Comparing research performance across universities has always run into the same problem: even when two institutions report a metric with the same name — “international collaboration rate,” “research income,” “citation impact” — they frequently mean different things by it, calculated from different data sources with different inclusion rules. Snowball Metrics is a community-owned methodology that exists to close that gap. It is not a product, a database, or a ranking system; it is a shared, publicly documented set of recipes that define exactly how a given research metric should be calculated, so that a number produced at one university means the same thing as the same-named number produced at another.
What Snowball Metrics actually is
Snowball Metrics is a methodology, published as a freely available Recipe Book, that specifies the precise calculation steps behind a defined set of research-management metrics — covering areas such as research income, collaboration, publication and citation activity, and commercialisation/knowledge exchange. Each “recipe” documents the metric’s definition, the data elements it draws on, and the calculation logic in enough detail that two institutions following it independently should arrive at comparable figures.
The key design principle is methodological agreement, not tool provision. Snowball Metrics does not tell an institution which database, CRIS, or analytics platform to use to produce a metric — it tells institutions how the metric should be defined and calculated regardless of the underlying system, so that outputs from different tools and different data sources remain comparable.
Origin and governance
Snowball Metrics began in 2010, when eight UK research-intensive universities started working together to agree a single, consistent method for calculating a set of research metrics that could inform institutional and funder strategy. The founding rationale was explicit: strategic decisions — about investment, partnership, or performance — are only as good as the comparability of the numbers behind them, and inconsistent, institution-specific metric definitions undermine that comparability.
An early phase of the work was carried out as a project with CASRAI, made possible by a donation from Elsevier, with input into the metric recipes from bodies including the Wellcome Trust, the Medical Research Council, and RAND Europe. That history is part of why Snowball Metrics recipes have historically been represented within CASRAI’s own data-dictionary work, and why the initiative’s governance has continued to coordinate with CASRAI and with euroCRIS (which is responsible for representing the recipes in CERIF, the standard data model used by many Current Research Information Systems) on interoperability between the methodology and the systems institutions actually use to hold research data.
Since that founding phase, Snowball Metrics describes itself as a bottom-up initiative owned by research-intensive universities — now spanning institutions in the UK, the United States, Australia, and New Zealand, coordinated through a Steering Group drawn from participating universities. This ownership structure is a deliberate design choice: the metrics are defined by the institutions that use them for their own strategic purposes, rather than being imposed by a funder, government agency, or commercial data supplier whose interests in how a metric is defined might diverge from the institutions being measured. Elsevier has supported the initiative (including through Scopus and SciVal, which can supply underlying data for some recipes), but Elsevier does not own or control the methodology itself, and the recipes are published for anyone to use with any compliant data source, free of charge.
What the Recipe Book standardizes
The Recipe Book sets out agreed, tested definitions across the general landscape of research activity that institutions typically report on, including:
- Research income and funding — how funded research activity is counted and attributed.
- Collaboration — including measures of international and industry co-authorship or co-funding.
- Publications and citation impact — bibliometric measures calculated in a defined, reproducible way rather than an ad hoc one.
- Commercialisation and knowledge exchange — outputs connected to licensing, spin-outs, and industry engagement.
Each recipe in the current Recipe Book documents not just a formula but the boundary conditions around it: what counts as an eligible output, what date range applies, what counts as a duplicate, and how edge cases should be resolved. That level of specificity is what makes the metric portable — an institution can hand the recipe to any analyst, at any institution, using any compliant data source, and expect a consistent result, which is the entire point of a shared benchmarking methodology.
Why standardized methodology matters for benchmarking
Cross-institution benchmarking is only meaningful if the underlying metric is calculated the same way everywhere it’s reported. Without an agreed methodology, a university might report a “collaboration rate” that includes only externally funded co-authored papers, while a peer institution reports the same-named metric using a broader definition that includes any multi-institution byline — producing numbers that look comparable on a dashboard but aren’t actually measuring the same thing. Snowball Metrics addresses this by making the calculation method itself the object of agreement, not just the metric’s name.
This matters for several groups of users:
- Research offices and strategy teams, who use standardized metrics to benchmark their institution against self-selected peer or aspirant groups without having to first reconcile differing internal definitions.
