Sixty unanswered reviews: the order to clear them in

Sixty unanswered reviews feels like one job, and Google has published three things that make it several. Repetition is a named rejection reason, and it covers similar content rather than only identical content. Moderation is documented as usually taking minutes but sometimes up to 30 days, so a burst of replies is a burst of pending moderation you cannot inspect. And a reply that fails does so quietly. What follows is an order to work in, with the published constraint behind each step named.

The three published facts the order is built on

First, repetition. Google’s v4 Business Profile reviews reference lists REPETITIVE among the reasons a reply is rejected, and describes it like this:

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.

The Maps content policy states the same rule, with a single item under it: “Posting the same content multiple times either from the same account or multiple accounts”. A catch-up session is one way to produce that pattern.

Second, timing. Google’s help page on managing reviews:

Google reviews your replies: This makes sure that they follow Google’s content policies. If your reply isn’t approved to be posted, you’ll be asked to edit it. Replies usually take up to 10 minutes to review, but sometimes a review can take up to 30 days.

Third, visibility. A reply’s state is published in the API and nowhere in the dashboard. PENDING “Indicates that the reply is pending moderation”, and the reason field is “Optional. Output only. The policy violation that resulted in rejection. Only populated if reviewReplyState is REJECTED.” A bulk session can therefore fail one reply at a time, with no visible signal.

What Google does not publish is a rate. There is no recommended number of replies per day, per session or per week on any of these pages. A figure of the “three to five a day” kind circulates in this category and is in nothing Google published, so we will not present it as a rule. Where this article suggests a pace, it is ours.

Step 1. Split the backlog before you reply to anything

Google’s own tips page argues against the uniform sweep:

Make your replies count: Instead of sending the same “thank you” to everyone, focus on reviews where you can share a helpful update or answer a question. This shows people that you are listening and genuinely care about their experience.

So the first pass writes nothing. Sort the unanswered reviews into three piles.

Reviews with text. These are the only ones where a reply can be built out of what the customer actually said, which is also the only reliable way out of REPETITIVE.

Ratings with no text. Nothing specific to respond to, so replies to these converge by default. The same page also says “As best practice, reply to reviews and share your perspective”, which points the other way, and Google reconciles the two nowhere. For a backlog we take the focus instruction as the operative one, because it is the one written about sending the same thing to everyone.

Reviews that are not reply problems. Both help pages carry the same line: “If you believe that a review violates Google’s content policies, you can flag it.” Flag those and write nothing under them. What flagging achieves is covered in removing a Google review and reporting a fake one.

Step 2. Work newest first

The only published statement bearing on order is a tip on the same page: “Respond in a timely manner: A prompt response shows that you value your customers’ feedback and are committed to high-quality service.” Nothing Google publishes supports oldest first. That is a bookkeeping instinct, and our own guide to replying to negative reviews suggests it in passing, which we are correcting here.

There is a second published reason to favour recent reviews. Google states that “Customers can still change their review after reading your reply”, and that constructive replies “may even encourage the customer to update their review”. Google attaches no age limit to either outcome, so the weight we put on recency here is our inference rather than Google’s.

Step 3. Build each reply out of its own review

Three published instructions apply to each reply, and a long sitting is where we would expect them to slip.

“Keep it short and simple: Customers value direct and honest responses.” The long reply is the one written after you stopped reading properly.

“Be conversational, not promotional: Your reviewers are already customers, so avoid using your response to offer deals or promotions.” The urge to settle a backlog with goodwill gestures runs into this, and the API publishes ADVERTISING_AND_SOLICITATION as a rejection reason.

“Never share a reviewer’s private information or engage in personal attacks.” The details you reach for to prove somebody really was a customer are the ones this covers.

One instruction we cannot resolve: the same page says “Personalize your reply: Address the reviewer by name and acknowledge their specific feedback”, while the API publishes PERSONAL_INFO as a rejection reason. Google does not reconcile them and we have not tested it, so we will not tell you that using the name is risky or that it is safe.

Step 4. Size the sitting yourself

Google publishes no rate, and REPETITIVE publishes no threshold: the word “similar” carries the rule and is defined nowhere in it. So no number we or anyone else offers is Google’s number.

Ours, and the reason for it: stop the sitting when you can no longer answer “what was different about that one” for the reply you just wrote. Similarity is the undefined term in the only rule that applies, and your own inability to tell two replies apart is the closest thing to evidence of it. A working suggestion, not a published limit.

Step 5. Then wait, and do not rewrite on day two

The stated outer bound is 30 days. A batch written in one evening is a batch sitting inside that window with no state shown anywhere, so a reply missing on day three is not evidence of anything.

Google documents the edit path plainly: “Under the reply you want to edit, select Edit”. It does not document whether an edited reply re-enters moderation, or how long that takes, so a batch of hasty edits is a batch of unknowns rather than a fix.

Step 6. Check the batch, because nothing else will

Google says that if a reply is not approved “you’ll be asked to edit it”, and does not say where or when that request appears. The dashboard exposes no state per reply, so the only free check is to open your public profile and read down it, confirming each reply is there. At backlog scale it is the step we would expect to be dropped first, and the only one that catches a silent failure.

Tools that read the API see the states directly; what that does and does not buy you is in review management software, and the full list of reasons a reply is refused is in what happens when Google rejects your reply.

What we will not tell you

There is no published replies-per-day rate, and no published threshold for how different two replies must be. Google does not say whether an edited reply is moderated again, or whether repetition is judged within one profile, across an account, or across a session. The nearest published sentence is about content review generally: Google takes into consideration “the content itself, account information, user actions and whether a pattern of harmful behavior exists”.

The short version

Split the backlog first, work newest first, write each reply from its own review, stop when the replies stop being distinguishable, wait out a window that can run to 30 days, then read your own profile to see which replies landed. The moderation window, the repetition rule and the read-back are each in a Google page you can open. Two parts are ours: the pace, because we found no published one, and newest first, which is our reading of a tip about responding promptly rather than an order Google prescribes.