CFO Dashboards Masterclass ft. The FP&A Guy
Speakers

Paul Barnhurst, Founder of The FP&A Guy

Kirk Kappelhoff, Senior Director of Strategic Finance, Drivetrain
The CFO has promised the board that churn will drop from 15% to 10% this year. At the quarterly review, he asks which accounts are at risk. Your dashboard tracks every customer you’ve lost, but nothing on it flags the next one. Then the head of Customer Success says her churn rate is three points lower than yours. After 20 minutes of trying to reconcile that, the CFO asks you to “take it offline,” and the meeting ends without any plans for how to keep even a single customer.
Leaders act on financial information when it reaches them with clarity and context.
Paul Barnhurst, founder of The FP&A Guy, learned this the hard way. He explained how when he joined Kirk Kappelhoff, Drivetrain’s Director of Strategic Finance, for a webinar on CFO dashboards.
Paul had spent six months on a project. His leadership team was fully aligned behind it. Then he presented it to the company’s global CFO. Ten minutes into the presentation, the CFO said, "This is a terrible idea. Why would we do this?" Paul was flabbergasted. Then, the CFO brought in the general manager, talked with him for five minutes, and approved the project after all.
Afterward, I asked my boss what happened, and he said: “Paul, your analysis was brilliant, your presentation lacked.” ~ Paul Barnhurst, Founder, The FP&A Guy
Paul had spent so much time on the data that he only shared what he thought the CFO needed, leaving out most of the story behind the numbers. In hindsight, he realized that “too much data, not enough context” is a common mistake. Today, it’s at the top of his list of the mistakes he sees most often, followed by no clear takeaway, KPIs that don’t match the business goals, and dashboards that are overdesigned or underexplained.
His webinar with Kirk is a masterclass in how to avoid those mistakes and how to build a CFO dashboard that a leadership team can act on.
No time to watch? This article walks through the framework Paul uses, with the audience Q&A and the full transcript at the end.
61:22The short version
An FP&A dashboard built for the CFO and leadership team isn’t worth much if the people reading it can act on what it shows. Here's how Paul's method takes you from a page nobody can actually use to a handful of numbers that support sound decision-making:
The core CFO dashboards: financial and operational
Kirk asked Paul which CFO dashboards every CFO should have as a baseline. Paul said the core dashboards fall into two groups: financial and operational.
The financial side gives the CFO visibility into cash flow and the core financials. The operational side covers sales and marketing, which are critical to nearly every business, along with an expense analysis and operational metrics. Paul added that department-specific dashboards are especially useful when the CFO needs a targeted conversation with one department or FP&A lead.
Kirk then walked through four CFO dashboards in Drivetrain, starting with the 13-week cash flow. He said it had recently emerged as one of the most important dashboards for Drivetrain’s customers for a couple of reasons. One is that companies are concerned about liquidity, and the other is that the view itself is simple, showing forecast vs. actuals and the variance between them. The other three dashboards showed budget versus actuals, the P&L, and headcount.
The mistakes that stop leaders from acting on a dashboard
Choosing the right dashboards is just the first step. Most of the advice Paul offered in the masterclass was about what keeps leaders from acting on them: the “killer mistakes.” The first is what he learned the hard way: too much data and not enough context. Here are the other four, along with Paul’s fix for each.
KPIs that don’t match the business goals
Paul’s starting point is the strategy. Companies typically have OKRs or a few key goals, and the question is whether each metric actually lets you influence them. Dashboards often try to “boil the ocean,” he said, so measure only what can actually be impacted and what matters to the business. If the goal is reducing churn from 15% to 10%, you’d watch NPS and time-on-product
Kirk added that a metric nobody can influence can still live in a report somewhere else.
Once the metrics match the goals, they should last as long as the goals do. Paul counts changing tracked metrics too often as a related mistake.
You don't change strategy and goals monthly, or if you do, you're probably not very successful. Set them for a year, revisit once or twice, adjust as needed.
Metrics should follow the same discipline, Paul said, so people have time to influence them. “If you picked them for the right reason in the first place, why keep changing them?”
Underexplained numbers
"I'd bet everyone listening has been in a meeting where someone says 'my number's different,'" Paul said. He explained that this happens when the methodology isn't agreed upfront. Paul's fix is a clear explanation of data sources, metric definitions, and visible formulas, so everyone works from the same page. In one of his previous roles, the CFO, CRO, and others couldn’t agree on what a booking was, and that had to be settled before any reporting could proceed.
Paul also said to show actuals beside the budget and latest forecast and the prior period, so it’s easy to see how the business is performing against a target. Last year gives you the baseline.
He said more teams now compare the budget directly with the latest forecast, too. The variance between them shows how far expectations have moved since the budget was set. Tracked over a full year, it helps explain a miss: Was the original budget unrealistic, or was the forecast off?
Overdesigned dashboards
"I've built some horrible dashboards over the years myself," Paul said. The principles he leans on come from Storytelling with Data by Cole Nussbaumer Knaflic: Think like a designer, keep it simple, use contrast effectively, make context clear, and ruthlessly reduce clutter.
If an element isn't adding to the story, it's just adding cognitive load. Every element should justify its presence. If it doesn't add value, the reader shouldn't have to look at it.
To illustrate the point, Paul put up a screen of digits and asked Kirk how many eights were on it. Kirk estimated that there were sixteen and was way off. The answer was nine, and even confirming it took a couple of seconds. The next version bolded the eights with a bit of shading, which made them much easier to spot. The last one used color, and that was easier still.