- Funders and government bodies, who receive institutional reporting that is more directly comparable across the institutions they fund.
- CRIS and research-information-system vendors, who can implement a documented, external methodology rather than each defining metrics independently — which is where the CERIF/euroCRIS interoperability work is most directly useful.
Snowball Metrics sits alongside — and is a methodological complement to, rather than a substitute for — the broader research-assessment reform conversation. Responsible-metrics frameworks such as the San Francisco Declaration on Research Assessment (DORA) and the Coalition for Advancing Research Assessment (CoARA) caution against over-reliance on any single quantitative metric, particularly journal-level proxies applied to individual researchers. Snowball Metrics doesn’t take a position on which metrics an institution should prioritize or how heavily quantitative indicators should weigh in an evaluation decision; it addresses a narrower, upstream problem — ensuring that when an institution does choose to report or compare a given metric, the number is calculated consistently. Used well, the two are complementary: DORA/CoARA-style principles guide whether and how a metric should be used in an evaluative decision, and a methodology like Snowball Metrics governs how the number is actually produced so that comparison is valid in the first place.
How Snowball Metrics differs from a bibliometric database or ranking
It’s worth being precise about what Snowball Metrics is not. It is not a bibliometric database like Scopus, Web of Science, or OpenAlex — see CASRAI’s bibliometric analysis methodology and workflow guide for how those data sources and the metrics computed from them actually work. It is not a university ranking system, and it does not publish comparative league tables of institutions itself. It also isn’t a proprietary analytics platform of the kind covered in CASRAI’s overview of faculty research-productivity benchmarking platforms. Rather, Snowball Metrics is the methodological layer underneath tools like these: a database supplies the underlying data, an analytics platform or CRIS may apply the calculation, and a ranking body applies its own separate weighting and aggregation choices — but Snowball Metrics is what makes it possible for the same input data, processed the same documented way, to produce a metric that a peer institution’s equivalent process can be checked against.
Frequently asked questions
Is Snowball Metrics a commercial product?
No. Snowball Metrics is a community-agreed methodology, and the Recipe Book that documents it is published for free use by any institution or analyst. It began with support from Elsevier and other funders during its founding phase, and Elsevier’s Scopus/SciVal products can be used as a data source for some recipes, but the methodology itself is owned and governed by the participating research-intensive universities, not by Elsevier or any single commercial supplier.
Who can use the Snowball Metrics Recipe Book?
The recipes are published openly and can be used by any institution, analyst, or system vendor, regardless of whether that organization is formally part of the Snowball Metrics governance structure. Using the recipes doesn’t require paying a fee or purchasing a specific product.
How does Snowball Metrics relate to CASRAI?
Snowball Metrics’ founding methodology work was carried out in part as a project with CASRAI, and the initiative’s governance has continued to coordinate with CASRAI (alongside euroCRIS) on how the recipes are represented for interoperability with research-information systems. CASRAI is not the body that governs or updates the Recipe Book itself — that sits with the Snowball Metrics Steering Group of participating universities.
Does Snowball Metrics replace responsible-metrics frameworks like DORA?
No. Snowball Metrics standardizes how a metric is calculated; it takes no position on whether or how heavily that metric should be used in an evaluative decision about a researcher or unit. Frameworks such as DORA and CoARA address the separate, evaluative question of appropriate metric use, and are typically treated as complementary to, not competing with, a calculation methodology like Snowball Metrics.
Key takeaways
- Snowball Metrics is a community-owned methodology (a published Recipe Book), not a database, product, or ranking — it standardizes how a research metric is calculated, not the data source used to calculate it.
- It began in 2010 with eight UK research-intensive universities and grew through a project involving CASRAI, with support from Elsevier and input from funders including the Wellcome Trust and the Medical Research Council.
- Governance is bottom-up: it is owned and steered by participating research-intensive universities across the UK, US, Australia, and New Zealand, not by a funder or commercial supplier.
- Recipes cover research income, collaboration, publication/citation activity, and commercialisation, and are free for any institution or system to adopt.
- It is methodologically complementary to responsible-metrics frameworks like DORA and CoARA, which address whether/how a metric should be used, not how it is calculated.
