activity cost pool

Lumping all of a factory's overhead into one giant bucket and spreading it by a single yardstick hides the fact that different activities cost money for completely different reasons. Setting up a machine, ordering materials, and inspecting finished goods are unrelated activities, each driven by something different. Activity-based costing first sorts overhead into separate buckets, one per activity. Each of those buckets is an activity cost pool — a grouping of overhead costs that all relate to the same activity.

An activity cost pool collects all the overhead costs caused by one activity so they can be assigned together using that activity's own cost driver. For instance, a 'machine setup' pool gathers the wages of setup workers, the cost of setup tooling, and machine idle time during setup; this pool is then divided by the total number of setups to get a cost per setup, and each product is charged for the setups it required. A typical ABC system runs several such pools — setups, purchasing, inspection, material handling, engineering — each with its own pool total and its own driver. Building good pools means grouping costs that move together for the same reason; mixing unrelated costs into one pool would just recreate the very blurriness ABC is trying to remove.

Activity cost pools matter because they are the organizing unit of activity-based costing: the accuracy of the whole system depends on choosing pools that are homogeneous (driven by one thing) and a driver that truly explains each pool. More pools generally mean more accuracy but more cost and complexity to maintain, so designers balance precision against effort, often grouping minor activities together. The common pitfall is creating pools that look activity-based but still bundle several unrelated cost behaviors under one driver — which quietly reintroduces the cross-subsidies that ABC was meant to expose.