Each month we at the tech support team see people reaching out because they’ve spotted something off in a DOI’s metadata and want it fixed. We get a steady stream of these, so I thought it’d be worth walking through what we usually see, who’s asking, and, most importantly, what the standard procedure is that we follow in such cases.
A few types of requests come up again and again. Someone notices a DOI is displayed on the publication landing page but is not actually registered. Someone else finds that a DOI resolves to a dead link or the wrong page entirely. Sometimes it’s about the author’s name in the contributor list that is wrong or missing, or the reference list that looks incomplete.
The people reaching out are just as varied. A lot of the time it’s the authors themselves, trying to get their own record corrected. Other times it’s a researcher who stumbled across the issue while reading or citing something. We also hear from librarians and from the general metadata users out there that look up our metadata and end up discovering these errors along the way.
Metadata travels far and wide
It’s easy to see why metadata accuracy matters so much to the scholarly community as a whole. Once a member deposits metadata with us, it’s openly available through our APIs, and from there it gets picked up and used by all sorts of research tools and services - you can get a sense of just how much traffic that involves in this post on “Rebalancing our REST API traffic” and a good explanation of who those downstream users actually are in this ticket of the month on “Who uses your metadata”. So if something’s inaccurate in our system, that error doesn’t just sit still, it spreads everywhere that record travels.
And yet, correcting it isn’t as simple as us just stepping in. Here’s the thing we always have to explain, since a fair number of metadata quality improvement requests come directed to us: members are the ones who’ve approved their content for publication and registration in the first place, which makes them the official gatekeepers of both that content and the metadata describing it. Because of that, Crossref doesn’t update, edit, or correct publisher-provided metadata directly - it is the publisher’s responsibility, and they’re the only ones who can authorize a change to it.
Helping our members with metadata improvements
Still, we’re, of course, not entirely hands-off. So to what extent can Crossref help with a correction? When our members join Crossref, they provide at least three different contacts to cover five roles: primary, voting, technical, metadata quality and billing contacts. These are the people we turn to when confirming any changes to the account or the metadata. You can find the full list of contact types and their responsibilities here.
This ensures that we have a designated contact for each of our members for technical and metadata quality issues. Depending on the nature of the metadata quality improvement request we receive, the first thing we do is confirm there is actually something to fix - sometimes what looks like an error is really just a timing issue, since it can take a little while for a new or updated record to show across all of our services (we’ve written about where and when your DOI will show up in the post “A DOI-namic timeline”). Once we’ve confirmed there is a genuine issue, we get in touch with the designated technical or metadata quality contact, pass along the information we received from whoever reported the issue, and request the correction. We do have timelines in place, and we follow up in cases where the contact isn’t responsive, including other contacts in the correspondence where needed, to try to resolve the issue in a timely manner.
That said, we act as an intermediary in metadata quality improvement cases, not the decision-maker. Please always keep in mind that the most straightforward way to correct or update metadata is to contact the member directly.
Now, there are a few rare cases where we do step in and update something ourselves, but these are genuinely the exception. It generally comes down to a member being unreachable, combined with something that poses a real problem for anyone landing on that DOI - like a link that’s broken with no fix in sight, or a DOI resolving to something clearly incorrect or inappropriate.
If you have a particular case you’d like help with, you can post in the Crossref Community Forum’s Metadata quality improvement category, reach out to us via the contact form or by emailing support@crossref.org and we are happy to follow the process described above.