Review Request Automation: How to Ask Every Customer, Legally
From the Lattice Management Journal: practical guidance on business systems, CRM and marketing technology.
Published October 19, 2026 / Updated October 19, 2026 / Lattice Management

Review request automation means your systems ask each customer for a Google review at the right moment, with no one having to remember. A real event triggers the ask, such as a completed job, a paid invoice or a finished appointment. The message goes out by text or email, and the customer lands directly on your review form. Done well, it gives you a steady flow of reviews. Done badly, it creates legal exposure and a pile of unanswered feedback.
This guide is for owners of small businesses and professional practices: contractors, clinics, consulting firms and similar teams. It covers how to build the workflow, which rules you cannot bend, and where automation should stop and a person should take over.
TL;DR
- Trigger the request from a real event in your CRM (job complete, invoice paid, appointment attended), not from a weekly batch.
- Ask every customer. Selecting only the happy ones is review gating, and it carries legal risk under the FTC’s 2024 rule.
- Never offer discounts or gifts for reviews. Both the FTC and Google prohibit it.
- Automate the asking, but route complaints, billing disputes and safety issues to a person before anything is posted publicly.
- Build the reply process first. A review volume you cannot answer works against you.
Table of Contents
- Why reviews are worth a system
- What review automation covers (and what it does not)
- The rules you cannot automate around
- How to build the workflow
- What to automate and what to hold
- Perspective: the review is an output of operations
- Measuring whether it works
- Beyond Google
- Common mistakes we see
- Where LATTICE fits
- FAQ
- Sources
Why reviews are worth a system
Reviews are where buyers check you before they call. In the 2025 BrightLocal Local Consumer Review Survey, 83% of consumers used Google to research local business reviews, and 74% consulted at least two review platforms before buying. Google is the first stop, but it is rarely the only one.
Most businesses do not lack happy customers. They lack a consistent moment of asking. The owner is busy, the technician forgets, the front desk is on the phone. A manual process produces reviews in bursts, usually after someone remembers on a Friday afternoon. An automated one produces them at the pace of your actual work.
That consistency matters for a second reason. A steady trickle of recent reviews looks more credible than forty from one month and none since. Automation gives you the trickle.
What review automation covers (and what it does not)
The phrase gets used for three separate jobs, and most tools handle only one or two:
- Asking: sending requests automatically after a qualifying event.
- Answering: drafting or sending replies to reviews that arrive.
- Watching: alerts, reporting and monitoring across platforms.
Only the first is what you searched for, but the other two decide whether the first pays off. Practitioners writing on Google review automation make the same point we see in client work: build the answering process before you scale the asking, because a growing backlog of unanswered reviews undermines the reputation you are trying to build.
The rules you cannot automate around
Before any workflow design, settle the compliance boundaries. They are short, and they shape every decision below.
No review gating
Review gating means screening customers first (“How was your experience?”) and only sending the happy ones to Google. It is a common feature in older tools, and it is now a liability. The FTC’s final rule on fake reviews, announced in August 2024, targets review suppression and lets the agency seek civil penalties against knowing violators. Legal commentary on the rule, such as this analysis from a law firm’s consumer products blog, lists review suppression and misrepresenting displayed reviews as covered practices.
The practical rule: send the same request to every customer, with the same link, regardless of how you think they feel. If you want private feedback as well, offer it as an addition, never as a filter in front of the public review.

No incentives
Do not offer discounts, free services or entries into a draw in exchange for reviews. Google’s Maps user-generated content policy says reviews must reflect a genuine experience, and incentivized contributions fall under its fake and misleading content rules. The FTC rule also addresses reviews that are bought or rewarded in ways that distort them. Thank customers in your message, but do not pay them.
Keep the ask honest
Google’s policies also change. If your workflow includes scripted language for staff, such as asking customers to mention a particular name or to review while still on your premises, check the current policy page before you launch. Review it at least once a year, because platform rules move faster than most business processes.
