A genuine review from a past client does more for a real estate business than almost any ad. It's social proof exactly where buyers and sellers look — Google, Zillow, your site — and it compounds: the agent with forty honest reviews looks established in a way the agent with four does not. The hard part was never the value of reviews; it's the discipline of asking every client, every time, in a way that feels personal rather than automated. That's the slow, repetitive work AI is genuinely good at removing.
But review-collection is also one of the few marketing activities with a federal rule written specifically about how you're allowed to do it. The Federal Trade Commission's Rule on the Use of Consumer Reviews and Testimonials took effect in late 2024, and it targets the exact shortcuts that easy, AI-powered outreach makes tempting: suppressing negative reviews, posting fake or AI-generated ones, and buying positive ratings with incentives. The discipline that makes this workflow safe is simple: AI drafts the request; the review stays in the client's own words, every client gets the same honest ask, and you own what you publish.
The three traps in AI review collection
There are three ways this goes wrong, and all three come from letting AI's speed do work that's supposed to carry your judgment and the client's honesty.
The first is review-gating — only sending happy clients to the public review site. Ask AI to "build a review campaign that filters out anyone who might leave a bad review," and it will happily branch your outreach on a satisfaction score: thrilled clients get the Google link, lukewarm ones get a private "tell us how we did" form that never sees daylight. That selective suppression of negative feedback is exactly the kind of distortion the FTC's reviews rule targets. The honest version asks every client the same way, with the same public link, regardless of how you expect them to rate you.
The second is fake or AI-written reviews. Once AI can write anything, the shortcut is obvious: have it draft the testimonial and ask the client to "just approve it," or generate a few glowing reviews to fill out a thin profile. Both are fake reviews — fabricated content presented as a real client's first-hand experience — and the rule prohibits them squarely. AI's only legitimate role here is to make it easy for the client to write their own review, never to manufacture one for them.
The third is incentives and Fair Housing in what you publish. Offering a gift card "for a 5-star review" conditions a reward on a positive rating, which the rule treats as buying a distorted picture; any incentive must be unconditional and disclosed, and most platforms ban incentivized reviews outright. And when you repurpose a review into marketing, AI "polishing" can quietly add steering language — "a lovely young family," "perfect for retirees" — or expose a client's name and personal details they never agreed to publish. Keep incentives off the rating, get consent before you publish, and strip anything about who the client is.
"Build me a review funnel that screens out unhappy clients, write a five-star testimonial each client can just approve, and offer a $25 gift card for every 5-star." In one afternoon you've set up review-gating, fake reviews, and a rating-conditioned incentive — three things the FTC's reviews rule was written to stop — and risked a Fair Housing problem the moment you publish a 'polished' one.
AI drafts a warm, specific request that goes to every client the same way, with the same public link. The review stays in the client's own words. No incentive tied to a rating. Before you republish any testimonial, you get consent and strip any protected-class or personal detail. Steady, honest reviews — fully compliant.
Review collection touches one rule with three sharp edges. No gating: the FTC's Rule on Consumer Reviews and Testimonials (effective October 21, 2024) targets suppressing negative reviews — ask every client the same way, same public link. No fake reviews: the rule prohibits fabricated or AI-generated reviews presented as genuine — the review must be the client's own words, never one you or AI wrote for them. No bought ratings: incentives conditioned on a positive rating are deceptive; any incentive must be unconditional and disclosed, and platforms like Google and Zillow often ban incentives entirely. Plus Fair Housing: never let a published testimonial steer or describe clients by a protected class. The rule for this workflow: AI drafts the ask; the client owns the review; you own what gets published. Your brokerage's policy and each platform's review terms govern.
Where AI actually helps — and where it must not
Drafting the request — AI helps
Turn a closed transaction into a warm, specific note asking the client to share their experience. Drafting a personal ask for every client is the slow part AI removes, so the request actually goes out every time.
Varying the message so it isn't a template — AI helps
Batch out personal-sounding variations keyed to one real detail from each transaction, so fifty requests don't read like the same copy-paste. The ask changes; the public link and the honesty don't.
Optional prompts that help the client write — AI helps
A couple of gentle "what stood out?" prompts can lower the friction of writing, so a busy client actually leaves a review. These help the client find their own words — they never supply the words for them.
Deciding who gets asked and how — you own this
Every client gets the same public ask, regardless of the rating you expect. Routing only happy clients to public sites is review-gating, and that's a judgment call AI must not be allowed to make for you.
