Most variance commentary fails for one simple reason: it answers the wrong question.
Executives aren’t asking “what changed?” They’re asking:
- Is it real?
- Why did it happen?
- Does it change our outlook?
- What are we doing about it?
Good commentary is short, specific, and decision-oriented. It doesn’t narrate a table. It makes the numbers usable.
What is variance commentary in FP&A?
Variance commentary is the written narrative that explains the difference between actuals and a baseline (budget/forecast/prior period) and translates that difference into:
- drivers (why it moved)
- forward impact (so what)
- actions and ownership (now what)
It’s not a summary of lines. It’s a decision brief.
Why do executives ignore variance commentary?
Common reasons:
- It’s too long and reads like an audit trail
- It restates what’s already in the table
- It’s vague (“higher spend,” “timing,” “mix”) without quantifying drivers
- It lacks a point of view (“should we care?”)
- It has no action, owner, or forward impact
If commentary doesn’t help a leader decide, it becomes noise.
What’s the best structure for executive-ready variance commentary?
Use a consistent 4-part structure:
- What moved (quantified, one line)
- Why (2–3 drivers, quantified where possible)
- So what (forward impact: run-rate / quarter)
- Now what (action + owner + timing)
If you can’t do all four, you’re not done.
What’s a good one-paragraph template?
Here’s a structure that works in most exec updates:
[Line/area] is +$X vs [baseline], primarily driven by A (+$Y) and B (+$Z). This appears [run-rate / timing / one-time] in nature. If it persists, it adds +$Q impact to [quarter / next 90 days]. We’re taking [action] by [date]; owner is [name/team].
That’s enough for most leaders to act or ask the right follow-ups.
How many drivers should you include?
For execs: 2–3 drivers max.
If the variance truly has more components, group them:
- “Three smaller items totaling +$X” is better than listing seven minor drivers.
If you need more detail, put it in an appendix or a drill-down note for FP&A/budget owners.
How do you avoid vague language like “timing” or “mix”?
Don’t ban the words. Make them specific.
Instead of:
Write:
- “Annual renewal posted this month instead of next (timing shift of +$X)”
Instead of:
Write:
- “Higher mix of senior contractors vs plan (+$X) and regional mix shift (+$Y)”
The rule:
If a budget owner can’t recognize the cause from the sentence, it’s too vague.
How do you quantify “so what” (forward impact) without overcomplicating?
Use two simple lenses:
- Run-rate impact: “If this continues, what is the monthly/quarterly delta?”
- Timing impact: “Does it reverse next period or is it permanent?”
A clean phrasing:
- “Primarily timing; expected to reverse next month”
- “Run-rate shift; expect +$X per month unless actions are taken”
Executives don’t need perfect. They need direction and confidence level.
What’s the difference between commentary for execs vs budget owners?
Same facts, different packaging.
Exec commentary
- focuses on material movements, forward impact, actions
- 2–3 drivers, high signal
- answers “do we care and what are we doing?”
Budget owner commentary
- focuses on controllability and specifics
- includes what you need from them (validate drivers, commit to action)
- may include more detail if it leads to action
If your exec note is written like a budget owner note, it will be ignored.
How do you write commentary that’s consistent month to month?
Consistency comes from using a stable taxonomy of driver types.
Examples (for expenses):
- timing / accrual
- rate / pricing
- volume / usage
- mix (role/level/region/vendor mix)
- scope change (new tools, new projects)
- one-time true-ups
- reclass/mapping
When you tag explanations consistently, leaders start to trust the narrative because it feels comparable over time.
What are common “commentary failure modes” in expense workflows?
- The “category dump”
“Software +$120K” with no driver breakdown (renewals? seats? tier?) - The “headline without ownership”
“Contractors up” but no action/owner - The “late discovery”
Commentary written after close as a scramble, not built continuously - The “false certainty”
Overconfident explanations without validation - The “no decision”
Explains the variance but doesn’t say what changes next
How do you handle uncertainty in commentary without sounding weak?
State what you know, what you’re validating, and when you’ll confirm.
Example:
- “Early read suggests this is driven by renewal timing (+$X) and seat expansion (+$Y). We’re validating vendor invoice timing and will confirm by [date].”
Executives trust transparent confidence levels more than confident guesses.
What should you do when a variance is real but not actionable?
Say that explicitly.
Example:
- “This is a true run-rate increase driven by contracted pricing changes (+$X). No near-term lever this quarter; we’ll incorporate into the forward view and address at renewal.”
“Not actionable” is a valid outcome. It’s better than inventing an action to look busy.
What are the most common mistakes in exec variance narratives?
- Leading with detail instead of the headline
- Listing drivers without quantifying the biggest ones
- No forward impact (“so what?” missing)
- No action/owner/date
- Writing one narrative for all audiences
- Treating commentary as a reporting requirement instead of decision support
What does “great” variance commentary look like?
You know it’s great when:
- execs forward it without asking for “the real story”
- follow-up questions get narrower and more decision-oriented
- the forecast update becomes faster because drivers are already clear
- the same driver logic holds month to month
- commentary results in fewer recurring surprises
In other words: commentary becomes a control mechanism, not a writing exercise.
Quick FAQs
How long should exec variance commentary be?
Usually 3–6 sentences per major area, plus a short “top movements” summary.
Should commentary be written before or after the forecast refresh?
Ideally it’s built as part of the refresh so the “why” directly updates the forward view.
Is it okay to reuse commentary language month to month?
Yes—structure and driver taxonomy should be consistent. Just don’t reuse vague sentences without refreshed driver evidence.
How many variances should you cover?
For execs: typically 5–10 key movements. Anything more belongs in detail views.
