The Margin Looks Healthy Until You Break It Apart

Aggregate margin is a starting point, not a conclusion.

When margin is reported at the business level, it reflects the average of all activity combined.
That average can look acceptable while individual products, contracts, or customer engagements generate returns that range from highly profitable to structurally loss-making.

The business does not see the spread, it sees the mean.

Pricing decisions are made against an average margin that does not reflect the economics of the specific work being priced:

-Resourcing decisions allocate capacity to engagements whose actual contribution has never been calculated.
-Sales decisions pursue volume in segments where the margin per unit of effort is below what the aggregate figure implies.

Each decision follows logically from the information available. The information was never granular enough to support the decision being made.

A reported margin of roughly 50% built from engagements ranging from 25% to 75% is not a 50% margin business, it is a business that has not yet separated its profitable work from its unprofitable work. The total reconciles, the distribution is invisible.

Every decision made at the product, contract, or customer level is made against a number that describes the average of the whole rather than the economics of the part.

Businesses that resolve this condition shift their cost tracking from the aggregate to the engagement level.

Labor, materials, and overhead are attributed to the unit of work they support rather than averaged across the total.

The outcome is a margin picture that shows where the business is genuinely profitable and where it is not.

Jeff Glick, CPA and Head of U.S. Operations at OCFO, joins us to discuss trust financial signals that fail to reflect what is actually happening inside the business.

City Shift Finance works with organizations on the structural conditions that affect revenue durability, margin visibility, reporting quality, and downstream financial decision making.