Agencies should avoid recommending review tactics that make the agency look like it is helping the client control review outcomes.
That includes tactics that control who reviews, what they say, what rating they leave, whether they revise a review, or whether negative feedback is kept away from Amazon.
The client may be asking for growth. The agency still needs to protect the account and the relationship.
Incentives tied to review behavior
The clearest risk is any benefit tied to review behavior.
That can include money, refunds, rebates, discounts, gifts, future access, entries, credits, or other benefits tied to whether someone leaves a review, what rating they leave, what they say, or whether they change or remove a review.
The FTC Consumer Reviews and Testimonials Rule Q&A says businesses cannot pay incentives for 5-star reviews on third-party review platforms, even if reviewers disclose the incentive.
Amazon’s Community Guidelines also draw an important boundary: products may be provided for free or at a discount and those customers may write reviews, but benefits cannot be conditioned on review behavior or review content.
Sentiment filtering
Agencies should avoid flows that route happy customers toward Amazon and unhappy customers somewhere else.
Customer support is legitimate. Product feedback is useful. The risk appears when the customer path changes based on sentiment in a way that encourages positive public reviews and suppresses negative ones.
That is not a review strategy an agency should normalize.
Review pods, swaps, or controlled reviewer groups
Agencies should avoid review pods, review swaps, friends-and-family review plans, employee reviews, controlled accounts, or groups where participants are expected to review each other’s products.
Amazon’s Anti-Manipulation Policy for Customer Reviews says any attempt to manipulate reviews, including false, misleading, or inauthentic content, is strictly prohibited.
For an agency, the practical rule is simple: if the reviewer relationship would be hard to explain to the client, legal, Amazon, or a future buyer of the brand, do not recommend it.
Positive-review asks
Agencies should not recommend language that asks for positive reviews, five-star reviews, favorable feedback, or reviews only from customers who had a good experience.
Amazon’s Customer Reviews tool page says sellers should not attempt to influence customer ratings, feedback, or reviews, and should not ask customers to remove negative reviews or post positive reviews.
That gives agencies a clean copy standard: ask for honest feedback, not positive feedback.
Rating or sentiment promises
Agencies should avoid promising ratings, sentiment, review language, or review outcomes.
The agency can forecast operational work. It can plan client education, review readiness, approved messaging, documentation, and reporting.
It should not present positive review outcomes as something the agency controls.
Unclear third-party workflows
Another risk is recommending a vendor without understanding how the customer-facing work happens.
The client may only see a clean dashboard or a simple pitch. The agency should still understand whether customers are contacted off-platform, whether review language is suggested, whether unhappy customers are treated differently, whether future benefits depend on review behavior, and whether subcontractors are involved.
If the workflow cannot be explained, it should not be recommended.
The practical takeaway
Agencies do not need to avoid review conversations. They need to avoid risky review recommendations.
The safer standard is clear: real customers, voluntary feedback, clean benefit boundaries, approved customer-facing language, no sentiment filtering, no review pods, no positive-review asks, and no promises about ratings or content.
That is the answer an agency can stand behind after the client call is over.