Cut anything that doesn't help the reader, and use bold, shading or color only to highlight the numbers that matter most.
No clear takeaway
Paul said a dashboard with no focal point or takeaway is the flip side of having a clear, solution-aligned recommendation.
Kirk made a related point: Storytelling gets skipped a lot, and people just jump into a report and read the numbers off the screen. He then quoted a favorite saying: “Never tell a number without a story, never tell a story without a number.”
Paul pointed to a framework that dates back to the 1970s: “what, so what, now what.” He said most of FP&A is good at the “what,” such as reporting that revenue missed by $200,000 because volume was down. The “so what” is why that matters. The piece people want most is the “now what,” meaning, what are we going to do about it?
The key in "now what" is offering an actual recommendation. Data storytelling in finance is fundamentally about driving decisions, giving calls to action. If all you do is present data, you're falling short on that piece.
Paul tied the recommendation to the tone you take with bad news. Piling onto bad news doesn’t help anyone, he said, while telling people where things stand and how you plan to address it lands very differently. His overarching rule is to “just tell it how it is.” If the problem is a model error, own it and fix it. A missed revenue target or a department overspend is something finance tracks, and most of the time, it isn’t finance’s fault.
“Our job in finance is to present numbers and own them alongside the business, not "you missed your numbers, what did you do wrong" but "we missed our numbers, here's what happened, here's an issue we found."
Getting the dashboard to the people who can act on it
When Paul talked about presenting the right numbers to the right stakeholders, two of his points were about delivery. The right people, meaning those who can actually influence the numbers, need to be the ones looking. And people don't need to see a number a month after it happened. It has to be timely enough to influence what is going on.
People won't look at reports they have to go hunting for; you want to push it to them and alert them to what matters.
Before a meeting, Paul often sets the agenda and gives* access in advance to the report itself or to the stakeholders who need to understand its assumptions. That way, the story has started before the meeting does, a habit that matters just as much for board presentations.
On cadence, Paul said it depends on the business. "Weekly is often fine; sometimes monthly." He added: "Don't assume everything needs to be real-time. There's a real cost to that."
Answering follow-up questions without leaving the dashboard
Many financial reporting tools offer dynamic dashboards that link the numbers on a dashboard to the detail behind them so a follow-up question can be answered in the same place it came up.
Kirk demonstrated this in Drivetrain’s executive overview dashboard. One click switches the timeframe (last quarter, current quarter, year to date), and every chart updates with it. From a month’s total, you can jump straight to the transactions behind it, then sort, filter, or export them to Excel. Or you can drill down one level at a time, from region to country to individual contracts. Kirk said this bifurcation “is really where the insight lives.” If a number still looks off, a comment box lets you tag a colleague on that chart or cell, and they’re pinged in Slack, Teams or email.
Paul agreed that this is what good drill-down and commenting should look like. Line-item commenting is especially valuable, he said, along with free-form notes, and the view resembles the pivot tables people already know. The easier that is, he said, the easier real analysis becomes.
Dynamics dashboards are just one of a robust set of reporting features Drivetrain offers as a comprehensive, AI-native FP&A platform.
To see how a dynamic dashboard works with your own numbers, book a demo.
Summary
- Track a few metrics tied to the strategy, with leading indicators first: Choose metrics the people reviewing the dashboard can influence. Set them for the year and revisit them once or twice, so the people responsible have time to improve the results.
- Agree on one definition for every metric, and show each one against a benchmark: Document the data sources, definitions, and formulas so everyone works from the same numbers. Show actuals beside budget, the latest forecast, and the prior period.
- Send the dashboard to the people who can act on it, as often as they need it: Flag what matters before you deliver it to them, because people won’t go hunting for reports. Set the cadence by how often they make decisions. Weekly is often enough, sometimes monthly. Real-time has a real cost, so don’t assume every metric needs it.
- Cut everything on the dashboard that doesn’t help the reader see the point: Remove gridlines, borders, and anything else that doesn’t help the reader. Save bold, shading, and color for the numbers that matter most.
- Present the numbers as a story that ends in a recommendation: Explain what happened, why it matters, and what to do next.
Questions from the webinar
How do you balance leading and lagging metrics on a sales or revenue dashboard?
Paul said to "start with the leading indicators people can act on, then layer in the lagging metrics that complete the picture," and to keep asking what people actually need to make a decision. On a sales or revenue dashboard, you'll generally want some lagging metrics, since you'll likely need the P&L figure alongside bookings and billings.
How do you get senior leaders to engage with the dashboard?
One principle that has stuck with Paul is BLUF, or bottom line up front. Lead with what people need to know plus any recommendations, then put the detail below. He recounted someone whose monthly Excel reports nobody opened until they led with the bottom line and stopped sending the full report. Then people started asking for the full report.
How should you analyze cohorts?
Paul said cohort analysis is critical, and the key is understanding what actually distinguishes one cohort from another. People often default to date-based cohorts, and that’s fine, but customer size or geography may be the real driver. So, dig into the data, then show the driver clearly.
What do you do when the data says one thing and leadership's gut says another?
Paul quoted the old joke, "in God we trust, all others bring data", but said there is genuinely something to intuition too. Most of the time, lean toward the data, but there are moments where going against it is right. Either way, know when to stop the analysis-paralysis loop: ive yourself a window, say two weeks or forty hours, then make the call.
