--- title: "SMS A/B testing: better review request conversion | Repto" description: "Repto tests the SMS text and send time on your own clinic's data, and the better version stays for good. Test after test, more of your review requests convert." url: https://repto.pl/en/features/ab-testing updated: 2026-11-10 language: en --- # A/B tests: Google review requests that keep getting better Repto tests the SMS text and send time on your own clinic's data, and the version that **brings in more Google reviews** stays for good. Repto checks every difference before anything changes, so test after test you improve the conversion of your review requests based on data, not gut feeling. What A/B tests give you At a glance - More Google reviews from the same number of visits. - SMS text and send time based on your own clinic's data. - Repto checks whether the difference is more than chance. - The better version stays for good, and the next test can improve it further. - The dashboard shows each version's results all the way to the Google review. Why ## How do A/B tests improve conversion? Each test compares two versions of the SMS: different text or a different send time. The better version stays for good and becomes the starting point for the next test. That way, test after test, **you get more Google reviews from the same number of visits**. What works at another clinic may not work at yours, and a “month before vs. month after” comparison misleads, because the number of visits, the season and the doctors' schedules change in the meantime. In a test, both versions go out on the same days to randomly chosen patients, so the only thing that affects the result is how the versions differ. How it works ## How does an A/B test run in Repto? 1. 1 ### You set up two versions Version A is usually your current SMS, and in version B you change the text or the send time. You can set up the test yourself in the dashboard, under A/B tests, or we can run it for you. 2. 2 ### Repto splits patients at random Each time messages go out, half the patients get version A and half get version B. Each patient gets only one of the versions. 3. 3 ### The dashboard counts the whole funnel For both versions Repto counts the same [funnel from SMS to review](https://repto.pl/en/features/review-analytics) as in the rest of the dashboard: SMS sent, links opened, visit ratings and the Google reviews that the [matching algorithm](https://repto.pl/en/features/review-matching) assigned to a patient. 4. 4 ### Repto checks whether the difference is more than chance At every step you see how much better or worse version B does, and how certain that result is. 5. 5 ### The better version stays for good With one click you move the winning version's text and send time into your settings. From then on every patient gets that SMS, and the next test starts from the better version. Statistics ## How does Repto tell whether a difference is more than chance? Over a few dozen messages, one version almost always does a little better, even when the two are identical. That's why, besides the percentages, Repto calculates **confidence: the probability that the difference is real**. The more messages sent, the more certain the result. When a test has been running for a long time and the difference is still too small to confirm, the dashboard shows that this change has little effect. That's the time to test a bigger one. Patients in the test ## What does a test change for patients? - ### Each patient gets one version Never both. The [SMS limit per patient](https://repto.pl/en/features/message-frequency) works the same during a test as without one. - ### The safety net works in both versions The test changes only the SMS. The [visit rating](https://repto.pl/en/features/patient-feedback) works as always: an unhappy patient describes the problem to the clinic first, and the manager gets an email and can get in touch with them. - ### A separate test at each clinic Each clinic has its own tests and results, so a group of clinics can try the same change at several locations. - ### You end the test whenever you want You switch it off with one click and every patient goes back to the SMS from your settings. FAQ ## Questions about A/B tests ### How do you raise the conversion rate of Google review request texts? + Test the text and send time on your own clinic's data and keep the version that brings in more reviews. In Repto you do this with A/B tests: half the patients get one version of the SMS, half get the other, and the better one stays for good. Test after test, you get more Google reviews from the same number of visits. ### How do you find out which review request text works better? + With an A/B test: over the same period, half the patients get one text and half get the other, and you compare how many Google reviews each one brought in. In Repto you set up both versions in the dashboard, and Repto does the rest by itself: it splits the patients, counts the results and checks whether the difference is more than chance. ### How long does an SMS A/B test take? + As long as it takes for the result to be certain. A test has no fixed end date and runs until you end it. A clinic with many visits gets an answer sooner, a smaller one needs more time. The dashboard shows at all times whether the result is decided yet. ### Who sets up the test: the clinic or Repto? + Either. The clinic account's admin sets up the test in the dashboard, under A/B tests, and the rest of the staff see the results. We can also run the test on your data for you: we set up both versions and keep an eye on the result. ### How do you know a difference in an A/B test is more than chance? + Repto calculates confidence, the probability that the difference between the versions is real. The dashboard names a winner only once the result is certain, and until then it says that it's too early to tell. ### Will a patient in the test get both versions of the SMS? + No. Each patient gets one of the two versions, and the SMS limit per patient works during the test the same as always. ## Other Repto features [Automatic SMS after the visit](https://repto.pl/en/features/review-request-sms) [Safety net for unhappy patients](https://repto.pl/en/features/patient-feedback) [Google review matching](https://repto.pl/en/features/review-matching) [SMS-to-review analytics](https://repto.pl/en/features/review-analytics) [QR codes for staff](https://repto.pl/en/features/qr-codes) [SMS limit per patient](https://repto.pl/en/features/message-frequency) [All Repto features](https://repto.pl/en/features) ## 15 minutes to see if it's worth it It's a short online call, in English. I'll look at your Google reviews and show you the dashboard. Then I'll work out the cost and a forecast: how many reviews you could collect from your visit volume at a typical conversion rate. ![Piotr Czupryński](https://repto.pl/main_page/me.jpg) Piotr Czupryński Founder of Repto or [Email kontakt@repto.pl](mailto:kontakt@repto.pl)