The review's words, incentives, and what you publish — you own this
The review stays in the client's voice, any incentive is unconditional and disclosed, and before you republish a testimonial you get consent and strip protected-class or personal details. This is law and judgment, not text generation.
What to capture from the transaction
A great review request is specific, and the specificity comes from the deal, not from the keyboard. Capture a little real detail per client — about the work and the transaction, never about who the client is:
- One genuine moment from the transaction — "the appraisal came in low and we renegotiated," "we closed in time for the school year," "they'd been trying to sell for a year before we listed." This detail makes the AI-drafted ask feel personal and earned.
- The public review link you want them to use — your Google Business Profile, Zillow, or brokerage page. The same link goes to every client; that consistency is what keeps you out of gating territory.
- The timing — the natural moment to ask (at or just after closing, when gratitude is highest) and your CRM's reminder so it actually happens.
- Whether the client has agreed to be quoted — note separately whether they're open to you using their words in marketing later, because a public review and a published testimonial are two different permissions.
- Nothing about protected characteristics — don't record, infer, or hand the AI anything about family makeup, age, race, national origin, religion, disability, or similar. It can't shape the ask and it can't appear in anything you publish.
The review-request workflow — step by step
Pick the moment and the public link
Decide when you ask (closing day or just after works best) and which public review link you'll send. The same link goes to every client — that's the simplest guard against review-gating.
Brief the AI with one real detail and your guardrails
Use the prompts below. Tell it up front: write the request only, not the review; personalize on the one transaction detail; no incentive language; no assumptions about the client; keep it to a single sincere ask with the public link.
Generate one request per client
Let AI produce a specific, warm ask for each person from your note. Fifty personal requests now take minutes, so the ask goes out for every closing instead of only when you remember.
Send the same honest ask to everyone
Send each client the request with the public link — the clients you expect to rave and the ones you're less sure about alike. If you want private feedback too, ask for it in addition to the public link, never instead of it for unhappy clients.
Let the client write their own review
Optional "what stood out?" prompts are fine to help them start; writing the review for them is not. Never send a pre-written testimonial to approve, and never generate reviews to fill a profile.
Before you republish, get consent and run the Fair Housing pass
To feature a review in marketing, confirm the client agreed, then cut anything that names protected characteristics or steers by neighborhood demographics, and remove personal details they didn't agree to publish. Log it in your CRM. Your brokerage's compliance review governs where required.
Prompt templates for review requests
Write a short, warm message asking a past real estate client to leave a review. The transaction: [buyer/seller, one-line context] One genuine moment from the deal: [your note, e.g. "appraisal came in low and we renegotiated to save the deal"] Where to leave the review: [your public review link] Rules: - Write the REQUEST only. Do NOT write the review or suggest what they should say beyond an optional gentle prompt. - Personalize ONLY on the transaction detail above. No assumptions about who they are (Fair Housing). - No incentive, no "5-star," no pressure on the rating. Ask for an honest review of any kind. - One sincere ask with the link. Make it easy to say yes and easy to ignore. Keep it under 110 words, in a warm, plain-spoken voice.
Give me 4 short, optional prompts I can include to help a busy client
write their OWN review if they're not sure where to start.
Rules:
- These are gentle starting questions, not a script and not a draft review.
- Open-ended ("What part of the process was easiest for you?"), never leading toward a rating.
- Nothing that asks them to mention a star count or to praise specific phrases.
- Keep each under 12 words.
I have written permission to feature this client's real review in marketing. Lightly format it for a graphic WITHOUT changing their meaning or words. The review (the client's own words): [paste verbatim] Rules: - Do NOT rewrite, embellish, or invent — fix only obvious typos/spacing. - Flag (don't delete) anything that names or implies a protected class (family status, age, race, religion, national origin, disability) so I can decide. - Flag any personal detail (full name, exact address) that may need removing. - Return the cleaned quote plus a short list of anything you flagged.
"Hi Priya — it was a real pleasure helping you and Sam get to closing, especially after the appraisal came in low and we had to renegotiate to keep the deal together. I'm glad it worked out the way it did."
"If you have a couple of minutes, would you mind sharing your honest experience in a review? It genuinely helps other buyers decide who to trust. Here's the link: [public review link]. No pressure at all, and whatever you write — good or bad — is helpful. Thank you either way, and don't be a stranger."