Speaker spotlight

Paul Barnhurst
Founder of The FP&A Guy
Paul Barnhurst is the founder of The FP&A Guy and a leading voice in the FP&A and CFO Tech space. With over a decade of finance experience across travel, banking, and SaaS, Paul has engaged with 100+ CFOTech companies in the past two years. He’s the host of popular FP&A podcasts, an advisor at Born Capital, and a seasoned trainer and consultant for finance teams worldwide.

Kirk Kappelhoff
Senior Director of Strategic Finance, Drivetrain
Kirk Kappelhoff is a financial modeling expert with a BBA in Finance & Accounting and nearly a decade of experience at Deloitte, EY, and KPMG, where he built models for pre-IPO companies, M&A transactions, and strategic planning initiatives. At KPMG, he led the Business Modeling Services team, specializing in equity stories and financial forecasts that help high-growth companies communicate their value to investors. At Drivetrain, Kirk writes about strategic planning, granular reporting, and modern FP&A best practices for rapidly scaling finance teams.
- 00:00
Paul Barnhurst
I'm doing great here with you, Kirk.
- 00:03
Kirk Kappelhoff
We are very excited to have you once again. It's always a pleasure getting to chat with the official FP&A guy. When did you get that moniker? When did that officially come into effect? Was that a recent thing, a couple years ago, or was it back in college?
- 00:18
Paul Barnhurst
No, no, much more recent. Probably about five years ago, I used it as a hashtag on LinkedIn. It just kinda stuck, and so I decided to name the business that, and away we went. It was really a self-thing — I thought, that'll be an interesting hashtag, kind of a way of distinguishing myself. Then people started referring to me that way on LinkedIn, and I said, I might as well lean into this, it stands out. So I just went with it.
- 00:52
Kirk Kappelhoff
Hey, good stuff. Well, either way, no matter what you're called, we're happy to have you and your phenomenal knowledge here. Today we're talking dashboards — a huge CFO Dashboards Masterclass, and we're excited to get a lot of your knowledge on this. For those who don't know, Paul is the founder of The FP&A Guy, a leading voice in the FP&A and CFO tech space, with over a decade of financial experience across travel, banking, and SaaS. He's engaged with over a hundred CFO tech companies in the past two years alone, hosts several popular FP&A podcasts and is a guest on many more, and advises OneStream Capital. He's a seasoned trainer and consultant for finance teams worldwide. Did I miss anything, Paul? If so, not important. Where are you calling in from today? I'm in Utah — I'd love to see where our listeners are joining from too. And you're in New York, if I remember right? That's correct — New York, the Big Apple itself. Definitely post in the chat where you're calling in from, and drop any questions you have — we want this interactive, and we'll respond mainly at the end but maybe throughout too. I'm Kirk, director of strategic finance at Drivetrain, and I have the pleasure of hosting Paul as we talk about CFO dashboarding. Today we'll go through three things: the CFO dashboard suite and what goes into each; a framework for presenting, with practical and real examples we've used at Drivetrain; and of course, we offer deep-dive consultations on modeling best practices — cohort modeling, TAM modeling, SaaS or non-SaaS. We love both learning from and teaching this community. Let's kick off with a poll: where do you currently create or manage your core finance dashboard, and what's your biggest challenge with creating dashboards? While you answer — Paul, where do you personally spend the most time dashboarding? Not necessarily your favorite, just where you find yourself doing it most.
- 03:23
Paul Barnhurst
Depends on the company and the tool they used — I've used Tableau, Power BI, Domo, some FP&A platforms (though usually I didn't have direct access there), and obviously built plenty in Excel. Today you don't really need much beyond a single business tool — a lot is happening in BI tools, and increasingly in dedicated platforms. Let's get into this. A few notes before we start — I understand you're a big fan of this Mark Cuban quote. What draws you to it? There are a lot of quotes like this — "what's measured improves," or "when performance is measured, performance improves." If you run a race for fun and don't track your time, you've got nothing to beat next time. Runners always want to set a new PR or challenge themselves — same in business. So you need to be able to measure and set baselines if you want to improve results. That leads nicely into the second point — presenting the right numbers to the right stakeholders. There are really three things there: first, measuring what can actually be impacted and matters to the business, not everything — dashboards often try to boil the ocean, I'll tell some stories on that later. Second, making sure the right people — those who can actually influence the numbers — are looking at them. And third, right timing: people don't need to see a number a month after it happened; it needs to be timely enough to actually influence what's going on. And that leads into clarity — we want clear numbers that make sense, which ties back to that boiling-the-ocean point. Exactly — we want to focus on what's important. We've all been there, focusing on too many things. One of my favorite quotes, from a boss of mine: "strategy is about closing doors" — you'll always have more opportunities than you can pursue. Same with dashboards: there's more data than anyone can really absorb, so you need to know what to push to the back versus keep upfront. Let's carry that into the eight high-level secrets to CFO reporting. What's happening on this page? What I like about this framework is its simplicity — doing things the right way, at the right time, in the right format, for the right people. It starts with getting the metrics right — nothing else matters if you pick the wrong ones. Then the data behind them. Format is comparatively easy to change, but still worth real thought — how many of us have had data put in front of us and had no idea what the person meant? Everybody's raising their hand. Then the right people, right medium — people won't look at reports they have to go hunting for; you want to push it to them and alert them to what matters. And that ties to right time and right frequency. It's a simple framework, but a great one for guiding dashboard and metric design. Definitely — I find myself circling back to this every time I build a new report: do I have the right metrics for these stakeholders, the right data? I go through and check each box, and often realize I'm missing something — then just adding that one missing chart changes the whole feel of the dashboard. One practical tip that just came to me: take an image like this, load it into ChatGPT, and ask it to turn it into a checklist. As you're building dashboards, ask yourself against that list — it's easy to forget things, and this is a simple way to make sure you're actually applying the framework.
