Regarding law departments, we often use the term “benchmark metrics” loosely. Start with “metrics.” They are something you can count that exists independently of the counting. The square feet of a law department’s office space is a metric; the amount paid in overtime to secretaries is a metric; the number of law firms retained in Canada is a metric. Before someone measured or counted, the feet, pay, and firms existed.
When you collect metrics over time for yourself or from several departments, and calculate the average or median or whatever of those metrics, you have created a benchmark metric. It is a benchmark because you can compare yourself to the pattern of metrics.
By contrast, if someone asks law department respondents “Will you increase your office square feet next year?” you can calculate the percentage that reply “Yes” and “No” and “Don’t know” but that finding is not a benchmark metric by my definition. Nothing existed until you asked the question, tallied the responses, and described the distribution of the responses. Similarly, “Please rank the following methods for managing outside counsel costs” does not produce what I think of as a metric, let alone a benchmark. It produces a tally which gives some insight but no comparison.
To benchmark your department’s metrics is to match how you stand comparatively based on metrics about existing and countable things. To find out distributions of opinions and views is something else.