Review gating is asking only your happy customers to post a public review while quietly routing unhappy ones somewhere else. Google prohibits it outright, and as of its April 2026 policy update the ban is enforced automatically rather than complaint-driven. If any part of your review-collection process treats a customer differently based on how satisfied they seem before they've written anything, that's the practice this post is about.
What counts as gating
The textbook version is a pre-screen survey: "How was your visit?" with a star tap, where four and five stars get a link to Google and anything lower gets routed to an internal form or nowhere at all. The customer never sees the option to post publicly unless the business has already decided they're likely to say something nice.
It doesn't have to be a formal survey to count. Staff who hand out review cards only to customers who seemed pleased are gating by hand. A checkout flow that only shows the Google link after a positive NPS answer is gating in software. The common thread isn't the tool. It's that the public review option depends on a sentiment check the customer didn't agree to.
What changed in 2026
Google made two separate policy moves in mid-April 2026. One built out enforcement: automated, AI-assisted detection aimed at scam and manipulation patterns, plus pre-publication screening on some review types. The other tightened what businesses themselves are allowed to do: it names review gating and incentivized reviews as violations, alongside newer bans like asking customers to name specific staff members in a review or collecting reviews on a shared, on-premises device.
The practical shift is that this stopped being a rule businesses could quietly ignore. Reporting since the update describes retroactive sweeps of existing reviews, public warning labels appearing on flagged profiles, and a later enforcement phase aimed at ranking penalties for repeat violators. None of that is a reason to panic if you're not gating, but it's a reason to actually check your process rather than assume "we've always done it this way" is fine.
What's still completely allowed
The policy doesn't ban asking for reviews, automating the ask, or using a QR code, email, or SMS to get there faster. It doesn't ban a rating step before the review either. What it requires is that every customer gets the same door: the same prompt, the same platform options, regardless of what they're about to say.
That's the distinction between gating and routing. A compliant flow asks everyone the same question, then lets the customer, not the business, decide where their answer goes. A five-star tap can open Google. A two-star tap can open a private message straight to the owner. Both are the customer's own choice, made after the same prompt, not a filter applied on their behalf. Our QR code review card piece covers the mechanics of a well-built version of this flow in detail.
Checking your own process
Three questions cover most of the risk:
- Does every customer see the same first screen? If the answer depends on a rating, a mood read from staff, or a "was everything okay?" gate before the link appears, that's the part to fix.
- Who decides where the feedback goes, you or them? The moment a business, not the customer, decides a low rating shouldn't reach Google, that's gating even if the intention is damage control rather than deception.
- Are you asking people to name staff or leave reviews on your own device? Both are separately banned now, independent of gating.
None of this means you have to accept every complaint landing on your public profile with no recourse. Routing a two-star customer to a private inbox is fine, as we cover in what to do with a bad rating, as long as the same choice was offered to the five-star customer standing next to them. The line isn't "can unhappy customers go private," it's "did every customer get offered the same options first."
Fixing a gated process without losing the upside
If you're running a filtered survey today, the fix is usually smaller than it sounds. Keep the rating step, it's useful and not the problem. Change what happens after it: instead of the business choosing the destination based on the score, show the customer both options and let them pick. Most genuinely happy customers still choose to post publicly when asked directly. You lose a small amount of curation and gain a process that survives an audit.
Starboard's rating flow is built around this exact shape: one prompt, then the customer's own choice of where it goes, with every rating reaching the owner's dashboard either way. It's less about picking a side and more about not deciding for the customer in the first place.
Whatever tool you use, the test is simple: if you can't show a regulator or a skeptical reviewer that every customer saw the same first screen, you're gating, whether or not that was ever the intent.