- 08:34
Kirk Kappelhoff
That's a great point — and with image-reading AI now, that makes life a lot easier. On actionable dashboards — the first piece is metrics that matter, tying back to your first point about showing the right people the right stuff. Can you expand on what these key, strategy-tied metrics actually look like?
- 09:00
Paul Barnhurst
A couple things: companies typically have OKRs or a few key goals, but different departments can end up influencing things in ways that conflict with each other. So it's important to boil it down — what's the overall strategy, what are our OKRs, and do these specific metrics actually let us influence those things? Say the goal is reducing churn from fifteen to ten percent — then you'd be watching NPS and time-on-product closely, since churn itself is a lagging indicator, while those are more leading indicators telling you what to actually do.
- 09:45
Kirk Kappelhoff
On NPS specifically — for those unfamiliar, what is it and how should people track it?
- 09:50
Paul Barnhurst
Net Promoter Score is a simple way to see whether customers are promoters or detractors of your product, usually on a ten-point scale — nine or ten is a promoter, seven or eight neutral, six or below a detractor. You take promoters minus detractors to get the score. So a hundred percent of respondents giving nine or ten gives you a hundred; if nine out of ten gave that and one detracted, that's ninety minus ten, an eighty. Neutrals also drag the score down slightly. We've seen a lot of customers really lean into this, since it's genuinely one of the best leading indicators of churn — if quarterly NPS surveys show a decline, that's something you need to address. One point on metrics that matter: I've heard people say if a metric doesn't affect strategy or behavior, leave it out — what's the point of showing something you can't actually act on?
- 11:15
Kirk Kappelhoff
Hundred percent agree — can they actually influence it? It's fine to have something in a report elsewhere, but that doesn't mean it belongs on a dashboard, where the goal is understanding business health and enabling decisions. On comparisons — obviously we track metrics that matter, but why is tracking them over time so important, and why does that get left out so often?
- 11:40
Paul Barnhurst
Sometimes it's smaller companies without a formal budget process — flying by the seat of their pants, "we'll grow ten percent, or maybe five, who knows." Having actuals, budget, latest forecast, and prior period (year or quarter) side by side makes it easy to see how you're actually performing against a target, even though plans sometimes turn out unrealistic. Comparing to last year gives you a baseline; your forecast is your best current guess, so you can ask, are we at least hitting that? I'm actually seeing more people track budget versus forecast directly now, not just actual versus either — how accurate was our first attempt versus our second, which is a nice additional lens. And looking at that over a full year tells you whether a miss is a forecasting problem or a market problem — most companies don't hit their original budget exactly; the real question is whether that gap reflects unrealistic expectations or a forecasting issue.
- 13:15
Kirk Kappelhoff
Across both of these, what cadence would you recommend for reviewing?
- 13:20
Paul Barnhurst
Depends on the business — financials need whatever frequency actually lets you influence outcomes. Some things are fine monthly or quarterly, others need weekly or daily tracking, occasionally even real-time. There's no universal right frequency — it depends on the purpose, how often you need to know something to act on it, and what your stakeholders actually want. Weekly is often fine; sometimes monthly. Don't assume everything needs to be real-time — there's a real cost to that.
- 14:49
Kirk Kappelhoff
Beautiful, that's really helpful — not a fixed answer, more about what's best for the company given bandwidth. We've been fairly theoretical so far — let's get more actionable. I believe you've got an exercise for us, Paul.
- 15:07
Paul Barnhurst
Before we jump in — anything from points one through seven you want to touch on first? We've covered why metrics matter and tying them to strategy. Before we get to dashboard examples, it's worth talking about visual design — I've built some horrible dashboards over the years myself. I love these quotes from Storytelling with Data by Cole Nussbaumer: think like a designer, keep it simple (how many of us find it easier to be complex — I do), use contrast effectively, make context clear, and ruthlessly reduce clutter. If you don't need the gridlines, remove them; don't need the box border, remove it; don't need a bunch of colors, cut them. If an element isn't adding to the story, it's just adding cognitive load. Every element should justify its presence — if it doesn't add value, the reader shouldn't have to look at it. I like that framing — anything not adding value is just more for the reader to process, which dilutes attention from what actually matters. And keeping it simple is genuinely harder than making something complex — complexity just means piling more on; simplicity requires real analysis of what actually matters most. Far from being the lazy option, simple is the harder path. Agreed — complex is easy, simple is hard. Ready to jump into the exercise? Feel free to answer in chat — Kirk, how many eights are on this screen? I'll ballpark it — four lines, I'll guess sixteen. Way off — let's see the next one. Five, six, seven, eight, nine — nine total. That one took a couple seconds. Now the third version — much easier, right? Went from a jumbled mess to something bolded, with a bit of shading. Now the last one — even easier still. Great example of the value of color, and it really drives home how much easier recognition becomes when done right — though even here we missed bolding the middle eight, showing there's always more polish possible. I'm actually