It’s the night before the board meeting. And the model’s wrong — again.

You thought it was final, but someone copied the wrong tab. Now the hiring plan doesn’t match the revenue plan, and the burn rate’s off. And you’re rechecking formulas at midnight.

Runway gives you one model and one source of truth. Everything stays up to date, so you’re not scrambling the night before. You’re ready the moment the question is asked.

That’s why teams like Superhuman and 818 Tequila trust Runway to walk into QBRs calmly.

See Runway in action

“You are like a bridge scientist!”

I meant it as a compliment.

Sandra was my first boss in industry. I'd left the Big 4 six months earlier for this gig. I'd been lucky to bag a great FP&A role supporting regional operators with their P&Ls despite having zero FP&A experience.

Sandra had put a couple of hours aside to show me how to improve the variance analysis and performance bridges I had been building. Which was another way of saying mine had been absolute crap.

She pulled out a notebook where she had handwritten how she defined and analyzed all the different lines on a performance bridge:

…to name just a few.

This notebook was a goldmine.

It was as detailed an operating system for the ‘A’ in FP&A as I’d ever seen. This was before the days of smartphones, so a photo wasn’t an option, and there was no copier nearby. So I scribbled down the key points as fast as I could.

I assumed everyone in FP&A was a "bridge scientist" with one of these notebooks, so I started maintaining my own.

But as I moved to other companies, I realized that no one else did this. Even in big companies, FP&A was more art than science. There was an alarming inconsistency in how language was used. When someone said "full year impact," did they mean 52 weeks? Current financial year? Something else entirely?

I came to realize how unusual Sandra's approach was. And how valuable.

When I became CFO and got the whole finance team to talk about variances in a consistent language, it was transformative. It was a force multiplier for the quality of the financial storytelling throughout the whole team.

It accelerated understanding and helped the businesses make decisions, and therefore, take action faster.

Welcome to part 7 of this 9-part FP&A series. This week, we are talking about how to build a variance bridge. This has been on my topic list since the day I started this newsletter. I wanted to wait for a special moment to unleash it. Today is the day.

But first, let’s recap the series so far:

Week 1 - Overview of FP&A - The FP&A Cycle

Week 2 - Long Range Planning

Week 3 - Budgeting

Week 4 - Monthly Management Accounts

Week 5 - Monthly Performance Reviews

Week 6 - Forecasting

And just in case you didn’t believe just how near and dear to my heart this topic is… as I started to write this, I realized I had way more to say than I could ever fit in a single post. So, I’ve had to adjust the plan for the rest of the series to accommodate.

So consider this the first of a 3-part series-inside-a-series on variance analysis:

1) This week, we’ll go through some key principles for building good bridges

2) Next week, we’ll get in even deeper and build a common system for preparing a bridge

3) Then, in the final week of this MEGA-series (no, I haven’t tired of caps lock even after 7 weeks), we will talk about how to keep those performance adjustments fair and ethical.

Ok, now that we’re on the same page, let’s get tactical. Super tactical.

Every finance team creates bridges. Most do it, at best, inconsistently. And at worst, just plain badly. Analysts are learning the wrong way from their managers. The managers learned the wrong way from people that are now VPs. Etc Etc. So this drive for consistency needs to start at the top.

I was very lucky to work for someone who took this sh*t SERIOUSLY in my first-ever role. And 20 years later, I have spent way too much time thinking about it.

What is a ‘bridge’?

A bridge is just a way of explaining how two numbers compare to one another, e.g., this year to last year, actual to budget, current run rate vs previous run rate, etc.

That could look like this:

Or in a ‘waterfall’ format like this:

Or, we can even get 2-dimensional like this:

And there are a bunch of other more wacky variants too.

What can bridges be used for?

The reason CFOs love these so much is their versatility. They can be used for:

1) Performance Variance Explanation

2) Strategic Planning & Board Communication

3) Investment or M&A Narratives

4) Revenue Walks

5) Departmental or Functional Reviews

6) People/Headcount Planning

7) Fundraising/Investor Communications

And they can be used in almost any setting too.

Bridges can be as powerful on an earnings call, in a boardroom, town hall, or even just giving your sales team a kick in the ass.

Used right, they can do all of those things. Their versatility as a storytelling device is their strength.

The role of bridges in storytelling

We recently covered storytelling in our April reporting series. But no compelling narrative survives a weak fact base. That fact base is the bridge.

Put simply, shit bridge… shit story. In fact, well-told stories on garbage analysis are dangerous. Because they are both false and convincing (ahem… LinkedIn… ahem).

Your bridges are the script for CFO storytelling. And your stories can only be as good as the script behind those stories.

Is the format of the bridge you choose important?

A bit. But format is 5% of the game. The thinking behind it is 95%.

I’ll take a bullet-point list of well-thought-out variances on a white slide over a polished investment banking-grade bridge presentation with lazy surface analysis, every single time.

Where most bridges fail

The key to a good bridge is to focus on business drivers and actions, more so than the line-by-line financial statement variances.

Let’s take an example.

We have a business with a P&L that looks like this:

This would be an example of a terrible bridge for a P&L that looks like that:

It offers no insight whatsoever, merely showing headline P&L variances in a different format, which in isolation mean nothing. While this is a particularly bad example, most bridges I see are closer to this than where they need to be.

Worked Example

Let’s develop this example further. The surface conclusion might be that we are selling well. But our margins are a problem, as are our expense lines.

