How to Turn Negative Reviews Into a Real Next Step with AI

Every guide on review management tells you to respond to negative reviews. Almost none of them tell you what the response should actually achieve.

The default AI reply — from most tools, including Inform before this feature — is a polite apology. "We're sorry to hear about your experience. We take this feedback seriously." It is inoffensive and completely inert. The unhappy customer has nowhere to go, and the next person reading the listing sees a business that acknowledges problems without solving them.

The replies that actually recover customers do three things: they acknowledge the specific complaint, they take ownership, and they end with a concrete next step. Something the customer can act on.

That last part is the one automation usually drops, because the next step is different for every business — and often for every location.

Handled by hand, this means:

• Watching for one and two star reviews as they arrive.
• Writing a bespoke reply that names the right contact for that branch.
• Remembering which location has which phone number.
• Doing it quickly, because a late reply reads worse than a quick one.

Miss a few days and the unanswered negatives pile up exactly where new customers look first.

The major platforms all generate competent replies to negative reviews. Birdeye's Review Response Agent reads sentiment, intent and business context and can pull custom tokens such as employee or location names into the response. That is genuinely good.

What almost none of them do by default is finish the reply with a route to resolution that is correct for the specific branch being reviewed. The reasons are structural rather than lazy:

• The tool does not know your escalation contact. A generic model will not invent a phone number, and it should not.
• One global contact is wrong the moment you have two locations. Giving a Belleville customer the head-office number is worse than giving them nothing.
• Human-in-the-loop approval is offered as the safety net, which is sensible for compliance but puts the work back on a person — and the person is who you were trying to free up.
• Per-location pricing means the businesses with the most branches, and therefore the most need for per-branch contacts, pay the most to get it.

Inform resolves the contact at reply time from the location's own Google listing. One setting, correct for every branch, no list to maintain.

Inform adds Review Remediation. When the weekly auto-reply answers a negative review, it ends with a real next step — a phone number, an email address, or a short custom instruction — instead of only apologising.

You configure it once, and it applies to negative reviews across every connected location.

In Org Settings, find the Review Remediation section. The enable toggle controls whether negative replies end with your contact.

It is on by default, but nothing changes until you choose a contact — so there is no risk of an unconfigured setting quietly altering your replies.

The dropdown offers four options, and the right one depends on how your business handles complaints:

• Each store's Google phone number — the dynamic option. Every location's reply uses that location's own phone number, pulled from its Google listing. This is the one most multi-location businesses want.
• Phone — one number for the whole organisation.
• Email — one email address for the whole organisation.
• Custom message — a short free-text next step, for when the right action is not a phone call.

If you run several branches, start with Each store's Google phone number. It means a customer complaining about your Belleville store is given the Belleville number, without you maintaining a list.

Once configured, negative replies change shape. Here is a one-star review about cleanliness and value, answered automatically.

The reply names the specific complaints, takes ownership, and then gives the customer that location's phone number and a reason to use it.

Remediation applies to reviews rated one or two stars. Three star and above are handled normally — a customer who left three stars and a mild note does not need to be told to phone the manager.

It runs as part of the weekly auto-reply, so if you already have auto-reply enabled there is nothing else to switch on.

Email is available as a static option — one address for the whole organisation — but there is no dynamic per-store email equivalent to the phone option. Google's business listings expose a phone number per location but not an email address, so a per-store email is not something we can pull reliably. If you need per-location routing, the store phone option is the one to use.

Reviews rated one or two stars. Higher-rated reviews get a normal reply without a remediation step appended.

Yes — that is the "Each store's Google phone number" option. Every reply uses the phone number from that specific location's Google listing, so you do not have to maintain a mapping yourself.

Use the Custom message option and write the next step you do want — visiting a feedback form, asking for a manager in store, or anything else that fits how you handle complaints.

No. The toggle defaults to on, but with no contact set the auto-reply behaves exactly as it did before. Nothing changes until you choose a contact type and value.

Negative reviews are the ones worth getting right:

• ✅ Replies that resolve — every negative reply ends with a real next step.
• ✅ Per-location contacts — each store can offer its own phone number automatically.
• ✅ Safe by default — nothing changes until you configure a contact.
• ✅ No extra workflow — it rides the weekly auto-reply you already use.
• ✅ Free to use — no add-on tier for handling complaints properly.

Related guides

• Learn how to make replies completely hands-free.
• Or see how to avoid the most common review management mistakes.