Every business runs on numbers. The uncomfortable question is whether those numbers mean what you think they mean.
Most management accounts are assembled by pooling overhead and spreading it across products, clients or divisions using whatever base is handy - revenue, headcount, labour hours. It is tidy, it reconciles, and it can be quietly expensive. The moment you blend and allocate, you lose the one thing a decision actually needs: a clean line from each cost back to whatever caused it.
Writing for the IMA, Brian Higgins argues the flaw runs deeper than imprecision. Conventional systems - absorption costing and much of activity-based costing included - lean on pooled and allocated costs, and the allocation base is chosen for convenience, not because it reflects how each line of business truly consumes resources. Worse, pooling destroys the audit trail. Once costs are averaged together you cannot trace them back to validate them, so you never really know whether a product makes money - only what the formula decided.
That is not a rounding error. It changes decisions.
Higgins describes a financial-services firm facing a projected $25 million profit shortfall as it migrated customers from a decades-old legacy platform to a new one. On paper the legacy business looked terminal: low, declining profitability. The firm had already tried the usual fixes - a Lean Six Sigma programme costing close to $1 million that, in the CFO's words, “simply did not move the needle,” a Big Five consulting engagement, and an activity-based costing project that collapsed under its own complexity.
The fault was in the map, not the territory. The firm allocated indirect and overhead costs by revenue. Because the legacy platform earned 51% of total revenue, it absorbed the majority of shared costs - including costs that had nothing to do with it. In truth, the old platform was subsidising the new one. Corrected, the profitability picture inverted: the ‘dying’ business was carrying the ‘accelerating’ one, and a migration strategy had been built on an artefact of the allocation method.
Fixing the costing surfaced a second, larger problem that pure financials would never have caught. Supporting customers was costing the firm roughly $24 million a year - close to a tenth of total spend - while more than 60% of customer sites had churned within two years for want of adequate support.
The reason was cultural. Sales staff had quietly taken over customer support because they did not trust operations to handle it. Inside the sales function that non-core work consumed 26% of effort but registered as just 11% of cost - a mismatch no money-only system could see. Support belonged to operations, which could do it at roughly half the cost per head. Handing it back let sales put 50% more effort into its real mission, generating revenue, and, according to the author, unlocked $45 million of additional higher-margin revenue at a company whose top line had been flat for years.
The point is not that any one methodology is magic. It is that financial data and non-financial, stakeholder-level input have to sit in the same frame. One tells you what is happening; only the other tells you why.
Higgins's proposed remedy is conceptually simple but demanding: assign costs directly. Treat every cost element as traceable - pulled straight from the general ledger and payroll -rather than pooled and averaged. Done properly, he reports, more than 95% of spending can be tied to specific activities and lines of business, leaving a genuine two-way audit trail you can actually validate. He cites returns across decades ranging from 50:1 to over 100:1. Read those as a practitioner's own account rather than an independent audit. The underlying discipline, though, stands on its own: you cannot improve what you have mismeasured.
The deeper lesson is about how organisations decide what to fix at all. Most begin by chasing the loudest complaint rather than the largest opportunity - the squeaky wheel gets the grease while the real leak goes unpriced. The better starting point is the set of linkages between cost, effort, profitability and customer sentiment, which rarely point where intuition does.
For founders and finance leaders the practical takeaway is smaller than any framework. Before the next consequential call - drop a product, chase a segment, migrate a platform - ask what your management accounts are quietly assuming, and whether the allocation base beneath them would survive a second look. Making numbers decision-useful before the decisions are made is precisely the finance process work we do at DualMinds.
Because the costliest number is rarely the one you got wrong. It is the one you trusted.
Source: What Your Costing Systems Are Not Telling You - Brian Higgins, IMA (Institute of Management Accountants), 2025.
