Variance Analysis for Humans (Not Accountants)

Your variance analysis is technically perfect. Every number ties. Every calculation is correct. Every account is properly categorized.

And nobody understands it.

The Communication Gap

Finance teams speak a language. Accountants understand it. Finance professionals understand it.

Board members don't.

When you present "favorable variance in SG&A driven by timing of marketing spend and accrual true-up," you're speaking accounting. You might as well be speaking Klingon.

Your audience nods politely. They don't actually understand. They move on.

What Humans Actually Want to Know

Non-financial stakeholders care about three things:

What happened? In plain English. "We spent less on marketing than planned."

Why? In causal terms. "The campaign we planned for May got pushed to July."

So what? In actionable terms. "We'll spend the money next month, so we're not really under budget for the quarter."

That's it. What, why, so what. Everything else is detail for people who ask follow-up questions.

The Jargon Problem

Finance jargon kills communication:

"Favorable/Unfavorable": Better to say "we spent less" or "we earned more."

"Variance": Better to say "difference" or "change."

"Timing": This word has lost all meaning. Say what actually happened.

"Accrual": Nobody outside accounting knows what this means. Skip it or explain it.

"True-up": See accrual.

"YoY/MoM": Say "versus last year" or "versus last month."

Jargon is efficient among experts. It's a barrier with everyone else.

The Narrative Structure

Good variance analysis follows a narrative:

Setup: Here's what we expected. "We budgeted $500K for marketing in Q2."

Conflict: Here's what happened. "We actually spent $420K."

Resolution: Here's why. "The product launch shifted to Q3, so we delayed the campaign."

Implication: Here's what it means. "Q2 looks good on paper, but Q3 will be over budget by the same amount."

Stories are memorable. Tables aren't.

Visual Communication

When in doubt, show it.

Waterfall charts: Show the flow from expected to actual.

Trend lines: Show whether the variance is a blip or a pattern.

Highlighting: Use color to draw attention to what matters.

Annotations: Add plain-English labels directly on charts.

A well-designed visual communicates more than a paragraph of text.

The Right Level of Detail

Different audiences need different detail:

Board summary: 3-5 key variances with plain-English explanations.

Executive discussion: Top 10 variances with context and implications.

Management deep-dive: Full variance detail for those who want to dig.

Most presentations err toward too much detail. Start high-level. Let questions drive depth.

Anticipating Questions

Good variance analysis preempts questions:

"Is this a one-time thing?": Address whether the variance recurs.

"Did we know this was coming?": Address whether it was expected or surprising.

"What are we doing about it?": Address planned response.

"Will this affect the forecast?": Address forward-looking implications.

If your commentary addresses these before they're asked, you've communicated well.

The Testing Method

Before presenting variance analysis, test it on someone outside finance.

Read your commentary aloud. Ask: "Does this make sense?"

If they say "sort of" or "I think so," rewrite it.

If they can repeat back what happened, why, and what it means, you've succeeded.

An Example Transformation

Before: "Marketing variance favorable $80K driven by timing of demand generation spend and SEO contract renewal delay, partially offset by unfavorable variance in events due to conference cost increases."

After: "Marketing was $80K under budget. We pushed some campaigns to next month, and our SEO contract renewal got delayed. Conferences cost more than expected, but not enough to offset the other savings. Net: we'll catch up on spending in Q3."

Same information. Different language. Different impact.

The Empathy Exercise

Remember: your audience has 47 other things on their mind. They're not going to decode your variance analysis. They're going to glance at it, extract what they can, and move on.

Your job is to make extraction easy. Plain language. Clear visuals. Obvious implications.

Variance analysis for humans isn't about dumbing down. It's about communicating up.

Your numbers are correct. Make sure your communication is too.