The Local Review Generation and Response Playbook

Two rulebooks apply, and they are not the same size. The Federal Trade Commission's Trade Regulation Rule on the Use of Consumer Reviews and Testimonials, codified at 16 CFR Part 465, targets fake or false reviews, buying reviews conditioned on sentiment, undisclosed insider reviews, businesses presenting controlled sites as independent, review suppression, and fake social media indicators. It is federal law with civil penalties attached, and it applies to the business whether or not a review platform notices anything. Google's contributor policy is a platform rule. It cannot fine you, but it can remove reviews, and, in our view, repeated breaches put the profile at risk.

The difference matters on incentives. The FTC's published questions and answers say the rule does not prohibit giving incentives for reviews, as long as there is no express or implied requirement that the review express a particular sentiment, and that incentives should be disclosed where a reader would otherwise not know. Google's policy, as we read it, goes further and covers reviews posted in exchange for money, products, or services; confirm the current wording. Following the federal floor is not enough on Google. A discount, a raffle entry, or a free service for a review is the thing to remove from the process, even where federal law might tolerate it.

Review gating is the other gap. Part 465 does not use the word and, per the FTC's questions and answers, contains no specific prohibition on asking only customers you think are happy. The FTC says that practice could still violate the FTC Act, and the rule does bar misrepresenting that displayed reviews reflect most or all submitted reviews when reviews are being suppressed based on rating or negative sentiment. Google's policy, as we read it, also discourages selective solicitation. Read the current text of the rule and the policy before you build the flow, because both can change, and have your own counsel review anything customer-facing. This guide is operational advice, not legal advice.

The compliant version of a review request is boring on purpose. The same message goes to every customer who completed a job, from the same trigger, with the same link. The trigger is a job status in your CRM or field-service tool, such as "completed and paid", not a satisfaction score. Nothing in the flow looks at sentiment before deciding who gets the link. If you also run a private feedback channel, such as a reply-to address in the message, offer it to everyone alongside the public link rather than as a fork that sends unhappy customers away from Google.

Timing is a judgement call, and the right answer differs by trade. A reasonable default is to send the request when the customer has seen the result and had a chance to confirm it works, which is the same day for a plumbing repair and probably a few days later for a roofing job or a kitchen remodel. Send one reminder if there is no reply, three to five days later, and then stop. A third message starts to read as pressure, and pressure produces the kind of rushed, emotional review you would rather not publish under your name.

Use the direct review link from the profile's dashboard, not a link to your own page that asks for a rating first. A short text works better than email for most service businesses because it opens on the phone the customer already has in hand. Messaging rules still apply to the text itself, which is covered in the missed-call text-back guide. Add a QR card for in-person jobs, handed over by the technician at the door, with the same neutral wording.

Most people who read your replies are prospects deciding whether to call, and they are reading the reply to a one-star review more closely than the five-star ones. That changes the job. The goal of a negative-review reply is to show how the business behaves when something goes wrong, in three or four sentences, without arguing the case in public. State what you understand happened, say what you are doing about it, and give a direct way to continue offline. Do not reveal customer details, job addresses, or anything covered by a confidentiality duty, even to defend yourself. Confirming someone was a customer can itself be more than they would want public.

Positive reviews need a reply too, but not a fresh essay each time. Thank the person, name the specific job or tech if the review does, and stop. Templates help a team keep the tone steady across locations, as long as each reply picks up at least one real detail from the review. Replies that repeat the same sentence forty times are easy to spot, and they signal a business that does not read its reviews. Keep promotional language, phone numbers inside the reply body, and keyword lists out of them. Google may moderate replies as well as reviews, so a reply that includes prohibited content may be held back.

Reply timing is a service-level choice. A same-business-day target for negative reviews and a same-week target for positive ones is a sensible starting point, written down with a named owner, as the multi-location profile guide recommends. Treat those numbers as a recommendation to adjust, not an industry benchmark.

Some reviews should not get a templated reply. Route these to a manager before anything is posted: allegations of injury, property damage, or illegal conduct; reviews that name a technician; reviews that mention a lawsuit, a claim, or a regulator; and reviews that look coordinated, such as several new accounts posting similar text on one day. The escalation owner decides between a reply, a flag, and a private call to the customer, and logs the decision so the next person does not repeat the discussion. Where a safety or legal exposure is possible, bring in whoever handles your insurance or counsel before replying in public.

Flagging is for content that breaks the platform's policy: spam, off-topic content, fake engagement, conflicts of interest, harassment, and the like. It is not a way to remove criticism that is accurate or merely harsh, and in our experience a platform is unlikely to remove a review just because the business disagrees with it. Keep a record of each flag with the date and the stated reason, because follow-up is often needed and the dashboard will not hold the history for you.

The FTC rule adds a hard limit to the other side of this. Its review suppression provision covers using an unfounded or groundless legal threat, a physical threat, intimidation, or a public false accusation made knowingly or with reckless disregard for its truth, in an attempt to prevent a review or have it removed. A threatening letter to a customer with a real, if unflattering, complaint is the kind of thing to avoid. Likewise, a non-disclosure or non-disparagement clause that stops customers from posting honest reviews can create legal risk under consumer-review laws. Have counsel check customer contract language before you rely on it.

Set this up once per business, then repeat the audit quarterly. Each step has a gate: nothing moves forward until the check can be shown, not merely claimed.

Track what the process controls. Those are request volume (jobs completed against requests sent), reply coverage (the share of reviews answered inside your own target), time to first reply, and the count of escalations and their outcomes. Review count and average rating are results, and they move for reasons outside the process, such as a bad month for the service itself. Report them next to the controllable numbers and do not set a target on the star average, because a target on the average is how teams drift into gating.

A rating that stays low after a clean process is information about the business, not about the review flow. Read the actual text of the last twenty reviews and sort the complaints. If three of them mention missed arrival windows, that is a dispatch problem. No reply template fixes it, and no number of extra requests outpaces it. This guide covers how to earn and answer reviews honestly. It cannot substitute for the service being good, and it cannot repair a profile whose basics, such as name and category, are wrong. The multi-location profile guide covers those basics.