Why the same reply under thirty reviews gets flagged, and what the repetition values mean

Google publishes three separate rejection reasons that a copied review reply can land on, and the strongest of them is REPETITIVE. Its published description is one sentence long, and the part of that sentence which matters is four words: repetition is counted across accounts, not only within one. Templates, saved snippets and reply generators all produce repeated text, and this is the published sentence they run into.

The three values, verbatim

There are 43 values in the PolicyViolation enum on Google’s v4 Business Profile reviews reference. Three of them describe repeated or meaningless content. Google’s descriptions, exactly as published:

Value	Google’s description
REPETITIVE	“Content that violates our policy on repetitive content isn’t allowed. Repetitive content includes identical or similar content added multiple times by one or more accounts.”
GIBBERISH_AND_REPETITIVE	“Content that violates our policy on unclear or repetitive content isn’t allowed. Unclear or repetitive content dilutes useful information others can find on Google.”
UNCLEAR	“Content that violates our policy on unclear content isn’t allowed. Unclear content includes random characters or other content that lacks meaning.”

The word “gibberish” appears once on that entire reference page: in the name of the second value. Its own description never uses the word, and describes unclear or repetitive content generally.

Three enum values, two policy rules

The policy those values point at is Google’s Maps User Contributed Content Policy, which puts both behaviours under one heading, “Unclear and Repetitive Content”, and then splits it in two.

On repetition:

We do not allow the posting of repetitive content on Google Maps because it dilutes useful information Google Maps provides its users.

with one item under it: “Posting the same content multiple times either from the same account or multiple accounts”.

On unclear content:

We do not allow the posting of unclear content on Google Maps because it dilutes useful information Google Maps provides its users.

also with one item: “Posting content that lacks meaning across languages, such as a random group of characters”.

So the API’s three values collapse onto two published rules. UNCLEAR maps to random characters, REPETITIVE maps to posting the same thing repeatedly, and GIBBERISH_AND_REPETITIVE sits across both and quotes the rationale the two share. A rejection could plausibly arrive under any of the three, and none of them tells you which sentence of your reply triggered it.

The clause that matters is “or more accounts”

Read the two sources side by side. Google’s enum says repetitive content includes “identical or similar content added multiple times by one or more accounts”. The policy page says the same thing in its own words: the same content posted “either from the same account or multiple accounts”.

Both published statements count repetition across accounts. Neither says anything about the repeated content having to sit on one business profile.

That is the part with consequences. If an agency writes one house reply and pastes it under reviews at forty client locations from forty different logins, the published description of REPETITIVE describes what happened. The same is true of any tool that ships a reply library to its whole customer base: the identical content is being added multiple times by more than one account, which is the exact wording.

We are not going to tell you that Google enforces it that way. Neither page says whether cross-profile repetition is evaluated together, and we have not tested it. What we can say is that the published wording does not carve out separate businesses, and that Google reasons about multi-account patterns elsewhere in the same policy: under fake engagement it prohibits “Content that has been posted from multiple accounts by or at the request of one person.”

What “similar” cannot be pinned to

REPETITIVE covers “identical or similar content”. The word “similar” appears exactly once on that reference page, in that sentence, and does not appear on the policy page at all. Neither page defines it.

There is no published threshold. Google does not state a percentage overlap, a word count, a number of repetitions, a time window, or any measure of how different two replies must be. Nothing on either page distinguishes this:

Thanks so much for the kind words! We really appreciate your support and hope to see you again soon.

from this:

Thanks so much for the feedback! We really appreciate your support and hope to see you soon.

They are not identical. Whether they are “similar” in the sense Google means is not answerable from anything Google has published. If you see a rule in this niche about how much a reply must vary, it did not come from these pages, and we will not invent one to fill the gap.

If something else writes your replies

This is the practical problem with generated replies, and it is not the complaint about them sounding robotic.

A reply generator given the same instruction and a five-star review with no text has almost nothing to vary on, because there is nothing in the input to be specific about. That is our reasoning and we have not measured it. Paraphrasing the template harder does not clearly escape the published wording either, because the wording already covers “similar” content. None of the vendors we have read in review management software address REPETITIVE at all.

The one thing that reliably is not identical or similar content is a reply built out of what the specific review said. That is the same conclusion as replying to a negative review well, arrived at from the moderation side rather than the persuasion side.

You will not be told it happened

None of this shows up where you work. REJECTED is documented as “Indicates that the reply was rejected”, and the reason field is “Optional. Output only. The policy violation that resulted in rejection. Only populated if reviewReplyState is REJECTED.” Both live in the API. The full list of reasons, including the large group that has nothing to do with your words, is in the published rejection reasons.

The enum is also explicitly open. Google documents PolicyViolation as “Represents standardized policy violations. New values may be added in the future. Clients should be prepared to handle unknown values.” A fourth value naming repetition would not be surprising.

What we have not verified

Three limits, stated rather than glossed over.

We do not know how often REPETITIVE actually fires on a reply. No published figure exists that we can check, and we are not repeating one from a vendor’s account of its own accounts.

We do not know whether replies are moderated by the same system as reviews. The enum is served for replies, but the policy it points at is written for contributions generally, and the repetitive and unclear sections contain no owner-specific or merchant-specific note, unlike several other sections of the same page.

And we do not know where the line sits. That is the honest end of this: Google published a rule against similar content and published no definition of similar.

The short version

Repetition has two names in Google’s published rejection list and a third adjacent value, and the strongest of them counts identical or similar content “by one or more accounts”. Every template-driven approach to review replies is producing exactly the pattern that sentence describes, at whatever scale it is run. The threshold is unpublished, so the only safe reply is one the review itself made necessary.