Benchmarking against far larger and better-resourced institutions produces a number and no course of action.
Comparable peers produce something you can actually do something about — which is the entire point of the exercise.
Choosing comparators
- Similar size and resource level.
- Similar disciplinary mix, since field composition changes everything.
- Similar development stage.
- Regional peers, often more informative than global ones.
- Include one institution slightly ahead as a direction, not a target.
Indicators worth comparing
- Output per researcher, not total output.
- Staff and student numbers.
- Completion rates and time to completion.
- Retention of early career staff.
- Collaboration patterns, including domestic first-authorship.
- Trends rather than levels.
Making comparisons valid
- Use the same counting rule for everyone.
- Use the same source and extraction date.
- Normalise for size.
- Account for field composition — publication rates differ enormously between disciplines.
- Note coverage differences, particularly for local-language output.
What benchmarking cannot tell you
- Whether the research is good.
- Why a difference exists.
- What to do about it.
- Whether the comparator's approach would work for you.
- It identifies questions; it does not answer them.
Turning it into action
- Investigate differences before acting on them.
- Ask the comparator institution — they will often tell you.
- Check whether the difference reflects data rather than reality.
- Focus on what you can change.
- A benchmark with no follow-up investigation is a number on a slide.
Common mistakes
- Comparing against aspirational rather than comparable institutions.
- Using different rules for different institutions.
- Comparing raw totals across different sizes.
- Acting on a difference without establishing its cause.
- Treating a single-year comparison as a trend.
One thing worth remembering
Benchmark against institutions you could plausibly become, not ones you admire.
A comparison with a much larger and better-funded institution generates a gap that cannot be closed and a discussion that cannot lead anywhere.
Câu hỏi thường gặp
How should comparators be chosen?
Similar size and resource level, similar disciplinary mix since field composition changes everything, similar development stage, and regional peers which are often more informative.
Which indicators are worth comparing?
Output per researcher rather than total, staff and student numbers, completion rates and times, retention of early career staff, collaboration patterns, and trends rather than levels.
What makes comparisons valid?
The same counting rule, source and extraction date for everyone, normalising for size, accounting for field composition, and noting coverage differences.
What can benchmarking not tell you?
Whether the research is good, why a difference exists, what to do about it, or whether the comparator's approach would work for you — it identifies questions rather than answering them.
Against whom should you benchmark?
Institutions you could plausibly become rather than ones you admire — a gap with a much larger institution cannot be closed or acted on.