Tools that work well for review requests
Every review you solicit and publish is your responsibility. Before you send or post anything, confirm you're asking every client the same way with the same public link (no review-gating), that the review is the client's own genuine words (no fake or AI-written reviews), that any incentive is unconditional and disclosed (and allowed by the platform), and that nothing you publish names, implies, or steers by a protected class or exposes personal details without consent (Fair Housing and privacy). AI tools do not understand the FTC's reviews rule, the Fair Housing Act, or each platform's review policies. That judgment is yours, and your brokerage's compliance policy governs.
A note on more reviews, honestly earned
The reason agents under-collect reviews is friction, and AI genuinely removes it — there's no longer an excuse to let a happy client close and never be asked. But the moment asking is free, the tempting next moves are the ones the FTC wrote a rule about: quietly screen out the clients who might not rave, have AI write the testimonials, and sweeten the ask with a gift card for five stars. Each one trades a durable asset — a wall of honest reviews — for a fragile, deceptive one.
Use AI to do what it's good at: turning one real detail from the transaction into a sincere ask in minutes, so the request goes out after every closing. Keep the decisions that carry the risk — who gets asked, whose words the review is in, what you publish — firmly in your own hands. A reputation built on real reviews is the one that actually compounds, and it's your license, not the tool's, on the line.
- Because the reviews that win clients come from asking every time — and this finally makes the personal ask fast enough to send after every closing.
- To turn one real moment from the transaction into a warm, specific request in minutes instead of skipping it.
- Because the real risk in review collection isn't the writing — it's review-gating, fake reviews, and incentives the FTC's rule now targets, and this workflow keeps all three in check.
Frequently asked questions
Can real estate agents use AI to ask clients for reviews?
Yes — for drafting the request. AI is good at turning a closed transaction into a warm, specific message asking a past client to share their experience, and at varying that message so it doesn't read like a template. What it must not do is write the review itself, decide who gets asked, or choose what testimonials you publish. A review you draft for a client to "just approve," or an AI-generated testimonial you post as if it were real, is a fake review — and the FTC's reviews rule specifically targets that. Use AI for the ask; keep the review in the client's own words.
What is review-gating and why is it a problem?
Review-gating means steering only your happy clients to public review sites while routing unhappy ones to a private channel so their feedback never shows up publicly. The FTC's Rule on the Use of Consumer Reviews and Testimonials (which took effect October 21, 2024) targets practices that distort the overall picture of reviews, and selectively suppressing negative reviews is squarely in that zone. AI makes gating trivial — it can branch your outreach on a satisfaction score in one prompt. The fix is to ask every client the same way, with the same public link, regardless of how you expect them to rate you.
Is it OK to offer a gift card for a 5-star review?
Conditioning an incentive on a positive rating is exactly the kind of practice the FTC's reviews rule treats as deceptive, because it buys a distorted picture rather than honest feedback. If you offer any incentive, it has to be unconditional — given for leaving an honest review of any rating — and the incentive should be disclosed. The safest version for most agents is no incentive at all: a sincere, well-timed ask converts well on its own, and it keeps you clearly on the right side of the rule. Your brokerage's policy may be stricter, and platform rules (Google, Zillow) often prohibit incentivized reviews outright.
How can publishing a testimonial create a Fair Housing or privacy problem?
When you turn a client's review into marketing, two things can go wrong. First, repurposing or "polishing" it with AI can surface protected-class signals — describing the client as a "lovely young family" or "perfect for a retired couple," or implying who a neighborhood is "right for" — which is the kind of steering language the Fair Housing Act prohibits in advertising. Second, a review may include the client's full name, the property, or personal details they didn't agree to see in an ad. Get the client's permission to use their words publicly, keep the testimonial about your service rather than about who they are, and strip any protected-class or sensitive personal detail before you publish.
Can AI write a testimonial for a client to approve?
No — that crosses the line into a fake review. A testimonial is supposed to be the client's genuine, first-hand experience in their own words; an AI-written review the client merely signs off on is fabricated content presented as authentic, which is precisely what the FTC's rule prohibits. AI's legitimate role is to make it easy for the client to leave their own review — a clear ask, a direct link, maybe a couple of optional prompts about what stood out — never to manufacture the review for them. If a client is busy, help them by lowering the friction, not by writing their words.
Get new AI workflow guides for real estate agents
We publish new workflow guides for real estate professionals. No spam, one email when something useful goes live.