“We built the model two years ago” is one of the phrases most likely to indicate a cost model problem. Not always — but often enough that it’s worth examining systematically.
The right update frequency is not the same for every business. It depends on how fast your cost structure changes, how decision-critical your model is, and what your data infrastructure supports.
The Four Update Types
Before discussing frequency, it helps to distinguish four types of model update:
- Full rebuild — recalibrate all cost pools, time equations, and allocation rates from scratch.
- Rate update — update cost pool totals (headcount changes, salary increases, new contracts) without changing the activity/time structure.
- Structural update — add or remove activities, cost pools, or product/client categories triggered by a business change.
- Output validation — compare model outputs to actual margin data to check for drift.
These have different effort levels and different triggers. A full rebuild is a quarterly or annual exercise. Structural updates are event-triggered. Output validation should be monthly.
Frequency Guide by Business Type
High-change businesses (startups, fast-growth companies, businesses in restructuring, companies with active M&A): full rebuild every 6 months; rate update quarterly; structural update within 30 days of any trigger event; output validation monthly.
Mid-change businesses (established mid-market, stable product mix, moderate hiring): full rebuild annually; rate update semi-annually; structural update within 60 days of trigger events; output validation quarterly.
Low-change businesses (stable operations, long-standing product lines, predictable cost structure): full rebuild every 18–24 months; rate update annually; structural update as needed; output validation semi-annually.
The Warning Signs That Your Frequency Is Too Low
Your update frequency is too low if:
- Your model shows results that contradict what your operations team experiences (“the model says this product is profitable but we all know it isn’t”).
- Significant changes have occurred (new product, new team, new process) that aren’t reflected in the model.
- You’ve stopped using model outputs for actual decisions because you don’t trust them.
- The last full rebuild was more than 18 months ago in a mid-to-high change business.
Any of these symptoms indicates that the model has drifted past its useful life in its current form.
Making Updates Sustainable
The main reason models don’t get updated is that updates feel like projects — requiring significant time investment each time. This happens when the model is not well-structured for maintenance.
A well-structured TDABC model (in CostCTRL or otherwise) makes rate updates a 2–4 hour exercise, not a 2-week one. The structure is defined once; the numbers update.
The key practices that make updates sustainable: document the data sources for every cost pool (so you know where to look); parameterise time equations (so updating one estimate flows through automatically); and build an annual model review into your finance calendar with a fixed owner.
If your model currently requires a consultant to update, the model structure needs to change — not your update frequency.
A Model That Updates Itself
CostCTRL connects to SAF-T and ERP data exports, which means the financial layer of your model can update automatically at each period close. The operational layer (time equations) needs human review — but the system flags when outputs deviate from the prior period by more than a defined threshold, prompting a check.
The goal is a model that tells you when it needs attention, rather than one you only consult when you think to.
The Profitability Health Check includes questions about your current update cadence and model maintenance practices — and benchmarks you against companies in your sector.