Is Google Scholar an academic database?

The academic community is not decided on this topic, calling Google Scholar sometimes a database and sometimes a search engine.

There is no general definition of criteria that have to be met by a service to qualify as an academic, bibliographic or citation database. While related services such as Web of Science, Scopus, or Medline all declare on their websites that their offered services are databases, Google Scholar does not do so, and that's the root of this dilemma.

Two key features of academic databases are missing in Google Scholar

Google Scholar does not search the entire public web, but limits it's scope to resources from academic publishers, universities, and academic repositories. While Scopus and Web of Science have an editorial team that manages a list of sources that are included in their database, Google Scholar adds to its index anything that looks like an academic article, research report, thesis, working paper, or book chapter according to its built-in algorithm. That's not a bad thing at all and greatly expands the academic universe.

But there are two key features that databases like Web of Science, Scopus, PubMed and others all offer and Google Scholar does not. First, a stable document identifier that uniquely identifies a document. That allows to just provide the identifier or a link containing the identifier for others to retrieve the same information that you have. Copying the link in the browser tool bar of your Google Scholar search is not the same thing, even if you currently only see one result. It's not guaranteed that the same search will give the same result. The reasons are plentiful: your search results might be influenced by your geographical location or your browsing history. Google Scholar is adding thousands of new articles per day, which will definitely change the results of search results.

A second key feature of academic databases is that documents that have been indexed once are not removed. Because of the way Google Scholar works it's not guaranteed that the document that you seen now in your results will also appear in future searches. Google Scholar constantly crawls the web for new sources, but also checks the ones that it has already added if they are still present. If Google Scholar finds that e.g. a working paper that is already indexed was removed from the university repository and there is no other copy available on the web, it will also delete it from its own search index. Again it all boils down to the fact that by simply taking the Google Scholar search URL in your browser tab it's not guaranteed that opening this URL again will give the same result.

Google Scholar is an academic search engine

Our conclusion is that Google Scholar should be referred to as an academic search engine an not an academic database. The main reason for this decision is that it lacks a stable document identifier and that it is not guaranteed that a once added document will also be shown in future search results.