Variance Analysis the CFO Actually Reads

Most variance reports are 14-page slide decks nobody opens. The analyst spends two days color-coding cells red and green, builds a 60-row reconciliation table, drops it into the monthly close folder at 11 PM Sunday, and on Monday the CFO scrolls to slide 3, asks "so what?", and nobody has an answer. The deck gets archived. The CFO emails the controller. The controller emails the analyst. The analyst rebuilds half of it from memory in a Slack thread.

This is the variance analysis trap. The fix isn't more detail. It's less.

A CFO who runs a real business doesn't want every line item color-coded. They want three bullets and a recommendation. They want to know what moved, why it moved, and what you're doing about it. That's it. Everything else is appendix.

This playbook walks through how to write that. Materiality thresholds so you stop chasing noise. Price/volume/mix decomposition so revenue commentary survives the first follow-up question. A root-cause framework so you name the decision, not the line item. The three-bullet executive format. And the no-surprises rule that protects your career when a $400K overrun lands in front of the exec team.

Set the Materiality Threshold Before You Start Writing

Before you analyze a single line, define what "material" means. Publish it. Get the CFO to sign off on it. Now you have an answer when somebody argues a $40K underspend deserves its own slide.

The standard rule for an FP&A team running a $50M-$500M business:

Material variance = 5% of plan OR $50K absolute, whichever is greater. Set per cost center.

The "whichever is greater" part is the trick. A flat percentage doesn't work because 5% of a $20M revenue line is $1M, but 5% of a $200K travel budget is $10K. You'd be writing a slide on a $10K travel overrun while a $300K revenue miss falls below your threshold. Wrong direction.

A flat dollar amount doesn't work either. $50K is a rounding error on enterprise SaaS revenue but a third of your marketing budget on a small product line.

Here's a worked example. Your plan looks like this:

Apply the rule. Enterprise revenue: -0.25% and -$50K. Below threshold on both. Ignore. SMB revenue: -7.5% and -$300K. Material. Lead with this. Marketing opex: +20% and +$80K. Material. Travel: +15% but only +$12K. Below the $50K floor. Ignore. Engineering payroll: tiny on both axes. Ignore.

You went from five lines to two. The two that matter.

The CFO can argue with your threshold once. They can't argue with it every month if you've published it as policy. Write a one-page memo: "Variance analysis materiality, FY26. We report and explain variances of 5% of plan or $50K, whichever is greater, per cost center. Below-threshold variances are tracked but not commented on unless they constitute a trend across three or more periods." Email it to finance leadership. Reference it in every variance pack. Done.

Decompose Revenue into Price, Volume, and Mix

"Revenue is down 8%" is not analysis. It's a temperature reading. The CFO will ask the obvious follow-up: "did we lose customers, or did we lose pricing?" If you can't answer that in the same breath, you don't have variance commentary, you have a number.

Price/volume/mix is the only revenue decomposition that survives a real CFO question. Every FP&A IC needs it in muscle memory.

The math:

• Price variance = (Actual price − Plan price) × Actual volume
• Volume variance = (Actual volume − Plan volume) × Plan price
• Mix variance = the residual when product lines have different margins or price points

A worked example. Your SaaS business plans Q1 like this:

Total new ARR is down $400K, or 6.25%. The single-line variance commentary would say "new ARR unfavorable to plan by $400K (6%) due to softness in new business." That's the kind of sentence that gets the deck closed.

Now decompose:

Enterprise price variance: ($75,000 − $80,000) × 50 = -$250,000. Enterprise ACV compressed $5K per deal. Probably from larger discounts at the deal desk, longer ramp schedules, or downsizing during procurement.

Enterprise volume variance: (50 − 50) × $80,000 = $0. We hit our deal count.

SMB price variance: ($12,500 − $12,000) × 180 = +$90,000. SMB ACV is actually up. The new self-serve tier may be pulling slightly larger initial commits.

SMB volume variance: (180 − 200) × $12,000 = -$240,000. We missed SMB by 20 logos.

The real story:

New ARR missed plan by $400K. Volume held in enterprise (50 of 50 deals) but price compressed $5K per deal (this is the deal desk policy working as designed). SMB hit price ($500 above plan per logo) but missed volume by 20 deals; SDR team flagged a top-of-funnel softness in February that we'll see resolve in April.

That's a totally different conversation. Volume held on the high-ACV motion. Pricing strategy is doing what we said it would do. The miss is concentrated in SMB demand-gen, which is fixable. The CFO can take that to the CEO without three follow-up questions.

The mix variance comes in when product margins differ. If enterprise gross margin is 78% and SMB is 65%, a deal-mix shift toward SMB lowers blended margin even if total ARR is on plan. Always check whether your variance is dollars or margin. They tell different stories.