pretty colorblind, so the fact that this was still striking to me is impressive — shows how effective it is at drawing attention even for someone who barely perceives color. Definitely going to employ this. Exactly — and always be mindful that some of your audience is colorblind; I'm sure you've seen charts where someone says "this is hard to read, can you adjust the colors?" Normally I'm the one creating those hard-to-read charts myself — it's part of the learning process for all of us. On selecting a dashboarding tool — my high-level view is focus on what you actually need. If you don't care about exporting, that criterion is moot; if you only need three chart types, a tool with ninety doesn't add value. So for me it's largely feature-based — what widgets are available, tables, bar charts, waterfalls, pie charts, do you even need those — but I can't deny aesthetics matter too; misaligned or poorly colored charts are genuinely distracting. Any thoughts on the presentation side specifically? It starts with your actual requirements — what do you truly need, versus what looks flashy in a sales demo. If you skip that homework, you end up with a bunch of bells and whistles you'll never use. So build the list, then ask: what's a must-have, what's nice-to-have but not essential, and what doesn't matter at all. How about the data and analysis side — sometimes the same tool, sometimes different? Your analysis tool needs to help you find real trends and patterns in the data — Drivetrain, for instance, lets you dig into that directly. No matter the tool, there's always some element of pulling things back into a spreadsheet for extra work. So the real questions are: how well does it integrate with everything else you're doing, how easily can you get into the data and ask questions, how often is it refreshed. On AI and automation specifically — what should people look for there? AI tools are increasingly good at helping assemble graphs, generating summaries, drafting basic variance commentary. The real test is whether it's making your life genuinely easier through natural language, or just there to look impressive. Some capabilities add real value; others you try once and think, I probably won't use that again. It comes down to asking which of your actual processes AI can meaningfully improve. And then there's collaboration — everyone knows the concept, but people rarely notice it until it's missing. Any collaboration horror stories, or things you look for there? Collaboration is something you genuinely want built in — can you comment on charts, on specific data points, link out to others easily, get alerts on changes? This gets overlooked constantly, but if your presentation, data, and AI are all great while collaboration is weak, the tool just won't get used. You want it to be effortless to tag someone — "this number looks off" or "why is this down" — and get a quick response, enabling people to actually act on what they're seeing.
- 24:55
Kirk Kappelhoff
Agreed — that quick-tag capability is huge. Let me show you how we've approached this at Drivetrain — can you see this executive overview, Paul?
- 25:05
Paul Barnhurst
Yep, I see it — perfect.
- 25:08
Kirk Kappelhoff
A few things that have worked really well in these dashboarding reports: one-click timeline switching — last quarter, current quarter, year-to-date — with all charts updating in tandem. That matters because you often want to archive a specific version, say Q1 versus Q2, locking in that exact timeframe. On analyzing the data itself, we've seen two main patterns: aggregating (rolling up to a time level — quarterly, yearly) or bifurcating. Bifurcation is really where the insight lives, and there are two approaches: jumping straight to the underlying transaction-level detail — here's the lowest-level detail behind this month's total, sortable, filterable, exportable to Excel — or drilling down one dimension at a time, say by region first, then by country, then down to specific contracts in a given country. That commenting box up top lets you tag people directly, pinging them in Slack, Teams, or email at either the chart or cell level — something we've heard is a major ask from prospective customers. And exporting — PDF download, or more commonly, scheduled report delivery: tag a few people, set a subject and message, choose a cadence like Monday/Wednesday/Friday mornings, and the report just shows up. Is that roughly what you meant by drill-down and commenting?
- 27:15
Paul Barnhurst
Yes — line-item commenting especially is valuable, plus free-form notes. In a lot of ways this resembles a pivot table people already know — rows, columns, different dimensions. The easier that is, the easier real analysis becomes, like writing monthly variance commentary or answering "why doesn't this country's number look right" by clicking straight into the underlying contract and tagging someone about it.
- 27:55
Kirk Kappelhoff
Spot on. Let's move to presenting the dashboard — storytelling. I find this gets skipped a lot; people just jump into a report and read off numbers on screen. There's a great line: never tell a number without a story, never tell a story without a number — probably truer for CFOs than anyone. What do we mean by the "what, so what, now what" framework?
- 28:20
Paul Barnhurst
It goes back to the seventies — I don't recall who coined it, but the idea is: most of us in FP&A are good at explaining the "what" — we missed revenue by two hundred thousand, volume was down. Fine, but so what? Why does that matter, what actually happened? And the piece people want most is "now what" — what are we doing about it. Dashboards and storytelling are different things, but dashboards can support the story. The key in "now what" is offering an actual recommendation — data storytelling in finance is fundamentally about driving decisions, giving calls to action. If all you do is present data, you're falling short on that piece.
- 29:15
Kirk Kappelhoff
Your thoughts on this framework here?
