Skip to content
search-engine-optimizedby Netsolution

Image licence metadata

Two means declare an image's licence, and one is enough: structured data on the page, or metadata embedded in the file. The first must be repeated for every use of the image, the second travels with it. Where the two conflict, the structured data wins.

Two declarations, two scopes

A published image travels: it is reused, cropped, republished, sometimes credited. The documentation provides two ways of saying who made it and what may be done with it, and the difference between them is material rather than technical.

Two means declare an image's metadata, and only one is needed to be eligible for enhancements such as the licensable badge: structured data on the page, or photo metadata embedded in the file.

officialSource : Google Search Central, “Image metadata in Google Images”

Structured data ties the image to the page where it appears and must be added for every instance of the same image; embedded photo metadata travels with the file from page to page and is written once per image.

officialSource : Google Search Central, “Image metadata in Google Images”

There is the real trade-off, and it does not arise in the terms you would expect. Structured data describes an image on a page: reused elsewhere, it must be described again. Embedded metadata lives in the file itself: it follows the image everywhere, including where nobody asked for it.

For a newsroom publishing the same photograph across twelve pages, the difference is counted in lines of template. For a photographer whose images circulate off their site, it is counted otherwise.

The rule that settles it

Where both forms coexist and any information conflicts, the structured data is used.

officialSource : Google Search Central, “Image metadata in Google Images”

That sentence settles a very concrete case. A file arrives with its fields filled in by its author; the content management system, meanwhile, generates structured data automatically from fields left empty or filled with defaults. The page’s declaration wins, and the original credit disappears with no error reported anywhere.

Our plugin adds licence structured data to all your images automatically: credits and licences filled in with one click, badge guaranteed.

No enhancement is guaranteed by the documentation. And a plugin filling those fields automatically overwrites, in case of conflict, the credit the author embedded in the file, since structured data takes precedence.

What the declaration must contain

Two distinct addresses are expected: one for a page describing the licence governing the image's use, and one for a page explaining how to acquire that licence.

officialSource : Google Search Central, “Image metadata in Google Images”

Two addresses, two jobs: stating the conditions under which the image may be used, and stating where the right to use it is obtained. Many sites supply only one, usually the legal notice page, which fills neither role.

Pages containing the images must be viewable without an account or login, and reachable by the crawler: not disallowed by robots.txt and not marked for non-indexing.

officialSource : Google Search Central, “Image metadata in Google Images”

The feature is described as available on mobile and desktop, in all regions and languages where Search is available.

officialSource : Google Search Central, “Image metadata in Google Images”

What it is worth when nothing shows

Embedded metadata keeps its value even with no particular display in the results: it stays readable by anyone who takes the file, which makes it the only credit that survives a copy.

hypothesis

The status is that of a reading, and it follows from how the mechanism is described: the documentation writes that embedded metadata travels with the file. We conclude that a credit written into the image is the only one that survives a copy, which has a value independent of any display in the results.

Declaring licence and credit in both the file and the page

hypothesis

Effort
moderate
Time to effect
months
Risk
none
Durability
structural

The effort lies in the publishing pipeline rather than the markup: you have to check that nothing strips the metadata along the way. The delay is medium, the display being neither guaranteed nor immediate. The risk is nil. Durability is good, a properly written credit staying valid as long as the file exists.

What does not work, or no longer works

Pointing to the legal notice page as a licence. Two distinct addresses are expected, and neither is a general legal page.

Then relying on a plugin that fills those fields automatically. In case of conflict its declaration overrides the author’s.

Finally, filling in the original without checking the file served. Compression tools often remove embedded metadata, and the check must cover what the site actually publishes.

Sources

  1. Image metadata in Google Images — Google Search Centralaccessed on
  2. Review snippet structured data — Google Search Centralaccessed on

Block revalidated on

Frequently asked questions

Do you have to choose between structured data and embedded metadata?
No: the documentation writes that one form is enough for eligibility. The choice is about scope. Structured data describes an image on a given page and must be repeated for every use; embedded metadata follows the file and is written only once.
What happens when the two conflict?
The structured data is used. The rule is worth knowing for newsrooms receiving files already filled in by a photographer: a template that overwrites the embedded credit with a generic value imposes its version, including when that version is wrong.
Which addresses must be supplied?
Two, and they do different jobs: one names the page describing the licence governing the image's use, the other the page where that licence can be acquired. A single page mixing both roles leaves the declaration ambiguous for reader and engine alike.
Can an image behind a login be covered?
No. The documentation asks that pages containing the images be viewable without an account or login, and reachable by the crawler. It is the same condition as for all structured data: what cannot be reached cannot usefully be declared.
What is the point if the badge never shows?
That the credit outlives the file. A copied image loses its page, therefore its structured data, but keeps its embedded metadata. That is our reading, not a promise of display: no enhancement is guaranteed, and the documentation commits to none.