Reconcile plan actual variance drivers

What it's for

Turns plan-versus-actual reviews into a reproducible explanation of interacting price, volume, cost, capacity, and timing drivers instead of an order-sensitive spreadsheet waterfall.

What you give it

Inputs split into evidence read from your connected systems, calibration your team owns, and numerical controls that affect precision but never the result's meaning.

{
  "drivers": [
    {
      "actual_value": 120,
      "category": "growth",
      "id": "volume",
      "planned_value": 100
    },
    {
      "actual_value": 12,
      "category": "growth",
      "id": "price",
      "planned_value": 10
    },
    {
      "actual_value": 7,
      "category": "cost",
      "id": "unit-cost",
      "planned_value": 6
    }
  ],
  "material_variance": 10,
  "reported_actual_value": 600,
  "reported_plan_value": 400,
  "value_terms": [
    {
      "coefficient": 1,
      "factor_ids": [
        "volume",
        "price"
      ],
      "id": "revenue"
    },
    {
      "coefficient": -1,
      "factor_ids": [
        "volume",
        "unit-cost"
      ],
      "id": "variable-cost"
    }
  ]
}

What you get back

This is the actual output of running the example above — computed by the same function the platform calls, not an illustration.

{
  "assumptions": [
    "The governed value model is a sum of declared multilinear terms evaluated once at planned and actual driver endpoints; the model, source ledger, units, scope, and period must be frozen before attribution.",
    "Each term's interaction change is allocated by the exact Shapley value of its planned-to-actual endpoint game, eliminating spreadsheet waterfall order but not proving that a driver caused the realized variance.",
    "Unmodeled drivers, changes in accounting classification, timing, scope, counterfactual, nonlinear functions outside the declared terms, and measurement error remain outside the decomposition.",
    "Driver and category contributions explain the submitted aggregate value model and must not be converted into named-person performance, compensation, hiring, or termination judgments."
  ],
  "category_diagnostics": [
    {
      "category": "growth",
      "favorable_variance_contribution": 310,
      "shapley_variance_contribution": 310
    },
    {
      "category": "cost",
      "favorable_variance_contribution": -110,
      "shapley_variance_contribution": -110
    }
  ],
  "configuration": {
    "base_value": 0,
    "favorable_direction": "higher",
    "material_variance": 10,
    "maximum_reconciliation_error_fraction": 0.01,
    "reported_values_supplied": true
  },
  "decision": "material_favorable_plan_actual_variance_explained",
  "driver_diagnostics": [
    {
      "absolute_attribution_share": 0.5238,
      "actual_value": 12,
      "category": "growth",
      "driver_change": 2,
      "driver_id": "price",
      "favorable_variance_contribution": 220,
      "planned_value": 10,
      "shapley_variance_contribution": 220
    },
    {
      "absolute_attribution_share": 0.2619,
      "actual_value": 7,
      "category": "cost",
      "driver_change": 1,
      "driver_id": "unit-cost",

How it works

Decision analysis — Turn uncertainty, cost and risk appetite into a defensible choice, with the reasoning left inspectable.

1. 1 Freeze one plan/actual period, source-ledger totals, driver endpoints, and a governed value model expressed as an intercept plus declared coefficient-times-factor-product terms.
2. 2 Evaluate every term at both endpoints and allocate each interaction change exactly across its factors using the Shapley value of the endpoint-change game.
3. 3 Reconcile allocations to modeled variance and the model to reported source values; abstain from explanation when the source reconciliation gate fails.

Before you trust it

Every tool in the catalog ships with the conditions under which its answer is meaningful — and the conditions under which it should abstain instead of guessing.

Calibrating it to your org

Calibration workflow

1. 1 Define the management decision, target outcome, aggregate unit, privacy boundary, cadence, and prediction/intervention horizon for this organization.
2. 2 Build a tenant-scoped historical cohort using only information available before each prediction or decision; preserve zero periods, censoring, assignment probabilities, and unresolved outcomes when the method requires them.
3. 3 Estimate statistical parameters on training history, but obtain costs, utilities, risk tolerance, practical-effect thresholds, capacity, and policy constraints from accountable decision owners.
4. 4 Validate on later time windows or held-out aggregate units at the deployment grain, against a simple baseline and the function-specific validation strategy.
5. 5 Deploy only if the returned decision clears evidence, overlap, calibration, robustness, and guardrail checks; warning, unsupported, schema-gap, and fallback decisions are abstentions.
6. 6 Monitor realized outcomes, data drift, coverage, and decision regret; recalibrate at a governed cadence or after a detected regime/definition change, never merely because a stakeholder dislikes the result.

Call it from your AI

You don't wire up 388 tools in your MCP client. The GitRevio MCP server exposes 18 tools, three of which let an agent search the catalog, read a tool's schema, and run it — so the assistant finds this one on its own.

gitrevio_capabilities_search
  { "q": "reconcile plantoactual value variance with an" }
  → finds "reconcile_plan_actual_variance_drivers"

gitrevio_capability_describe
  { "capability_id": "reconcile_plan_actual_variance_drivers" }
  → returns the input schema and agent guidance shown on this page

gitrevio_capability_run
  { "capability_id": "reconcile_plan_actual_variance_drivers", "arguments": { ... } }
  → returns the result shown above

Works in Claude Desktop, Claude Code, Cursor, Cline, Continue.dev, Goose and Aider. See the MCP server.

Related tools

See every tool in Causal evidence & experiments →