This is a compliance area. If your business is regulated (healthcare, financial services, legal), confirm with your own counsel or compliance officer what you may say in a review request and how you may respond to reviews from patients or clients, since confidentiality rules apply to replies as well.
How to build the workflow
The design below works whether your CRM is a purpose-built system or a set of connected tools. What matters is that the trigger comes from a record of something that really happened.
Step 1: Pick the trigger event
The best trigger is the moment a customer has just received value and the record proves it:
- Job marked complete (contractors, home services)
- Appointment marked attended (clinics, salons, consultants)
- Invoice marked paid (professional services)
- Project milestone signed off (agencies, consulting firms)
Event-triggered requests generally outperform batch sends, because the customer still remembers the work. A request sent three weeks later reads like a form letter. This only works if your operational system is the source of truth, which is why review automation fails in businesses where job status lives in someone’s head or on a whiteboard.
Step 2: Choose the delay
Hours, not weeks. For a repair or visit, a few hours to a day is usual. For longer engagements, such as a consulting project, wait until after delivery and the final conversation. Test it. Start with one delay, measure the response, then adjust.
Step 3: Choose the channel
Text messages are read faster than email for most local services, but email suits professional services where the relationship already runs through an inbox. Make sure you have consent to message customers by text and honor opt-outs. If you are unsure of your obligations, check with an attorney familiar with US messaging rules.
Step 4: Write the message like a person

Keep it short, name the person or team who did the work, and give one link. Do not hide the link behind three clicks. Consumers have become sharp at spotting inauthentic content: a Chatmeter survey found that 53% mistrust reviews that look AI-generated. Use automation to deliver the message, but write it yourself, and never draft reviews on the customer’s behalf.
A workable template:
Hi Dana, thanks for choosing us for your kitchen remodel. If you have two minutes, a Google review helps other homeowners find honest contractors. Here is the link: [link]. If anything fell short, reply to this message and I will personally follow up.
That last line is not gating. The review link is visible to everyone, and the offer of direct contact simply gives unhappy customers a faster route to you.
Step 5: Use your direct review link
Generate the link from your Google Business Profile so customers go straight to the review box. This depends on a complete, verified profile, which we cover in our Google Business Profile setup and management work.
Step 6: Send one reminder, then stop
One gentle follow-up after several days is reasonable. Three is harassment. Suppress anyone who has already reviewed, who has opted out, or who has an open complaint on file.
Step 7: Route the replies
Every new review should create a task or alert. Positive ones can get a prompt, personal reply from the owner or office manager. Negative or complex ones need a human decision first.
What to automate and what to hold
A useful rule from practitioners is to automate the request for every customer, then hold sensitive responses for a person. The table shows how we split it.
| Task | Automate? | Why |
|---|---|---|
| Review request after a real event | Yes | Consistency across every customer |
| One reminder to non-responders | Yes | Low effort, low risk |
| Alert when a review arrives | Yes | Speed of response |
| Draft for a positive reply | Assist | A person edits and posts |
| Reply to a negative review | No | Needs judgment and context |
| Anything involving billing, safety, legal threats or health details | No | Risk of escalation or privacy breach |
| Filtering who gets asked | Never | Review gating |
AI drafting can save time on routine thank-you replies. It should not post unsupervised. A generic reply copied under thirty reviews tells readers nobody is paying attention.
Perspective: the review is an output of operations
We see the same pattern in nearly every small business we work with. The owner buys a review tool, connects it to nothing, and wonders why requests go to the wrong people at the wrong time. The tool was never the problem. The problem was that the business had no reliable record of when work was finished.
At LATTICE, we treat review requests as one of the touchpoints in a connected system, alongside lead capture, follow-up, booking and payment. When the CRM knows a job is complete, the request goes out. When a review comes in, someone is notified. When a complaint is open, the request is suppressed. None of that is clever. It is connective tissue, and it only works when the underlying workflow is clean.
Our view is also that the human voice is the asset. Automation should remove the forgetting, not the personality. We would rather a client send a plain, slightly imperfect message from a real person than a polished one that reads like a template. Customers can tell, and the data on mistrust of AI-sounding content backs that up.