- 29:18
Paul Barnhurst
We've laid out a more detailed step-by-step view of the "so what, now what" pieces that I follow a lot. I might start by setting the agenda and providing access in advance — to the report itself, or to the stakeholders who need to understand the assumptions behind it. Either way, setting that agenda mentally means you've started telling the story before you've technically begun. One of the frameworks I teach a lot in storytelling is ACE: Audience (know who you're talking to, set the agenda accordingly), Content, and Execution — your voice, body language, delivery. How many of us have sat through a presentation with fascinating content but a boring presenter, or the reverse — boring material but a fabulous presenter? It makes a huge difference. The second point I find genuinely valuable: adopt a positive tone focused on solutions. There's research showing people who focus on the negative and complain often show measurable brain changes — gray matter decreasing — while a positive tone can do the reverse. If I walk in with bad news and pile onto it, I'm not helping or being part of the solution. If I take a positive approach — "here's where we are, here's our plan to address it" — that lands very differently. That doesn't mean lying: if you're fifty percent off with nine months left, you don't say "we'll definitely hit it, just work harder" — you focus on realistic solutions, being a realist but an optimist in tone. Walking in negative impacts everyone; we've all been in meetings that turn into a gripe session and help nobody. Having that call to action, that recommendation, ties directly into the positive-tone point. The other thing I've found valuable is backing claims with data — wishy-washy consultant-speak only gets you so far. If I make a claim, I need data behind it to be convincing. You need both the data and the narrative around it — there's a great book on how, when telling something, you're really replacing someone's existing story with a new one, blending data and facts with the impact of emotion. Storytelling with Data is fabulous on exactly that balance — data alone isn't enough, but paired with the right narrative and emotional framing, you get much higher success rates. This next one — delivering bad news — I've seen people struggle with, since nobody wants to be the messenger who gets blamed. My overarching rule: just tell it how it is. Most of the time it's genuinely not your fault — if there's a model error, own it and fix it going forward, but if it's a missed revenue target or department overspend, you're the one tracking it, not the one who caused it.
- 33:40
Kirk Kappelhoff
Agreed — our job in finance is to present numbers and own them alongside the business, not "you missed your numbers, what did you do wrong" but "we missed our numbers, here's what happened, here's an issue we found." Maintaining that transparency and framing it as "we" — team ownership, not individual blame — matters a lot.
- 34:05
Paul Barnhurst
Otherwise you get the classic pattern: "I" exceeded the numbers, "we" missed them.
- 34:15
Kirk Kappelhoff
On defining the numbers themselves — this ties in nicely, since clear definitions matter so much. I've seen teams disagree on something as basic as CAC — one team defining it as sales-and-marketing-this-month over new-customers-this-month, another insisting it should be sales-and-marketing-from-three-months-ago given a ninety-day lead time. That kind of mismatch creates real confusion, so having a slide that clearly defines each metric and its underlying assumptions — by region, customer, product, whatever's relevant — matters a lot.
- 34:50
Paul Barnhurst
Agreed — having a clear explanation of data sources, metric definitions, and visible formulas is critical so everyone's working from the same page. I'd bet everyone listening has been in a meeting where someone says "my number's different" because methodologies weren't aligned upfront. I remember a role where the CFO, CRO, and others literally couldn't agree on what constituted a "booking" — we had to resolve that fundamental definition before anything else in reporting or dashboarding could proceed. This is especially tricky in SaaS, where there's no GAAP-equivalent standard rulebook for these metrics yet. Lastly, comparability across periods — the metrics we track need to be clearly defined, transparent, and tracked consistently over time, otherwise there's no reference point to show growth or decline. One addition: don't be afraid to annotate big one-time items in the notes rather than let them distort the picture — think about how many of you had to report the year after COVID, wildly up or down depending on your business, almost nobody flat. Providing that commentary prevents a lot of confused questions later. And this last point is a newer one I've started adding recently —
- 37:28
Kirk Kappelhoff
Like you, Paul, I'm a huge user of AI tools and genuinely love them, but there's real nuance in the quality of answers you get, especially with data analysis — I've had it merge documents and miss line items or decimal points from PDFs. So double-check everything, especially as you lean into AI more — hallucinations are real, and you need to keep that in mind.
- 37:55
Paul Barnhurst
Hundred percent — think of it like an intern. You'd never hand your boss an intern's work without checking it first; same applies to AI. I love that framing — AI is the intern. On the flip side — the killer mistakes people typically make. Too much data, not enough context is one of my own historical weaknesses — I'm a detail person, and I'll tell a story that gets at exactly this. I'd spent six months on a project, fully aligned with my leadership team, everyone signed off, and we walked in to present to the global CFO. Ten minutes in, he says, "this is a terrible idea, why would we do this?" I was flabbergasted — everyone had approved it. He brought in the general manager, talked with him for five minutes, and ended up signing off after all. Afterward I asked my boss what happened, and he said: "Paul, your analysis was brilliant, your presentation lacked." I'd never brought the CFO along on the context — I'd spent so much time in the data that I only shared what I thought he needed, not the full story.
- 39:20
Kirk Kappelhoff
Interesting — so the story itself can actually detract from otherwise strong data if it's missing that context.
- 39:28
Paul Barnhurst
The next few mistakes tie into that same theme: no focal point or takeaway (the flip side of having a clear, solution-aligned recommendation); misaligned KPIs versus business goals, which comes back to showing the right stakeholders the right things; and overdesign or underexplanation. On misaligned KPIs specifically — the real issue isn't misalignment so much as actionability. Are they leading indicators? I've seen entire dashboards that are nothing but lagging metrics — fine if the goal is purely financial reporting, since those are inherently lagging by nature, but if the goal is a dashboard people can actually act on, you need real leading indicators in there. The next set is related: changing tracked metrics too frequently, not explaining the "why" behind number changes, and overly complex visualizations. On changing metrics too often — I remember a meeting where our boss wanted new metrics and we ended up with a list of forty or fifty. Every meeting somebody wants to change something, and it makes it impossible to focus. You don't change strategy and goals monthly — or if you do, you're probably not very successful. Set them for a year, revisit once or twice, adjust as needed. Metrics should follow the same discipline — give people real time to actually influence them, and if you picked them for the right reason in the first place, why keep changing them?
