Benchmarking that creates useful context

What benchmarking is - and what it is not

A performance figure on its own is not a benchmark. It becomes a benchmark when it is placed in a structured comparison: with a relevant group, a market, a historical period, an agreed standard or another meaningful reference point. Benchmarking can help an organisation see where it differs and where further investigation may be worthwhile. It does not automatically explain why a difference exists, and it does not by itself improve performance. That distinction matters. The benchmark creates context. People still need to interpret the context and decide whether action is appropriate.

Benchmarking is not the same as ranking

A ranking orders participants. A benchmark compares results in context. The two can overlap, but they serve different purposes. In many association and network programmes, the objective is not to declare a winner. It is to help each participant understand its own position and development. A value above the average is therefore not automatically good, and a value below it is not automatically bad.

Start with the question, not the KPI

A common mistake is to begin by collecting every metric that might be interesting. A stronger benchmark starts with the question the organiser or participant needs to answer. What should a participant understand after receiving the result? Which decision or discussion should the comparison support? Which differences are meaningful enough to investigate? Only then does it make sense to select the indicators and underlying data.

Five conditions for a useful benchmark

Reference groups: more specific is not always better

Peer groups can make a benchmark more relevant when participants operate under meaningfully different circumstances. But every extra segmentation reduces the number of observations in a group. That can affect stability, confidentiality and the usefulness of the comparison. The right question is therefore not 'How many peer groups can we create?' but 'Which distinction materially improves interpretation?'

There is no universal minimum number of participants

The number of participants needed depends on what the benchmark is trying to do. A market estimate, a confidential remuneration comparison and an internal network benchmark can require different approaches. Heterogeneity, confidentiality, distribution, reference-group design and the intended use of the result all matter. A single fixed minimum would create false certainty.

The average can hide important information

An average is easy to communicate, but it may not represent a typical participant when the distribution is skewed or contains extreme values. Depending on the research question, a median, percentile, range or distribution can provide more useful context. The choice of statistic should follow the data and purpose rather than habit.

Historical comparison answers a different question

Peer comparison asks: how do I compare with others? Historical comparison asks: how am I developing? Both can matter. An organisation can improve while remaining below a reference group, or compare favourably with peers while its own trend deteriorates. A useful benchmark can make those perspectives visible separately.

Method changes can break historical comparability

Recurring benchmarks become more valuable as history accumulates, but only if the meaning of the data remains sufficiently consistent. Changing a definition, question, calculation or participant population can change the trend. Methodological improvements are sometimes necessary. They should be documented so users know when a time series remains comparable and when it does not.

Changing participants can distort market trends

A group result can move because the market changed - or because different organisations participated. For recurring market monitoring, a consistent panel can help isolate development among organisations that participated in the relevant periods. This improves one aspect of comparability but does not automatically make the panel representative of the entire market. VvDN uses consistent participation across consecutive quarters for particular development analyses, illustrating the distinction between trend consistency and whole-market representation.

Validation comes before comparison

A benchmark can only be as credible as the data entering the comparison. Validation can identify missing values, implausible entries, inconsistencies between fields or unexpected movements over time. But validation is not the same as silently correcting data. Where a value needs explanation or adjustment, the process should make clear who is responsible for that decision.

Participant value and organiser value can coexist

Many association benchmarks have two audiences. The participant wants to understand its own organisation. The association wants to understand the sector or member group. One research programme can support both, but the outputs do not need to be identical. An individual benchmark report can focus on relevant peer comparison while the organiser uses aggregated results for market insight, member services or policy work.

Not every benchmark needs individual feedback

Individual reporting is valuable when the participant's own position is central to the research purpose. It is not a requirement for every study. A market monitor, for example, may primarily create aggregated sector insight, while another programme returns extensive participant-specific reports. The output should follow the purpose rather than a fixed benchmark template.

Benchmarking can become a recurring member service

A well-designed benchmark often becomes more useful after the first edition. Definitions improve. Participants learn what information is needed. Historical comparisons become possible. The organiser can build recurring communication or advisory activities around the results. This is why the operational design matters as much as the first questionnaire. A benchmark that depends on one person's spreadsheet knowledge may be difficult to sustain even when the methodology itself is strong.

Different benchmark programmes require different operating models

C2C customers do not all outsource the same work. Rödl retains its specialist methodology and operates the benchmark itself using C2C technology. BMF and SRA perform substantial parts of their own research with C2C support. Kantoorcijfers is closer to a research partnership. For VvDN, C2C operates much of the recurring process while the association uses aggregated market insight. The practical lesson is that benchmark methodology, technology and research execution are separate choices.

Examples from recurring benchmarks

Core Benchmarking topics

Useful benchmarking starts with a clear definition of what is being compared and why. C2C’s COMPARE approach provides a structure for moving from the research question and definitions to comparison, reporting and recurring improvement. Within that process, indicators, ratios and reference groups need to fit the question, while the choice of mean, median or percentile should reflect the distribution of the data.

Recurring programmes also require attention to consistent definitions, methodology changes, historical continuity and documented validation. When participants change, consistent panels can help interpret trends, with their limitations stated clearly. Individual reports should turn those comparisons into relevant feedback without overwhelming the reader. Insight, interpretation and action remain separate stages: a benchmark does not automatically improve performance.

A current version of the COMPARE approach

C2C has used the COMPARE concept for many years to structure thinking about successful benchmarking. The underlying idea remains useful: benchmarking is a process, not a single calculation. For the new Knowledge Centre, the older COMPARE material should be rewritten as a current practical guide rather than republished as an old ebook. The updated version should retain the durable methodology, use current customer examples and terminology, and remove outdated product positioning or legal wording. This can become one of the cornerstone articles in the Benchmarking topic cluster.

From knowledge to implementation

The Knowledge Centre explains the methodology. The Solutions pages show how C2C applies it to performance benchmarking, market monitoring, remuneration and other research programmes. The Survey Tool pages explain how collection, validation, analysis, dashboards and workflows support the process technically. Customer stories show how different organisations divide the work in practice. These layers should link to each other without duplicating the same page in four different forms.

Start with the comparison you need

A useful benchmark does not start with a template of standard KPIs. It starts with the question you want participants or the sector to understand, the data needed to answer it and the comparison that gives those data meaning. If you are designing a new benchmark or reviewing an existing one, that is the best place to begin.