If you are choosing where to begin, start with the CRM and the event that marks a job as done. The review automation comes almost for free after that.
, Lorenzo
Measuring whether it works
Track a few numbers monthly, and resist the urge to track everything:
- Requests sent versus reviews received: your conversion rate on the ask.
- Average rating and its trend: watch for drift, not a single month.
- Time to first reply: how quickly your team responds after a review is posted.
- Reply rate: the percentage of reviews that get an answer. Aim for all of them.
- Source of new customers: ask new leads how they found you and record it in the CRM.
Your average rating may dip slightly when you start asking everyone, because quiet, mildly satisfied customers now speak up alongside the delighted ones. That is what an honest sample looks like, and it is also what makes a rating believable.
Beyond Google
Google is the priority, but BrightLocal’s finding that 74% of consumers check at least two platforms is a reminder to keep other profiles current. Trades may care about the BBB, hospitality about Tripadvisor, healthcare about specialty directories. You do not need an automated campaign on each. Keep your business information consistent, and send customers to the platform that matters most for your category. Our business listings and citation management service handles the consistency side, and reputation management covers how you handle what people say once it is posted.
Common mistakes we see
- Batch sending from a spreadsheet. The request arrives weeks after the job, and response rates fall.
- Asking only satisfied customers. This is gating, whatever the tool calls it.
- Reward programs. A coupon for a review looks generous and breaks the rules.
- No reply process. Reviews pile up unanswered.
- Wrong link or an unverified profile. Customers cannot find where to post, so they give up.
- Automation without data hygiene. Duplicate contacts mean one customer gets three requests.
- Ignoring complaints in the queue. A customer in a dispute should never receive a cheerful review ask.
Where LATTICE fits
If your jobs, appointments and invoices already live in one system, review requests are a small addition. If they do not, the better investment is fixing the system first. That is what we do: design and build the CRM, booking, automation and website layer, then support it after launch so it keeps pace as the business changes.
Our review automation for small business service connects requests to the events in your own workflow, and our CRM systems give those events a reliable home. Not sure where your gaps are? Book our free systems audit and we will map where follow-ups, bookings and reviews are leaking today. You can also see real work and testimonials before you decide.
FAQ
How can I automate Google reviews?
Connect your CRM or scheduling system so that a completed job, attended appointment or paid invoice triggers a text or email containing your direct Google review link. Send it to every customer, add at most one reminder, and set an alert for each new review so a person can reply. Replies to sensitive reviews should stay manual.
Is review gating illegal?
The FTC’s 2024 final rule bans review suppression and allows civil penalties against knowing violators, and selectively asking only happy customers falls into the practices regulators are concerned about. Google’s policies are also against it. Because the details matter, ask a business attorney to review your process if you operate in a regulated field or are unsure.
Can I offer a discount in return for a review?
No. Google’s policies require reviews to reflect a genuine experience, and rewarding reviews is treated as fake engagement that can lead to profile restrictions. The FTC rule also covers purchased or rewarded reviews. Thank customers sincerely and leave it there.
Is it better to leave reviews on Google or Yelp?
For most local businesses, Google is the first priority, since 83% of consumers in BrightLocal’s 2025 survey use it to research local reviews. Yelp or niche platforms can matter in particular industries or cities. Ask where your own customers look, and keep those profiles accurate.
Should I let AI write my review replies?
Use it for first drafts of simple thank-you replies, then edit before posting. Do not let it respond unsupervised to complaints, billing questions or anything involving health, safety or legal matters. Those need a person who knows the customer.
Sources
- Local Consumer Review Survey 2025, BrightLocal
- FTC announces final rule banning fake reviews and testimonials
- Final Rule Announced: The FTC Strengthens Its Enforcement Capacity Against Deceptive Reviews and Testimonials
- Maps User Generated Content Policy, Google
- Chatmeter 2024 survey findings
- Google Review Automation: The Complete Guide (2026)