- 41:15
Kirk Kappelhoff
With all that theory in place, let's talk about the essential CFO dashboard suite. Any individual pieces you'd prioritize, or is this the full baseline every CFO should have?
- 41:30
Paul Barnhurst
I think this covers the core — there are two sides: the financial side, where a CFO needs visibility into cash flow and the core financials, and the operational side. You've covered that — sales and marketing are critical to nearly every business, plus expense analysis and operational metrics. Then there are vertical-specific dashboards, especially useful when the CFO needs to have targeted conversations with a specific department or FP&A lead.
- 42:00
Kirk Kappelhoff
Let me jump back into Drivetrain and show a couple of the ones we've seen work best from that essential suite. The thirteen-week cash flow has been, by far, the most important dashboard we've seen recently — driven by liquidity concerns and its inherent simplicity. Excel really struggles here — actuals versus forecast, the variance between them, and drilling into that — something notoriously difficult to do well in a spreadsheet. A dedicated tool makes that much easier, including drilling into the underlying calculations and figures, not just the formula logic. Similarly with budget-versus-actuals — drilling to the customer level for your own analysis, or letting AI answer that question directly. Here's something we've recently built and continue to expand: as I'm doing my own written analysis, I get the option to refine, shorten, or elaborate the text with AI, then drop it straight into the report — what used to take real manual time now happens almost instantly, with room to validate and adjust as needed. From there you can drill deeper — customer level, department level — or not, depending on the need. This next one's a specific use case tied to scheduled report delivery, typically sent out daily. On the P&L side, you'll typically see something like this: tabular data you can drill into, alongside charts calling out key figures over whatever timeframe you choose. What's valuable isn't just the charts — you can build charts anywhere — it's the ability to see the calculations behind them, comment directly, and click into the data at both a headcount and P&L level. Headcount is another area that's notoriously painful in Excel — adding and removing lines, start and end dates, promotions — whereas a purpose-built tool handles that natively. Whether you want to drill by department or track headcount movement over time, that's where the real answers live. Any thoughts or additions to that dashboard suite, Paul?
- 44:15
Paul Barnhurst
You covered it well. I'd just add that dashboards will look different for every business — take a three-statement model, for example: American Express, Google, or Amazon aren't running a traditional three-statement view the same way a smaller company might; they might focus on P&L plus a couple of balance sheet items. So keep in mind what we've shared is a general framework — every business is unique, and you have to decide what's actually most important and refine from there.
- 44:50
Kirk Kappelhoff
Looking forward to the questions coming in — before we get there, one more poll: would you like to explore a Drivetrain demo based on what you've seen, and did you find this webinar a good use of your time? We're always trying to improve these. On Drivetrain itself — here's a quick overview of some of our customers; we're currently the highest-rated FP&A tool on G2, with some customer feedback shown here. We'd love to show you more about consolidating data, the modeling side (which closely mirrors Excel if you're an Excel person yourself — I certainly am), and the reporting side we've covered today. A few resources: we're always glad to have these conversations about the right way to model — as you mentioned, that varies company to company; sometimes indirect cash flow is the right approach, sometimes direct, sometimes people don't even know the difference. Either way, we're happy to have that conversation and share materials. Let's move into Q&A — Paul, anything jumping out at you from the chat?
- 45:50
Paul Barnhurst
A few questions here — one in particular: how do you decide the balance between leading and lagging metrics for sales and revenue dashboards? Good question. For me it starts with what metrics they can actually influence — always lead with those — but you also need the lagging financial metrics for full context and to understand actual performance. So start with the leading indicators people can act on, then layer in the lagging metrics that complete the picture. The real question to keep asking is: what's most important, what do they actually need to make a decision? Sometimes that's only one or two lagging indicators, sometimes it's mostly leading ones — though you'll generally want at least some lagging metrics, since you'll likely need the actual P&L figure alongside bookings and billings and the differences between accounting and sales views.
- 46:50
Kirk Kappelhoff
I think I'd do that answer a disservice trying to add to it — that was great. Here's another one, from Kyle: what strategies would you recommend to generate engagement with dashboards? My senior leadership champions data-driven decisions but I struggle to get their attention.
- 47:10
Paul Barnhurst
One principle that's stuck with me: BLUF, bottom line up front. Give people what they need to know immediately — any recommendations — then let the supporting graph or detail follow below if they want more. I remember someone sending monthly Excel reports that nobody ever opened, until a colleague told them: give a great summary, lead with the bottom line, make the email quick and actionable — and once they stopped sending the full report, people started actually asking for it. So ask yourself: are you just delivering data, or delivering a story, an action, a recommendation, and a summary that lets people act quickly? If not, you'll likely keep seeing low engagement. Another thing I've seen work well: whatever people ask questions about is usually what they're most interested in — lean into that. I ran a weekly report for one client where two or three specific slides kept generating all the interest, so I made the presentation more about those, going deeper — it became one of my favorite meetings of the week, and clearly more valuable for them too. Another one, from JK: what are the trade-offs in leaving out low-impact KPIs, when they technically check boxes like actionability, ownership, benchmarks, forecasting relevance, controllability, and strategic alignment? How would you rank these in importance generally? They're definitely not all equally important — there's always a hierarchy. I'd start with: is it actionable? If a report is meant to drive action and something doesn't support that, it goes in an appendix or a secondary dashboard. Second: is there clear ownership? If nobody owns the underlying data, that's a problem to fix even before deciding whether to present it. Benchmarks I'd treat as table stakes — you should be able to compare against a forecast, budget, or prior year by default, not something to actively rank high. Forecasting relevance is more business-specific, harder to generalize. Strategic alignment should be a starting filter — if there's no tie-back to strategy or company goals, even loosely, it's genuinely low-impact. Another one: can we create views for different departments or roles in Drivetrain, and control data visibility?