Build a price/volume/mix worksheet once and reuse it monthly. Don't recompute the formulas from scratch every close.

Use the "Why Did This Happen?" Framework — Root Cause, Not Symptom

Here's a sentence that adds zero value: "Marketing overspent $80K." The CFO knows that. They can read the variance column. What they don't know is why.

Apply the five-whys. Keep going until you hit a decision somebody made or an external event that actually happened. Stop when the answer names a person, a date, or a strategic call.

Bad commentary: "Marketing overspent plan by $80K."

One why: "Marketing overspent plan by $80K due to higher demand-gen spend."

Two whys: "Marketing overspent plan by $80K due to higher demand-gen spend in March."

Three whys: "Marketing overspent plan by $80K because we accelerated demand-gen into March."

Four whys: "Marketing overspent plan by $80K because we accelerated demand-gen into March to cover a Q1 pipeline gap."

Five whys: "Marketing overspent plan by $80K because we pulled forward Q2 demand-gen to cover the Q1 pipeline gap caused by the SDR hiring freeze in November."

Now the CFO has something to act on. The overrun isn't a marketing problem. It's the cost of plugging a sales-staffing decision from four months ago. That changes the conversation in the exec review from "rein in marketing" to "the SDR ramp is six weeks behind, here's what it cost us."

Name the decision. Name the date. Name the person who made the call if it's politically safe to do so. Variance commentary that names "macro headwinds" or "general softness" or "timing" is the corporate-finance equivalent of saying "stuff happened." It tells the CFO you didn't do the work.

Write the Three-Bullet Executive Variance

Every material variance gets three bullets in the exec deck. Not a paragraph. Not a slide. Three bullets.

Bullet 1, What moved: number, direction, materiality threshold. Bullet 2, Why: root cause, named decision or named external event. Bullet 3, What we're doing about it: action, owner, timing.

That's it. If you can't fit it in three bullets, you don't understand the variance yet.

Three real examples.

Revenue miss (before / after)

Before:

Q1 revenue was unfavorable to plan by $1.2M (4.2%). The miss was driven primarily by softer-than-expected enterprise renewals as well as some delays in expansion bookings. The team is monitoring the situation closely and we expect to recover in Q2 as several large opportunities progress through the pipeline. Additional color is provided in slides 7-12.

After:

• Q1 revenue missed plan by $1.2M (-4%). Volume held; the entire miss is enterprise renewal price compression averaging 6%.
• Driver: new deal desk discount policy launched January 15 is producing larger discounts on multi-year renewals. Working as designed; we accepted lower ACV in exchange for longer terms (avg term up from 18 to 27 months).
• Action: tracking renewal NRR through Q2 to confirm thesis. No corrective action proposed. CFO + CRO review scheduled May 12.

Opex overrun (before / after)

Before:

Sales & marketing opex came in $200K above plan in March, primarily driven by elevated contractor spend and unbudgeted event costs related to our Q1 industry presence. We are reviewing our discretionary spend processes to ensure tighter controls going forward.

After:

• S&M opex over plan by $200K in March (+8%): $120K contractor overrun, $80K unbudgeted event sponsorship.
• Driver: contractor overrun is the agency we hired in February to cover for the open SDR roles (decision Lisa made Feb 8 to keep top-of-funnel intact). Event sponsorship is the SaaStr add-on the CRO approved on March 3.
• Action: contractor spend tapers in May as two of three SDR hires close (offers out). Lisa to confirm full tapering by May 30. No further events through Q2.

Headcount underspend (before / after)

Before:

Engineering payroll was favorable to plan by $180K in Q1 due to slower-than-anticipated hiring. The team continues to actively recruit for several open roles.

After:

• Eng payroll favorable $180K in Q1 (-3%): three open backend roles still unfilled at quarter-end (planned to fill by Feb 15).
• Driver: VP Eng paused offers in mid-Feb after CTO redirected the team from infrastructure to AI agent platform. Re-scoping JDs took six weeks.
• Action: revised JDs went live April 18, two offers out this week. Underspend reverses in Q2 if both accept; full Q2 catch-up unlikely. Updated forecast in flash report.

Notice what's gone: hedge words, fluff, "we're monitoring," "primarily driven by." Notice what's there: dates, names, the specific decision, the specific action, the specific person on the hook.

That's the format. Memorize it.

Walk Every Variance Past the Business Partner First

The single biggest career multiplier for an FP&A IC is the no-surprises rule. Every material variance gets a conversation with the budget owner before it lands in the CFO deck. Always.

