TDABC and GPK are, in modern practice, the two methods that serious profitability work actually runs on, and they converge far more than they differ. TDABC (time-driven activity-based costing) models cost through time, costing each transaction with a capacity cost rate and a time equation, and reaches naturally to the customer and order. GPK (Grenzplankostenrechnung, German marginal planned cost accounting) models cost through a cost-centre resource structure, with an explicit fixed-and-proportional split and a multi-level contribution-margin P&L. Both prize practical capacity, both surface unused capacity, and both run off operational and ERP data. This page sets them side by side and shows when each is the right tool.
TDABC and GPK are the two most rigorous "serious" management costing systems in modern practice, and they converge more than they differ. Both prize practical capacity and make unused capacity visible, and both run off operational or ERP data. The difference is how they model cost. TDABC models cost via time, using a capacity cost rate (cost divided by practical capacity, typically 80 to 85 per cent of theoretical) and time equations in minutes per transaction; it is lighter and faster to stand up and reaches customer and order level naturally. GPK models cost via a cost-centre resource structure with an explicit fixed-and-proportional split and a multi-level contribution-margin P&L; it is heavier but very strong for marginal pricing and mix across the whole organisation. RCA is effectively the bridge between them. ---
The core difference
The cleanest way to see the difference is to ask what each method uses as its modelling currency.
TDABC
TDABC models cost in time. Introduced by Kaplan and Anderson (HBR, November 2004; book, 2007), it needs just two parameters: a capacity cost rate, which is the cost of a resource group divided by its practical capacity (usually set at around 80 to 85 per cent of theoretical, to leave honest slack), and time equations that express how many minutes a transaction consumes. Multiply the two and you have the cost of an order, a customer or a process step. Because it works in minutes off ERP and CRM data, TDABC scales to thousands of transactions, surfaces unused capacity as the gap between supplied and used minutes, and reaches the customer with ease.
GPK
GPK models cost in cost-centre resource structure. Built from cost-type, cost-centre, product and contribution-margin elements, its defining discipline is splitting every cost into fixed and proportional across hundreds or thousands of cost centres, under a strict principle of causality. Only proportional cost flows to products; fixed cost is held back into a multi-level contribution-margin P&L. It is heavier to build and feed, and it leans on integrated ERP, but it is very strong for marginal pricing and mix across the whole organisation. Its international cousin, RCA, itself blends GPK with ABC-style drivers, which is why RCA reads as the bridge towards TDABC.
Side by side
| Dimension | TDABC | GPK |
|---|---|---|
| Origin | Kaplan and Anderson, HBR 2004, book 2007 | Germany, Plaut and Kilger, from the late 1940s |
| Modelling currency | Time (minutes per transaction) | Cost-centre resource structure |
| Two key parameters | Capacity cost rate; time equations | Cost-type, cost-centre, product, contribution margin |
| Practical capacity | Yes (typically 80 to 85% of theoretical) | Yes (core to the fixed-proportional logic) |
| Unused capacity | Surfaced as unused minutes | Surfaced via fixed cost held back |
| Reaches the customer | Naturally, at order level | Possible, but resource and product oriented |
| Weight to stand up | Lighter, faster | Heavier, data-hungry |
| Best decisionKey row | Customer and cost-to-serve profitability | Marginal pricing and mix across the org |
A worked contrast
Take an illustrative business, CaP Logistics (figures illustrative). TDABC costs an order directly: a capacity cost rate of, say, EUR 0.856 per minute (illustrative), multiplied by the minutes a time equation predicts for that order's pick, pack, check and dispatch. Add up the orders for a customer and you have a clean cost to serve, plus a visible block of unused capacity wherever supplied minutes exceed used minutes. That customer-level reach is TDABC's home ground.
GPK costs the same operation from its resources. For each relevant cost centre it derives a proportional rate, the cost that genuinely varies with output, and holds the fixed cost back into separate blocks. Those blocks drop into a multi-level contribution-margin P&L, so a pricing or mix decision rests on the cost that actually changes with volume. The two methods are looking at the same plant through different lenses, time versus resource structure, and RCA is effectively the bridge, blending GPK's resource modelling with ABC-style drivers of the kind TDABC refined.
When to choose which
Reach for TDABC when you need fast, scalable customer and cost-to-serve profitability, and your environment is mid-size to large with many transactions. It stands up quickly, scales to thousands of orders, surfaces unused capacity and reaches the customer naturally, which makes it the strong default for distribution, logistics, healthcare and financial services.
Reach for GPK when you run a capital-intensive, controlling-heavy, ERP-deep manufacturing business, especially in a German-speaking context, and your central question is clean marginal cost for pricing and mix across the whole organisation. GPK rewards that depth with marginal rigour few methods match. If you want that resource discipline in an international setting, look at RCA, the cousin that blends GPK with ABC-style drivers.
Because the two converge so much, the choice is often less either-or than it looks. Both make practical capacity and unused capacity visible, both run off operational data, and RCA sits deliberately between them. Pick TDABC for speed and customer reach, GPK for cost-centre depth and marginal pricing, and RCA when you want a foot in both.
Frequently asked questions
Are TDABC and GPK really that close?
Closer than most people expect. Both are rigorous "serious" costing systems that prize practical capacity, make unused capacity visible, and run off operational or ERP data. The main difference is modelling currency: TDABC works in time (minutes per transaction), while GPK works in cost-centre resource structure with a fixed-proportional split. RCA sits between them as a bridge.
Which is faster to implement, TDABC or GPK?
TDABC, generally. It needs only two parameters, a capacity cost rate and time equations, and runs off ERP and CRM data, so it is lighter and faster to stand up and scales easily to thousands of transactions. GPK is heavier and data-hungry, built across hundreds or thousands of cost centres, and typically leans on integrated ERP.
Which one reaches the customer?
TDABC reaches the customer naturally, at order level, which is why it is strong for cost-to-serve work in distribution, logistics, healthcare and financial services. GPK is resource and product oriented; it can inform customer decisions but is built around cost centres and a contribution-margin P&L rather than around the customer directly.
Where does RCA fit between them?
RCA (resource consumption accounting) is effectively the bridge. It is GPK's international cousin and itself blends GPK's resource and cost-centre modelling with ABC-style drivers, the same family of drivers TDABC refined. If you want GPK's resource discipline with more activity and driver flexibility, RCA is the path between the two.
Can a business run both TDABC and GPK?
It can, because they look at the same operation through different lenses and both rest on practical capacity. A firm might run GPK-style cost-centre discipline for marginal pricing while using TDABC for fast customer and cost-to-serve profitability, with RCA as the conceptual bridge. Each answers a question the other does not.
Go deeper
Want to know which of your customers actually make you money? Start with a Profit Check.
Start a Profit Check