Cost analysis with sector data: how to approach it

An industry average is a mirror, not a standard. Put your costs next to those of your industry and you see where your business differs. Whether that is good or bad, the number does not tell you. That takes a careful approach: the right comparison group, figures classified the same way, corrections for what is structurally different, and only then the question why. This article describes that approach for owners who do it themselves, for advisers who do it for a client and for controllers comparing business units or subsidiaries with their sector.

What a benchmark is, and is not

As a management method, benchmarking took off at Xerox and was captured by Robert Camp (source: Robert C. Camp, 1989). The subtitle of his book says what it is about: the search for industry best practices that lead to superior performance. The number is the start; the value lies in understanding why others do it differently.

Know two limits. An industry average is an average: it says nothing about the spread between companies. And statistics use their own definitions, which do not always match your general ledger. That is why a good analysis starts with classifying, not comparing.

The approach in six steps

1. 01

   Choose the right comparison group

   Pick the industry that best matches what the business actually does; the registered activity code is not always enough. An installer focused on maintenance has a different cost profile from one focused on new builds. Start with the sector data and go as specific as the figures allow. For a multi-activity company, compare each business unit with its own sector.

2. 02

   Express your figures as a percentage of revenue

   Amounts are not comparable, ratios are. Divide every cost line by net revenue. Use a full financial year or a rolling twelve months, never a single month.

3. 03

   Classify your costs the way the statistics do

   This is where it often goes wrong. Statistics offices use fixed definitions. Statistics Netherlands, for example, counts subcontracted work as cost of sales and rent and lease of buildings as housing costs (source: CBS, April 2026). If you own your premises, those costs largely sit in depreciation instead. Read the definitions behind the figures you use and map your accounts to them.

4. 04

   Correct for what is structurally different

   The main correction is owner pay. Statistics Netherlands explains that wages only exist where there is an employment relationship; the income of a self-employed person is mixed income from work, which includes profit (source: CBS, March 2021). A sole proprietorship therefore shows lower personnel costs and a higher result on paper than a comparable company with a salaried director. Compare like with like, or add a market-rate salary for the owner. Correct also for owned versus rented premises, leased versus owned assets and one-off items.

5. 05

   Look at the trend, not one year

   A difference in one year can be chance. A difference that grows for three years is a pattern. Compare your development with the industry's: are your personnel costs rising faster than at comparable companies, or are they rising everywhere?

6. 06

   Explain, decide and follow up

   Every large difference gets an explanation: a deliberate choice, a structural disadvantage or a problem. Only the last needs action. Record it and check after a quarter whether the ratio moves.

Average and spread

An industry average hides how far companies are apart. Read a difference as a question, not a verdict. A small difference says little; a large difference you cannot explain deserves attention. The larger the comparison group, the more an average is shaped by a mix of company sizes and business models, so also check whether the group fits your own model.

Where to find the figures

The mapping table and the explanation matrix

You know the six steps. Below you will find the mapping table that links common ledger accounts to the right industry cost line, a checklist of the corrections most often needed, and the explanation matrix: for each type of difference, the possible causes and the questions that tell you which one it is.