forecasting anchors revenue projections to the specific actions the business is taking: hiring, marketing investment, product launches, market expansion. Start by identifying five to ten drivers with the teams responsible for them. Estimate the revenue impact of each. Update monthly as you learn more. It is a business exercise, not a finance exercise.
What does it actually take for FP&A to earn a seat at the strategic table?
Business acumen and psychological safety. Acumen means understanding how the company makes money beyond the P&L. Safety means building a team environment where problems get flagged early rather than buried until the board meeting. The forecast is only as good as the team’s willingness to tell the truth about it.
Where can I learn more about building FP&A skills beyond forecasting?
You can explore Christian’s courses and workshops at InsideFPA.com. And if you want to understand how FP&A skills fit into the broader path to CFO, read Controller vs Head of FP&A: Which Path Leads to CFO Faster? on The Diary of a CFO blog.
Go deeper on the team and leadership side in The Strategic CFO vs. the Transactional Finance Function: How to Close the Gap on The Diary of a CFO blog.
About The Author:
Wassia Kamon is a CFO and the host of The Diary of a CFO, where she interviews finance and business leaders on strategy, risk, and leadership. She writes about finance leadership and governance in small and mid-sized organizations, including what works, what breaks, and how leaders manage growth and complexity without burning out.