Turn Guest Reviews Into Referrals
A 20-minute check to see if guests can actually recommend you
This is the fastest way to see whether your guests are doing any marketing for you.
If they cannot describe you, you are invisible the moment they leave.
The Question
Can a guest describe you to a friend without opening their phone?
If they need to search, the referral usually does not happen.
“It was really nice” is not a referral. It is a dead end.
Step 1: Sort (10 minutes)
Pull 10 five-star reviews. Make two piles (or two columns if you are using a spreadsheet).
Use whichever platform has the most. If you have fewer than 10, use what you have.
If a review does not contain a noun, it is not Sticky.
If you are debating whether something counts, it does not.
Count each pile.
Step 2: Fix Your Checkout Message (10 minutes)
One place. One change. Your checkout message.
If you found a repeating detail:
Add it to your checkout message. This is not marketing. It is helping guests not miss something worth remembering.
A coastal inn added one line: “Before you go, sit by the fire pit at dusk.” Soon, the fire pit started showing up in new reviews.
If nothing repeated:
Pick one thing guests would not find on their own. Add it to your checkout message.
The beach path behind the grocery store
The bakery with no sign (blue door, before 9am)
The taco truck that only shows up on Thursdays
This can be simple. The best light in the morning. The quietest spot. A shortcut. You know something guests do not.
A cabin owner added: “Walk past the red barn to the overlook before you pack up.” It started showing up in reviews.
Your Result
Sticky count: _____ / 10
Repeating detail: _______________________
Line I added to checkout: _______________________
Most operators start with 0 to 2 Sticky reviews. That’s normal. Now you know what to fix. Give it a few stays. You will see it show up in reviews.
You now know whether guests are spreading word of mouth for you. Or not.
Referrals do not fail because guests do not care. They fail because nothing sticks.
This checks one place referrals break: memory. It is not the only one.
Before booking: Can travelers believe what they see?
After checkout: Are you getting guests back direct?
If you want AI to speed this up, paste your reviews into ChatGPT or Claude with: "Sort these into Sticky (specific detail) vs. Generic (vague praise). Count each." Manual works just as well. Use this only if it saves you time.
The sticky vs generic framework is smart, especially that noun test. I've seen places do the oposite though, where their checkout message over-promises and guests feel misled. The sweet spot seems to be highlighting something real they might miss, not trying to manufaacture a moment.