Page 19 of the SEP refers to benchmarking as a potential way to generate robust data: ‘Other sources of robust data may include benchmarking against peer research units [..]’. Although this sounds simple, the SEP does not explain what exactly is meant by this technique/methodology and precisely how it should be operationalised.
There are several objections to benchmarking in the strict sense of comparing performance based on quantitative criteria (time, costs, yield). The main objections are that it is not really possible to make such a comparison, and that attempting to do so is at odds with the spirit and purpose of the SEP. For a more detailed explanation of these objections, see the explanation below.
For this reason, if you want to use benchmarking, it is advisable to start from a broader definition of the term and to take a more qualitative approach as a way to position the research unit. With this approach, aspects such as the position, mission and strategy of the research unit are compared to or mirrored in general terms against similar units/institutions. Such positioning can be extremely useful and can even take the form of a case study.