How to choose a research impact and bibliometrics tool

Research impact and bibliometrics tools answer a narrower question than a generic search engine: not just "what has been published on this topic," but "how has it been cited, funded, shared and reused, and how does that compare across researchers, institutions or fields." Research offices, librarians, funders and individual researchers all use these tools, but for different jobs — literature discovery, formal impact reporting, institutional benchmarking, or tracking attention outside academic citation altogether. Picking the wrong one usually means picking a tool built for the wrong job, not a bad tool.

Literature search or management-facing analytics

• Literature discovery and citation search. Scopus, Web of Science, Dimensions, OpenAlex and Semantic Scholar are all, at core, searchable indexes of publications and their citations, aimed at a researcher looking for and citing related work.
• Management-facing benchmarking. SciVal is built on Scopus data specifically to answer institutional questions — how does our output compare with named peers, which topics are gaining momentum, who are potential collaborators — rather than to support day-to-day literature search. If your actual users are research-office staff producing funder reports, not researchers searching for papers, a benchmarking layer like SciVal (or Dimensions Analytics) matters more than raw search features.
• Network visualization on top of a citation index. VOSviewer is not a database at all — it's a free desktop tool that imports data from Scopus, Web of Science, Dimensions, PubMed, OpenAlex or Lens.org exports and turns it into co-authorship, co-citation and keyword-co-occurrence maps. You need one of the indexes above before VOSviewer has anything to visualize.
• Attention beyond citations. Altmetric tracks how a paper is discussed in news, social media, policy documents and blogs, producing an Attention Score that measures visibility and discussion, not research quality or scholarly influence — a genuinely different signal from anything the indexes above report.

Editorial curation versus broad coverage

Web of Science's Core Collection applies editorial selection criteria before a journal is added, which institutions cite as an argument for quality over raw breadth; it also underpins the Journal Citation Reports and the journal impact factor itself. Scopus generally indexes a larger set of journals and conference proceedings with a lower bar for inclusion, trading some selectivity for coverage. Neither is objectively "more accurate" — the right choice depends on whether your field is well served by curated journals (favoring Web of Science) or spans conferences, regional and interdisciplinary venues that a broader index picks up better (favoring Scopus). Citation counts for the same paper will differ between the two because they draw on different underlying sets of sources, so pick one as your institution's system of record rather than mixing counts from both in the same report.

Free and open versus paid institutional subscription

OpenAlex, Semantic Scholar, VOSviewer and the free tier of Lens.org and Dimensions cost nothing and are funded by nonprofits, grants or philanthropic support rather than subscriptions — OpenAlex by OurResearch, Semantic Scholar by the Allen Institute for AI, VOSviewer by Leiden University's CWTS, The Lens by the nonprofit Cambia. Being free doesn't mean being identical in coverage: OpenAlex and Semantic Scholar cover preprints, non-Western scholarship and, in Semantic Scholar's case, computer science and biomedical fields especially well, but both can still trail Scopus or Web of Science in disambiguation quality and coverage of long-tail journals in some fields. Scopus, Web of Science, SciVal and full institutional access to Dimensions are paid subscriptions with no public price list, aimed at institutions that need consistent, defensible numbers for formal reporting.

Does the tool connect publications to anything beyond citations?

Dimensions differentiates itself by linking a publication to the grant that funded it, the patents that cite it, the clinical trials that use it and the policy documents that reference it, all in one graph — useful for a funder or research office tracking impact across the research lifecycle rather than citation count alone. The Lens does the same specifically for patents, combining scholarly literature search with global patent data so technology-transfer offices can trace research into later inventions, something neither Scopus nor Web of Science does natively. If your question is purely "how many times was this cited," any of the citation indexes answer it; if it's "did this research lead anywhere — funding, a patent, a policy," Dimensions or The Lens are built for that question specifically.

How pricing works

Scopus, Web of Science and SciVal are sold as institutional subscriptions with no published list price — expect a quote scaled to your institution's size and the modules included. Dimensions and The Lens are freemium: free public search with reduced functionality, and a paid institutional analytics tier sold by quote. OpenAlex, Semantic Scholar, VOSviewer and Altmetric's public attention badges are free; Altmetric Explorer, the institutional version, is quote-based. None of these publish self-serve pricing, so budget for a sales conversation regardless of which tier you land on.

A shortlist by situation

• If you need a selectively curated index for formal impact reporting and already use EndNote or InCites, look at Web of Science.
• If you want broader coverage, especially alongside institutional benchmarking, look at Scopus and SciVal.
• If you're a funder or research office tracking publications through to grants, patents and policy, look at Dimensions.
• If you need free, bulk-accessible scholarly metadata with strong preprint and non-Western coverage, look at OpenAlex.
• If you want fast, free literature triage with AI-generated summaries, especially in CS or biomedical fields, look at Semantic Scholar.
• If your focus is innovation and IP — connecting literature to patents — look at The Lens.
• If you need to visualize co-authorship or co-citation networks from data you already have, look at VOSviewer.
• If you want to track online attention and public discussion separately from citations, look at Altmetric.

Questions to ask a vendor or in a trial

• Which journals, conferences and regions does your index cover well, and where does coverage thin out for our specific fields?
• Can we export raw data for our own reporting and for tools like VOSviewer, or are we locked into your dashboards?
• How do you handle author and institution disambiguation, and how much manual cleanup does that leave us?
• If pricing is bundled with another product (Scopus with SciVal, for example), what does each part cost standalone?
• How is your underlying data funded or governed, and does that affect what we're allowed to do with exported data?

Common mistakes

• Comparing citation counts for the same paper across two different indexes and treating the discrepancy as an error rather than a coverage difference.
• Buying a management-facing benchmarking tool when what researchers actually need is better literature search, or the reverse.
• Assuming a free, grant-funded tool has worse data than a paid one in every field — coverage varies by discipline, not just by price.
• Mistaking an Attention Score for a measure of research quality rather than online visibility.
• Standardizing on one index for internal reporting without checking your funders or evaluators expect numbers from a different one.

For two direct comparisons, see Scopus vs Web of Science and OpenAlex vs Semantic Scholar. Every tool in this category is listed at every tool in this category.