The Research Excellence Framework (REF) is the United Kingdom's national assessment of research quality, and REF 2029 will once again shape funding, reputation, and strategy for UK institutions. The institutions that approach it calmly are the ones whose underlying research information is already reconciled and current. The ones that approach it in crisis are those rebuilding the evidence base from spreadsheets months before the deadline.
What REF actually demands of your data
REF assessment rests on outputs, their correct attribution to units of assessment, and the narrative of environment and impact. Each of those depends on a clean, complete, and consistently structured record of institutional research — exactly the data that is fragmented across global indexes and internal systems at most universities.
A research-information readiness checklist
- Output completeness. Every eligible output captured, including those in open-science sources a single index may not cover.
- Correct attribution. Author and unit affiliations disambiguated so outputs map to the right unit of assessment.
- Deduplication. The same output never counted twice or recorded inconsistently across systems.
- Impact evidence. Traceable links from research to societal outcomes, including SDG alignment, maintained continuously rather than reconstructed.
- Environment data. Collaboration, funding, and people data consistent with the output record.
Why continuous beats last-minute
REF cycles span years, but readiness is decided in the months before submission only if the data was neglected in between. A RIMS keeps the record continuously reconciled, so REF preparation becomes verification and narrative work — not data reconstruction. The same reconciled dataset also serves QS, THE and ARWU submissions and accreditation, removing the duplicate effort of preparing each in isolation.
From data to narrative
REF is not only a counting exercise; it rewards a credible account of environment and impact. That narrative is far stronger when it is evidenced by a defensible dataset — collaboration maps, impact alignment, and output trends that leadership can stand behind. SDG mapping and a single source of truth are the practical foundations.
Frequently asked questions
When should REF 2029 data preparation start? It should never stop. Continuous reconciliation means there is no separate "start" — only verification before submission.
Does open-science coverage matter for REF? Yes. Outputs missing from a single curated index still count; incomplete capture understates eligible output.
Getting ready
Discover RIMS keeps REF-relevant data — outputs, attribution, collaboration, and impact — continuously reconciled across Scopus, OpenAlex, ORCID, Crossref, and Scimago, so REF 2029 is a verification exercise, not a reconstruction.