Activity-based costing (ABC) assigns overhead through activity drivers instead of flat volume rates - and most ABC implementations fail under their own weight. We are ABC consultants of a specific kind: we build costing models on ABC's principles using its time-driven successor, and we rescue classic ABC models that have stopped being maintained. 150+ models since 2010.
ABC was right about the problem
ABC's core insight, formulated in the late 1980s, still stands: overhead is not caused by volume, it is caused by activities - setups, orders, inspections, deliveries - and different products and customers consume those activities very differently. Spreading overhead by volume hides that difference and misprices everything.
If your company allocates 40%+ of its costs with a single key, ABC's diagnosis applies to you, four decades later, unchanged.
Why classic ABC projects die anyway
The method's weakness was never the idea - it was the data collection. Classic ABC asks employees to estimate, by survey, how their time splits across activities. In practice:
- Surveys are subjective, political and always sum to 100% - nobody reports idle capacity;
- Re-surveying every quarter is so painful that the model freezes; within two years it describes a company that no longer exists;
- Every new product, channel or activity multiplies the survey matrix - complexity kills the model exactly where it was most needed;
- The model ends up understood by one analyst. When they leave, it dies.
This is not our opinion - it is why Kaplan, ABC's own co-creator, developed time-driven ABC (TDABC) as its replacement: time equations estimated from operational data instead of surveys, with capacity made explicit.
Three situations, three ways we help
1 · You have an ABC model that's dying
We audit what exists, keep what's right (the activity dictionary and driver logic are usually reusable), and migrate the survey-based allocations to time equations fed by your transaction data. Typical migration: 4-6 weeks, and the quarterly survey ritual disappears.
2 · You're starting from zero
We skip the 1980s entirely and build time-driven from day one - see our method. You still get everything ABC promised: cost per activity, per product, per customer - without the survey debt.
3 · You need ABC for compliance or costing standards
Some sectors (healthcare reimbursement, public tenders, transfer pricing) require recognisable ABC structures. We build models that satisfy the requirement while running on time equations underneath - auditable on the surface, maintainable underneath.
Classic ABC vs. time-driven ABC
| Classic ABC | Time-driven ABC | |
|---|---|---|
| Driver data from | Employee surveys | Operational systems (ERP, WMS, billing) |
| Idle capacity | Hidden (surveys sum to 100%) | Explicit - you see unused capacity and its cost |
| Update cycle | Quarterly survey campaigns | Automatic, with the data |
| Complexity handling | Each variant = new activity | Each variant = one term in a time equation |
| Typical lifespan | 2-3 years, then abandoned | Indefinite - it's a system, not a study |
Frequently asked questions
What's the difference between ABC and TDABC?
Both assign overhead through activities. Classic ABC estimates activity consumption with employee surveys; TDABC calculates it with time equations derived from operational data. TDABC was developed by Robert Kaplan - ABC's co-creator - precisely to fix ABC's maintenance problem.
Can we reuse our existing ABC work?
Usually yes, and you should. The activity dictionary, driver logic and cost-pool structure of a well-built ABC model transfer directly; what gets replaced is the survey-based data collection. Migration is typically faster than building from scratch.
Why do ABC projects fail?
Survey fatigue, hidden idle capacity, and exploding complexity. The model freezes because updating it hurts, and within two years it describes a company that no longer exists. The failure is in the data collection method, not in the activity-based idea.
How long does an ABC-to-TDABC migration take?
For a model with a sound activity structure, 4 to 6 weeks to a live time-driven model - including validation against the old model so finance can see exactly what changed and why.
Have an ABC model on life support?
Start with the free Profit Check, or book a scoping call and bring the old model - we'll tell you honestly what's worth saving.