We have a cost problem, right? Let’s keep digging…

Now, we overlay the units sold.

Turns out we actually sold 2,500 units in the period vs a budget of 1,800. That's +700 (+39%). This changes the picture, as it becomes clear gross profit per unit have been diluted significantly, partly compensated by the benefit of extra volume:

This insight changes the shape of the bridge significantly. Now it might look more like this:

Is this an operational issue?

Let’s start with the VP of Ops to see what the hell is going on.

“The sales team forced through a one-off 1,000 unit order in the last week of the month to close a volume gap and hit their quota. We moved heaven and earth to get that order shipped before the end of the month. Doing so with only $25k of overtime expense. It might have even unlocked a new level of capacity for us. I didn’t know we could be that efficient. I don’t see the sales pricing, so not sure how that order was priced, but I hope it was worth it.”

You thank her and speak to the VP of Sales, who corroborates the story:

“Yes, we firesold some inventory to that one customer to hit our sales quota. We accelerated a sale from next month into this month.”

This gives the story a totally different complexion. Now let's look at the full unit economics vs budget:

How you attribute overhead variance, etc., to the volume can get complicated here (we’ll save that for another time). So we’ll keep it simple and assume that all sales variances are assessed against the budgeted margin.

If you break down the variance by driver, you build this picture:

This story is completely different. Three big things are going on here:

1) The business missed its core volume target, which cost $67 of EBIT

2) An ill-informed ‘firesale’ cost $378 at standard margins

3) But the business coped well operationally and managed to mitigate much of the effect. Overperforming standard margins by $269

At the start of this example, it had looked like a cost problem. But as we’ve dug deeper, that isn’t the case at all. There are three takeaways that would feed into the storytelling for the business:

1) There is a core sales volume issue to fix - that is the root cause problem in the business

2) The decision-making in response to that core volume problem (firesale) was dreadful. It cannot happen again.

3) On the plus side, the business’s ability to respond and scale was very impressive, mitigating some of the poor decision making. But more importantly demonstrating that the business operations are ‘ready to grow.’

With the right storytelling, the focus is on a completely different part of the business. Much more actionable.

It’s an extreme case, but it makes the point: superficial analysis leads to bad framing, misdiagnosis, and ultimately the wrong action.

A good bridge moves beyond surface-level analysis of P&L headline variances. A good bridge is analyzed by driver.

It even goes beyond the core atomic data of the business. Data alone isn’t enough. Data wouldn’t have told us that the sales team above accelerated a sale from next month. That needed dialogue.

A good bridge needs business context. It is not a spreadsheet exercise.

Sidenote: This is why I’m skeptical of AI-driven variance analysis tools, at least in the short term. Most CFOs still struggle to pinpoint root causes, let alone train a bot to do it. Because often the answer isn’t in the data, it’s in the conversation.

Time to find a common language

This is the exciting bit. Well… if you are a nerd like me, anyway.

The biggest problem with bridges is consistency. Which comes from a lack of a common language. A taxonomy.

So let’s start to build that here, in the rest of this piece, and next week. Courtesy of Sandra’s little book and the 20 years since during which I’ve learned and put this into practice.

Full Year Effects vs Part Year Effects

I hope there is a lot you can take from this piece. But if there is only one thing, make it understanding full-year effects vs part-year effects.

Let me give you an example:

Your business has 10,000 customers paying $20 per month. It is December 25, and you are preparing your 2026 budget. From September 1st, 2025, you implemented a price increase from $18 to $20 per month. You intend to implement a further $5 per month price increase from October 1, 2026.

Now let’s say we are bridging the 2025 full year actuals to the 2026 full year budget:

This would give you two bars on the year-over-year bridge:

1) The Full Year Effect (FYE) of the 2025 price increase on the 2026 numbers: $2 x 10,000 x 8 months = $160,000. The price was $18 from January 1st to August 31st 2025.

2) The Part Year Effect (PYE) of the 2026 proposed price increase on the 2026 numbers: $5 x 10,000 x 3 months = $150,000. You would also note here that the full annualized effect of the 2026 price increase would be $600,000 (with $450,000 being the FYE in 2027).

Here that is explained visually:

Does this matter?

Yes. It matters a lot.

FYEs have happened. They are sunk. A mathematical effect of a decision already taken. But a PYE is something that hasn’t happened yet. Meaning it could be subject to a change in decision, or execution risk, or a macro event or any number of other impacts.

This principle could apply to price increases, major customer wins or losses, cost price increases, or anything driving movement in your financials. And it has been particularly vital the last few years in analyzing inflation effects.

If you want your bridges to be real and actionable and your CFO storytelling to be more powerful, this is an important change. Because HOW you talk about an FYE and a PYE should be completely different.

A PYE is something you need the business to smash with some kind of verb: decide, execute, mitigate, ideate, etc. It needs commitment and action. A FYE is the opposite. It’s done. Beyond a tool for helping people understand the numbers, they are essentially backward looking (even if they have an effect in future periods.)

Bridges are the glue between planned and actual performance. They are the storyboard for your investor calls. The scripts for your internal narrative. And so much more.

It’s time we, as a finance community, got consistent in how we prepare them. That started today … we’ll continue next week.

:::::::::::::::::::::::::::::::::::::::::::::::::::
:: Thank you to our sponsor ::
:: RUNWAY ::
:::::::::::::::::::::::::::::::::::::::::::::::::::

If you enjoyed today’s content, don’t forget to subscribe.

Stay crispy,

The Secret CFO