Allocations Quick Start

This walkthrough takes you from raw cost data to a working allocation model in roughly 30 minutes.

What You’ll Need

A project with at least two tables:
- Values to allocate — the costs (or revenues, or volumes) you want to spread. One row per source unit, one column with the amount.
- Driver data — the basis for spreading. Headcount, square footage, transaction counts, revenue — whatever you want to allocate by.
A dimension that ties source and target together (cost centers, departments, products — whichever taxonomy fits your model).

If you don’t have all of that yet, the Tables and views and Dimensions guides cover loading the inputs.

Open the project that holds your values and driver tables.
Create a new workflow. Allocations always run inside a workflow — they don’t operate on tables directly outside of one.
Add an Allocation step. Inside the workflow, add a step from the Allocation category. The most common starting point is Allocation Rules for straightforward driver-based spreading.
Configure the source. Point the step at your values table and pick the column holding the amount to allocate.
Configure the driver. Point the step at the driver table and pick the column holding the driver values.
Map the dimension. Identify which column on each table represents the dimension members. The allocation step uses these to match source rows to driver rows.
Run the step. The output is a new table with one row per spread amount.
Inspect results. Check that the totals match what you expected — sum of allocated amounts should equal sum of source amounts (within rounding tolerance).

Spreading recursively — if a target itself contains drivers for further allocation, see Recursive allocations.
Tagging rows for allocation — to drive which rows allocate to which targets, see Rule-based tagging.
Investigating unexpected results — if totals don’t reconcile or specific rows look wrong, see Troubleshooting allocations.