Snowball Metrics present as a totalizing grand narrative. For now, let me simply list some of the ways in which this is so, with little or only brief explanations.

1. Snowball metrics are a tool for commensuration, “designed to facilitate crossinstitutional benchmarking globally by ensuring that research management information can be compared with confidence” (p. 5 — with all references to page numbers in this PDF).
2. Snowball metrics are based on consensus: “Consensus on the ‘recipes’ for this first set of Snowball Metrics has been reached by a group of UK higher education institutions” (p. 8).
3. Despite the limited scope of the above consensus, however, Snowball Metrics are intended to be universal in scope, both in the UK “We expect that they will apply equally well to all UK institutions” and “to further support national and global benchmarking” (p. 8).
4. Snowball Metrics are presented as a recipe, one to be followed, of course. The word occurs 45 times in the 70 page PDF.
5. Other key words also appear numerous times: agree (including variations, such as ‘agreed’) appears 31 times; method (including variations, such as ‘methods’ or ‘methodology’) appears 22 times; manage (including variations) appears 15 times; impact appears 16 times, 11 times in terms of “Field-Weighted Citation Impact.”
6. Snowball Metrics are fair and “have tested methodologies that are freely available and can be generated by any organisation” (p. 7).
7. Snowball Metrics are ‘ours‘ — they are  “defined and agreed by higher education institutions themselves, not imposed by organisations with potentially distinct aims” (p. 7).

To sum up, using their own words:

The approach is to agree a means to measure activities across the entire spectrum of research, at multiple levels of granularity: the Snowball Metrics Framework. (p. 7)

Coming in the next post (4 of ?), I present an alternative ‘framework’ — let’s call it Snowflake Indicators for now.