Count the covers you did last week. Now count the Google reviews that came in. For most independent restaurants the second number lands somewhere under two percent of the first, and it says almost nothing about the cooking.
Run the arithmetic on an ordinary week. Two hundred covers, maybe four new reviews. That isn't a struggling restaurant, it's a completely normal one. Leaving a review is volunteer work, and the people who volunteer without being asked skew heavily towards the ones who are annoyed. Everyone else had a nice dinner, paid, and got on with their evening. They were never going to open the app on their own.
So the job isn't earning more goodwill. You probably have plenty of it already. The job is converting goodwill you've earned into something a stranger can read three weeks from now, at nine on a Friday night, while deciding between you and the place two streets over.
Ask after they leave, not while they're paying
The standard advice is to ask at the end of the meal. A card on the table, a line at the bottom of the receipt, a server mentioning it while they run the card. It works, but barely, and the reason is obvious once you watch it happen: the diner is mid-conversation, splitting a bill, finding a coat, thinking about the babysitter.
Ten minutes later they're in the car or on the platform with nothing to do. That is the moment. It's also the moment almost no restaurant ever reaches, because reaching it means having some way to contact a diner who has already walked out the door.
Whatever you use to bridge that gap matters less than the timing itself. A QR code at the table works well as the capture point because the phone is already out. The invite then arrives afterwards, when there's nothing competing for their attention.
At the table
Diner scans and rates in a few seconds. No app, no login.
If something's wrong
You hear about it while they're still in the room, not on Monday.
Shortly after
A short message with a direct link to your listing.
The review
Their words, not yours. Good or bad, it's theirs.
The sequence that works, and the order it has to happen in.
The mistake that gets listings penalised
Here is where a lot of otherwise sensible advice goes wrong, including plenty of it sold as software.
The tempting move is to rate diners first and then only invite the happy ones to Google. It sounds clever. It filters your public profile and it feels like quality control. It is also a direct violation of Google's review policies, which prohibit discouraging negative reviews and selectively soliciting positive ones. The practice has a name in the industry, review gating, and Google has been explicit about it for years.
Selectively asking only satisfied customers for reviews breaches Google's policy, whether you do it by hand or a tool does it for you. Penalties range from reviews being stripped to the Business Profile being suspended, and a suspension takes far longer to undo than a bad review takes to fall off the first page.
So what's the legitimate version? Split the two jobs apart. Capturing feedback privately and inviting people to review publicly are separate things, and only one of them should be conditional.
Private feedback is for you. It exists so that a cold plate on table six becomes something you fix at 8:40pm rather than something you read about on Sunday. Fixing it changes the diner's actual experience, which is allowed to change what they write, because you improved the meal rather than filtered the audience.
The public invite goes to everyone. Someone who had a mediocre night and still wants to say so publicly is entitled to, and a profile with a handful of three-star reviews among the fives reads as considerably more trustworthy than an unbroken wall of perfect scores. Diners are good at spotting the second thing.
Fix the meal, not the sample. One is service. The other is against the rules and reads as fake anyway.
Write like a person, once
The invite itself does more work than people expect. Compare two versions.
The first is a table tent that says "Review us on Google!" with a QR code. Diners have learned to read that as a request made for the restaurant's benefit, and they skip it at roughly the rate you'd guess.
The second arrives on WhatsApp twenty minutes after they've left, mentions the restaurant by name, thanks them for coming in tonight specifically, and contains one link that opens the review box directly rather than dumping them on a search page to hunt for you. That reads as a follow-up from somewhere that noticed they were there.
Two practical notes. Link straight to the review form, not to your listing, because every extra tap costs you a meaningful share of the people who intended to follow through. And send it once. A reminder three days later converts poorly and irritates the diners you least want to irritate.
Don't draft the review for them
Suggesting text for diners to paste in is a shortcut that fails twice. Google's guidelines treat it as manipulation, and it doesn't survive contact with a reader. Reviews written from a template have a particular flatness to them, the same three adjectives, no mention of a specific dish or a specific server, and prospective diners discount them instantly even when they can't articulate why.
The reviews that actually sell tables are the specific ones. "The lamb was worth the wait and Priya sorted out my daughter's allergy without making a thing of it" does more than fifty generic five-star ratings. You cannot write that sentence for someone. You can only make it easy for the person who lived it to write it themselves.
Reply, and reply like you read it
A profile with eighty reviews and no owner replies looks abandoned. Replying signals to prospective diners, and to Google, that somebody is home.
Replies don't need length. They need one specific detail from the review, which is the only thing that distinguishes a real reply from the copy-pasted "Thank you for your feedback! We hope to see you again soon!" that appears under every review of every restaurant in the country. If you're answering a complaint, name the thing that went wrong and say what changed. If you can't say what changed, the reply isn't ready yet.
What a working month looks like
Nothing here is complicated, which is why the restaurants that do it well tend to be the ones that made it automatic rather than the ones with the biggest marketing budget.
- Feedback captured at the table, every service, without a server having to remember to ask.
- Problems surfacing to whoever is on the floor while the table is still occupied.
- One invite per diner afterwards, sent to everyone, linking straight to the review box.
- Replies going out within a day or two, each one referencing something the diner actually said.
Do that for three months and the arithmetic from the top of this piece changes shape. Not because you asked harder, but because you stopped losing the goodwill you'd already earned somewhere between the table and the car park.
RatingEcho automates the sequence for restaurants that don't have anyone free to run it by hand each night: QR feedback capture at the table, private routing so problems reach you immediately, WhatsApp review invites, and reply drafts in the diner's own language that you approve before anything posts. Nobody is ever blocked or filtered out of leaving a public review, for the reasons above.