Variance Analysis

SkillCommerce & finance

Lets your agent break down financial variances between actuals, budget, forecast, and prior periods into quantified drivers with commentary.

Use Variance Analysis in Claude, ChatGPT or Ahel Desktop

Free. Sign in, add Variance Analysis and connect your AI. About a minute.

Also: Claude Code · Cursor · Codex

Then ask your AI: use the Variance Analysis skill

Details

Instructions available. Your AI can read the instructions. Execution depends on the setup they require.

Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.

Variance AnalysisStart free
About this skill

Decompose actual, budget, forecast, and prior-period financial variances into quantified drivers and management commentary. Use for price-volume-mix, rate, headcount, spend, margin, and waterfall analysis.

What this skill tells your AI

The instructions your AI receives, as published by rongxinzy/rongxinai in SKILLs/zhiyuan-expert-manager/presets/finance-accounting-expert/skills/variance-analysis/SKILL.md and read by ahel’s review.

Prepare the comparison

Confirm entity, account hierarchy, period, currency, unit, actual/budget/forecast versions, sign convention, and materiality. Validate that compared datasets use consistent scope and mappings.

Calculate absolute variance = actual - comparison and percentage variance using a documented denominator. Treat zero or sign-changing denominators explicitly rather than presenting misleading percentages.

Decompose drivers

  • Price/volume: volume effect = (actual volume - baseline volume) x baseline price; price effect = (actual price - baseline price) x actual volume.
  • Mix/rate: quantify shifts among products, customers, channels, regions, grades, or contract types.
  • Headcount: separate headcount, compensation rate, mix, hiring timing, attrition, bonus, and benefit effects.
  • Spend: separate fixed, volume-driven, discretionary, contractual, one-time, foreign-exchange, and timing effects.

The starting value plus all signed drivers must equal the ending value. Show any residual as an unresolved item, not “other” without explanation.

Narrative

For each material variance state:

  1. amount, percentage, favorable/unfavorable status, comparison, and period;
  2. quantified primary and offsetting drivers;
  3. the business cause and evidence;
  4. whether the effect is timing, one-time, structural, or uncertain;
  5. outlook, action, owner, and forecast implication.

Avoid circular statements such as “revenue increased because revenue was higher.” Distinguish proven cause from plausible hypothesis.

Waterfall and output

Limit a waterfall to a decision-useful set of drivers, verify it reconciles, and accompany it with a driver table. Rank follow-up by absolute impact, unexpected direction, recurrence, trend, controllability, and materiality. State all data limitations and required review.

Signals

GitHub stars
154
Forks
3
Last commit
Oct 2026
Advanced
Item type
skill
Key
variance-analysis-rongxinzy
Source
github.com/rongxinzy/rongxinai