Averages hide where the money is really made.
A single company profit number tells you almost nothing about where margin comes from or leaks away. This question reveals how finely you can slice profitability, from the whole business down to the individual transaction, and how much detail you lose to averages.
“How granular is your profitability analysis: company, segment, product-customer, or transaction?”
What does granularity of profitability analysis mean?
It means the smallest unit at which you can measure profit. At the coarsest level, you have one profit number for the whole company, an average that hides every difference inside it. As granularity increases, you can see profit by segment (product line or region), then by product and customer together, and finally by individual transaction (order, SKU, or shipment). Each finer level reveals variation the level above averaged away: a profitable product line can contain loss-making products, and a good product can be sold at a loss to particular customers in particular orders. Time-driven activity-based costing is what makes the finest levels possible, because it can attach cost to each transaction, not just each period.
The average is a story about no one.
A single profit figure describes an entity that does not exist: the average customer, the average order, the average product. Real decisions are made at a finer grain, and at that grain the averages routinely mislead. A healthy segment margin can hide products sold at a loss; a good product can be unprofitable through one channel or to one customer.
Granularity is what turns cost data into decisions. The finer the unit of analysis, the more precisely you can act: reprice this product for that customer, drop this order type, change these terms. Coarse analysis forces blunt, across-the-board moves that help the average and hurt the specific.
The limit is not ambition but capability. Reaching transaction-level profitability requires attaching cost to each transaction, which is exactly what time-driven activity-based costing is built to do, and what spreadsheets and period-level allocations cannot.
From one number to the transaction.
As granularity matures, a single company profit number gives way to segment analysis, then to product-and-customer detail, and finally to transaction-level profitability where every order can be measured.
How fine is your analysis?
Question 3 assesses the smallest unit at which you can measure profitability. Each level cuts a finer cell and reveals variation the one before it hid.
“We have one profitability number for the whole business.”
Profit is known at the company level and nowhere finer. The single figure is an average across every product, customer, and channel, so it can look healthy while masking large pockets of loss inside it. Any decision below the company level is made blind.
- One profit figure stands in for a diverse business
- Losses inside the average are invisible
- Decisions below company level lack any profit basis
- Good and bad performance cancel out in the total
“We can see profitability by segment, product line, or region.”
Profit is split into meaningful segments, the first real cut into the average. This reveals which lines or regions lead and which lag, and is often where the first uncomfortable surprises appear. But within each segment, individual products and customers are still averaged together.
- Segments reveal variation the company total hid
- Averages still operate inside each segment
- Winners and losers within a line remain merged
- The cut is useful but not yet actionable at the edge
“We can analyse profitability by product and by customer together.”
Profitability is measured at the intersection of product and customer, so you can see not just which products earn but which customers they earn from. This is where most of the actionable insight lives: the same product can be profitable to one customer and a loss to another.
- Profit is seen at the product-customer intersection
- Same-product, different-customer variation becomes visible
- Insight is now specific enough to act on
- Data and cost modelling must support the finer grain
“We can measure profitability down to the individual transaction.”
Profit is measured at the atomic level: the individual order, SKU, or shipment. Every transaction carries its own cost to serve, so the whale curve, mix effects, and marginal decisions can all be seen with precision. This is the level time-driven activity-based costing is built to support.
- Transaction-level costing needs disciplined data and TDABC
- Detail must be summarised well, or it overwhelms
- The model must stay transparent and maintainable
- Granularity serves decisions; it is not detail for its own sake
Practical steps, level by level.
Quick Wins
- Split your single profit number into a few meaningful segments
- Choose segments that reflect real decisions: line, region, or channel
- Look for the first surprise, the segment that is weaker than assumed
- Bring that split to the leadership table as a starting point
Structural Improvements
- Break segments down to individual products and customers
- Measure profit at the product-customer intersection, not just each alone
- Attach cost to serve so the intersection reflects true net profit
- Surface the same-product, different-customer differences and act on them
World-Class Practices
- Model profitability at the transaction level with TDABC
- Let each order, SKU, or shipment carry its own cost to serve
- Summarise the detail into decisions, not just dashboards
- Keep the model transparent so the granularity stays trusted and used
Where fine detail pays back most.
Every business benefits from finer analysis, but transaction-level granularity pays back hardest where volume is high and each transaction differs.
| Industry | Granularity Signal | Key Insight |
|---|---|---|
| Distribution & Logistics | Per shipment | Every order and drop differs in size, distance, and handling; only transaction-level profit shows which shipments actually pay. |
| Manufacturing | Per SKU and run | Short runs, changeovers, and custom SKUs vary enormously; product-and-transaction detail reveals which really earn. |
| Retail & E-commerce | Per basket | Returns, delivery, and promotions make baskets of equal value differ in profit; only fine granularity separates them. |
Are your averages hiding the real story?
Take the free Profitability Health Check to see how granular your profitability analysis really is, and what the averages are hiding at the segment, customer, and transaction level.