The script:

"Hey, I'm seeing a $120K overrun on contractor spend in your March actuals. Before I write commentary on it, can you walk me through what happened? I want to make sure I'm telling the right story to finance leadership."

That's it. Open question. No accusation. You're collecting context, not auditing them.

What you get back is the real story, the kind that doesn't show up in NetSuite. "We held two contractors over from February because the FTE backfill slipped." "The agency invoice for January came in late and got booked to March." "I authorized the spend because the alternative was missing the launch deadline and the CRO was good with it."

Now you can write commentary the budget owner will not flinch at when it appears in the exec deck. They've already seen it. They've already approved the framing. If the CFO challenges the variance in the meeting, the budget owner has your back instead of being blindsided.

Three tactical rules for partner pre-reads:

1. Do them before you finalize the deck, not after. If your deck goes out at 9 AM Monday, the partner conversations happen Friday afternoon. Not Monday at 8 AM.
2. Send the actual draft commentary you intend to publish. "Here's what I'm planning to write — does this match what happened?" Not a paraphrase. The exact words.
3. Document who you talked to and when. Not for politics. For your own forecast accuracy. Variances explained by the actual operator are 2-3x more accurate than variances you guessed at from the GL.

Partners who get blindsided in exec reviews stop trusting finance. Partners who never get blindsided forward your commentary to their team and tell them to read it. That's what you want.

Write Commentary That Explains Decisions, Not Just Numbers

There's a difference between accounting-speak and business commentary. Accounting-speak describes what happened to the numbers. Business commentary explains the strategy that produced the numbers.

Bad: "Revenue unfavorable to plan by $1.2M (4%)."

Better: "Revenue down $1.2M (4%) on enterprise renewal price compression."

Best: "We held volume but lost $1.2M to enterprise discount expansion. The new deal desk policy is working as designed; this is the cost of the longer-term commitments we'll see compound in NRR through 2027."

The third version ties the variance to strategy. It tells the CFO: this isn't a problem, this is a planned trade-off, here's the upside that justifies it. The CEO reads that and nods. The CFO doesn't have to defend the number to the board because the number defends itself.

Every material variance has a decision behind it. Find the decision. Write the commentary in terms of the decision. If there is no decision behind it (if it's just a slip), say that too. "We missed because we missed. The team didn't execute on Q2 renewal outreach. CRO is on it." Honest beats spun every time.

What to Leave Out

Things that do not belong in variance commentary:

• Below-threshold line items. If it doesn't clear materiality, don't comment.
• Year-over-year comparisons when the business model changed. If you raised prices, launched a new product, or restructured cost centers, YoY tells you nothing useful.
• Color-coded heat maps with 80 cells. Nobody reads them.
• Anything the CFO can derive from the three bullets. Don't repeat yourself.
• Hedge language: "we're monitoring," "trends are mixed," "primarily driven by." Either you know why, or you don't.
• The variance to budget AND the variance to forecast AND the variance to last year, all in one table. Pick one anchor (usually current forecast for in-year, plan for board reporting). Footnote the others.

The goal is signal, not coverage.

Templates Worth Reusing

Three-bullet variance template:

[Line item] [direction] plan by $X (Y%):
• [What moved — number, threshold, scope]
• [Why — named decision or named event, with date]
• [Action — what, who, by when]

Materiality threshold worksheet: one row per cost center, columns for plan, threshold % (default 5%), threshold $ (default $50K), and "report?" calculated as a flag.

Price/volume/mix sheet: one tab per revenue stream. Inputs: plan price, plan volume, actual price, actual volume. Outputs: price var, volume var, mix var (if multi-product). Reuse monthly.

Partner pre-read script: "I'm seeing a $X variance on [line]. Before I write commentary on it, can you walk me through what happened? I want to tell the right story." Send Friday for Monday review.

The Promotion Test

Here's how you know your variance commentary is working. The CFO opens the deck. They read your three bullets. They forward the email to the CEO without rewriting a word. The CEO reads it on their phone between meetings, replies "got it, thanks," and that's the end of it.

When that happens, you've done it right. You've turned two days of close work into thirty seconds of executive understanding. You've named the decision instead of the line item. You've protected the budget owner from a surprise. You've explained the strategy, not just the number.

That's the skill that gets a financial analyst promoted. Not Excel. Not tools. Not modeling speed. The ability to walk into an exec review and have the CFO say "Camellia's commentary covered it, let's move on."

Three bullets. Beats fourteen slides. Every time.

Learn More

• Building a Three-Statement Model from Scratch
• Board Prep: Numbers Into Narrative
• FP&A Metrics: Forecast Accuracy, Model Velocity, BP NPS
• Path From Financial Analyst to FP&A Manager
• Financial Analyst Job Description Template