How will Google Search handle multiple Rich Snippet entries

Question

Background

• I've started to use Article markup (https://schema.org/Article) on a WordPress website. The aim is to add as much as possible properties.
• By using a (popular) plugin, basic Article data is parsed as JSON-LD which Google Rich Results Test is marking as eligible.
• I feel that the plugin does not cover enough properties, so I started to add additional properties using (mainly) microdata.
• By disabling the plugin I lose other functionality, which I feel is not worth adding manually.

Without spending too much time on this (traditional question being: is it worth it), I have the following problems and questions:

Problems

By using this approach, I'm getting two sets of Article entries in the Google Rich Results Test (one for each of the syntax)

To, maybe, be able to answer better:

• The IDs of the two entries are different (is that maybe the reason for the two entries? - Edit: I've matched the ID's but two entries are still created)
• Other properties (e.g. the updated date) in the microdata might be different
• There will be additional properties in the microdata

Question

Will Google use the first entry, the second entry (i.e. latest variables) or a combination of the two entries?

Possible outcomes

• Fix the problem
• Continue with two entries based on how Google handles the entries
• Remove the microdata altogether

Answer 1

If you use the same syntax, then you can merge entities via id. But Google does not merge them between mixed syntaxes.

If there are multiple entries for the same entity, I have tended to see Google pick one and ignore the others.

Sounds like your solution is to change your customisations to json-ld and include the same id as generated by the plugin.