- 50:20
Kirk Kappelhoff
Absolutely — that's largely role-based access control, but I'll take it. On a headcount or vendor model page, you don't want everyone seeing everyone's salary, so you can restrict that down to the line-item level by department, cost center, or region — fully personalized. One feature we've recently released: as an admin, you can view the tool as if you were a specific user without admin access, so I could view it "as Paul" and confirm the permissions are set correctly. Speaking of salary visibility — Paul, you're in trouble here.
- 50:55
Paul Barnhurst
Guess I won't be living in Drivetrain much longer.
- 50:58
Kirk Kappelhoff
On the P&L side, you'd also filter by entity or currency — more a visibility control than a permissions one. About eighty percent of our customers are multi-entity, seventy percent of those multi-currency, so you'll want to slice by intercompany transactions, eliminations, or specific entity mixes. So yes, the answer is a resounding yes on custom views. Another one: can we discuss analyzing cohorts — it's used across industries. Do you have strong opinions there, Paul? I can also show something in Drivetrain.
- 51:35
Paul Barnhurst
Cohorts are critical, and the key is understanding what actually distinguishes one cohort from another. Often people default to date-based cohorts, which is fine, but if that's not the meaningful driver, maybe customer size or geography defines your cohorts better instead. First, dig into the data enough to understand what genuinely differentiates one cohort from another; then make sure you can show that clearly, since people want to understand what's driving a cohort's performance up or down and how to influence it — whether that's a revenue waterfall, a churn cohort, or something else entirely.
- 52:20
Kirk Kappelhoff
I've done a lot of cohort modeling myself, and it gets genuinely hard in Excel. We've simplified this in Drivetrain around two core components: the "horizontal" — new cohorts starting each month, usually informed by historicals but still a judgment call — and the retention curve, describing how each cohort behaves after inception. You can adjust the starting cohort size for any given month, and that ripples through; if I bump a cohort up two hundred percent, every subsequent month tied to that cohort doubles accordingly. That alone gives you the customer-count basis of the cohort, but if I layer in price — say fifteen dollars flat — I can duplicate that cohort line as "revenue," multiply the price by the cohort-months figure, and get a revenue cohort view instead, in whatever currency makes sense. The underlying formula essentially says: if this month is after the cohort's start, apply the retention curve; if it's the starting month, set the base number; otherwise zero. We've simplified that into a straightforward multiplication rather than the nested IF/INDEX-MATCH formulas most people build in Excel — a lot of our SaaS customers use exactly this kind of structure, and it's dramatically easier than doing a forty-year cohort spend-down by product entirely in spreadsheets, which was genuinely one of the worst modeling exercises I've ever had to manage.
- 54:15
Paul Barnhurst
I'll happily leave that one to you all — in Excel, no less. Another question: how do you avoid decision paralysis when the data says one thing but leadership intuition says another — data versus gut, what are your thoughts? There's the old joke: "in God we trust, all others bring data" — but there's genuinely something to intuition too. When they don't align, dig in initially, but at some point, with experienced people, you have to decide what you trust more, since we've all seen cases where data was wrong and cases where gut instinct was wrong. Most of the time you should lean toward trusting the data, but there are moments where going against it is the right call. The key is knowing when to stop the analysis-paralysis loop — give yourself a defined window, say two weeks or forty hours, then make the call. If you're never willing to override data with judgment, that's a problem; if you're purely data-driven with zero room for judgment, that's also a problem — sometimes gut instinct really is the right signal.
- 55:40
Kirk Kappelhoff
I've gone against data before myself, usually when I suspected it wasn't clean — and when data's clearly wrong, I won't let it distract from otherwise clean analysis. Outside that, I'd say I mostly follow the data.
- 55:55
Paul Barnhurst
Agreed — that should be the default most of the time, with exceptions for gut instinct or genuinely dirty data. If you're not following the data most of the time, though, it's worth asking yourself why.
- 56:10
Kirk Kappelhoff
As we wrap up, Paul — one thing you'd leave everyone with?
- 56:15
Paul Barnhurst
Start simple — applies to financial models, dashboards, everything. I heard someone say once: start simple, because your business will make it complex on its own. When building a dashboard, focus on what's genuinely important, keep it simple, and accept that people will always ask for more — then decide deliberately what actually deserves to be added. Be willing to have open, constructive conversations — it's okay to say "I don't think that fits here, but here are some alternatives" as long as you can back it up. Ultimately, keep having those conversations so what you deliver genuinely has impact.
- 56:50
Kirk Kappelhoff
That's a great place to end. Paul, thank you so much as always — it's a pleasure having you, and we'll definitely have you back, maybe to dive into something more modeling-focused and less dashboard-y next time. Really appreciate your time.
- 57:05
Paul Barnhurst
Pleasure to be here, thank you everyone — really enjoyed this. Feel free to reach out to Drivetrain to learn more about the product, or reach out to me directly. Appreciate everyone carving out time today — always a pleasure. Thanks, Kirk, see you again soon.
- 57:15
Kirk Kappelhoff
Thanks, Paul.





