OpenAlex and COKI Gain Ground as Free Alternatives to Web of Science and Scopus

The research reporting officer I spoke with spent the first week of February reconciling two records for the same mid-career plant scientist. Web of Science counted 14 publications and 312 citations. Scopus counted 17 publications and 389 citations. The difference traced to conference proceedings and a Brazilian journal indexed in only one database. She needed a number for a promotion committee, and the number depended on which subscription her institution could still afford.

The same author appeared in OpenAlex with 21 works and 402 citations. The higher count was not wrong; it included preprints, book chapters and Portuguese-language journals that the commercial indexes had not yet ingested. That extra coverage didn't solve her problem. It changed the question from which figure to report to which source a university is willing to trust.

The Practical Test: Open Data in Hiring and Promotion

That shift is now visible in research offices, national evaluation bodies and a few university rankings. Two open infrastructures have moved from side projects to plausible alternatives for institutions that can no longer justify paying for Web of Science and Scopus while budgets contract.

OpenAlex is a free and open scholarly catalog run by OurResearch, the nonprofit behind Unpaywall. It launched on 3 January 2022, weeks after Microsoft retired Microsoft Academic Graph, and absorbed those records before layering in metadata from Crossref, ORCID, ROR, DOAJ, Unpaywall, PubMed and other open registries. OpenAlex now indexes more than 250 million scholarly works, assigns unique IDs to works, authors, institutions, venues, funders and concepts, and offers the whole dataset through an API and monthly snapshots.

OpenAlex is not a polished portal in the way Scopus is. It is a data service. Researchers and administrators access it through an API, monthly data snapshots, or third-party tools built on top of it. A structured query can return every work attributed to an ORCID, a ROR institution ID, or a concept such as agroforestry, with citation counts, open-access status and funding acknowledgements attached.

COKI, the Curtin Open Knowledge Initiative, sits one layer above that raw catalog. Led from Curtin University, COKI builds open dashboards that aggregate publication, citation, collaboration and open-access metrics for universities, countries and research funders. A dean can pull up an institution profile and see what proportion of its outputs are openly available, which partners appear repeatedly, and how citation patterns shift by discipline, without a commercial login.

COKI's dashboards are built from similar open sources, including OpenAlex, Crossref and Unpaywall. Because the underlying scripts are open, an institution can inspect how a percentage is calculated and replicate it locally.

The strongest institutional signal arrived in June 2024, when the Centre for Science and Technology Studies at Leiden University released the first CWTS Leiden Ranking Open Edition, built entirely on OpenAlex. Every indicator in that edition can be recomputed from public data and open scripts, something no subscription-based ranking has offered. It did not replace the classic Leiden Ranking overnight; it ran alongside it, letting research managers compare what open data does and does not change.

Early adopters describe the tradeoffs in concrete terms.

• Coverage breadth. OpenAlex tends to find more works and citations outside high-income, English-language journals, partly because it ingests Crossref references and registered preprints. The plant scientist's Brazilian journal was the difference.
• Metadata quality. Author disambiguation and affiliation mapping are improving but uneven. Some institutions appear fragmented across name variants, and wrong attributions can surface in automated reports.
• Citation meaning. An OpenAlex citation count is not directly comparable to a Scopus or Web of Science count because source coverage differs. Committees using one number per candidate without context inherit new biases.
• Timeliness. OpenAlex often picks up newly published works and preprints faster than curated indexes, but speed can outpace retraction and errata updates.

Commercial indexes aren't passive here. Scopus AI has raised citation accuracy questions of its own, as this earlier analysis documents. The point is not that open data is clean and commercial data is flawed. The point is that every index encodes choices about what counts, and open catalogs at least make those choices inspectable.

Principles such as DORA have urged institutions to judge research on its own merits rather than on journal-level metrics. A longer look at research assessment reform beyond DORA shows how slowly those principles become operational criteria, even when committees want change.

COKI's dashboards push in that direction by reporting open-access rates, collaboration intensity and citation concentration rather than a single prestige score. The reporting officer who tested them found they answered a different set of questions than she was used to asking. Her old reports showed where her university's authors published. The COKI view showed what proportion of the university's accepted manuscripts were actually readable by anyone, and which disciplines were most dependent on paywalled journals.

For PhD-track researchers, the difference turns up in routine checks. A hiring committee that queries OpenAlex by ORCID may see a broader record than the candidate's Scopus list, including preprints, data sets, or chapters that commercial indexes never covered. That can advantage interdisciplinary and non-English-language work, but it also means candidates should verify how their ORCID-linked outputs appear in OpenAlex before an application screen.

She kept Scopus for one narrow task: verifying disciplinary journal coverage for tenure letters in a specialist subfield. For everything else, her office now uses OpenAlex as the base record and COKI for portfolio-level reporting. The annual report separates the data source from the performance indicator, so one vendor's coverage choices no longer shape a department's evaluation by default.

The cost differences are substantial, but they are not the whole story. A mid-sized university can spend tens of thousands of pounds or dollars each year on citation database licences before it pays for analytics add-ons. OpenAlex is free, and COKI's public dashboards are free. The real investment is staff time: cleaning affiliation records, checking false matches and writing documentation for committees that have never seen an OpenAlex ID.

What should be measured next is less glamorous. The rate at which non-English journals, retracted works and institutional affiliation changes flow into open catalogs compared with commercial ones. The number of research offices still running dual systems because they do not yet trust open metadata. The frequency with which a changed citation count changes a personnel decision. Those are the metrics that will show whether OpenAlex and COKI have become genuine alternatives or merely cheap supplements.

Photo by Trnava University on Unsplash