Overview
You can make data-driven marketing decisions by A/B testing your SMS messages in flows. You can A/B test your SMS messages in a few ways. You can look at the timing, the content, the number of messages, and whether an email or SMS converts better and choose the cadence that is the most engaging to your customers.
Because Klaviyo does not pick a winner for flows, you'll need to monitor performance over a set amount of time to gather the data you need to determine a clear winner. The amount of time that you run your test will depend on your sending frequency. For example, if you send SMS messages twice a month maximum, you will need to run the A/B test longer than if you send six times a month.
Branching for SMS
Before you build out the below tests, you’ll need to either add in flow filters if you want only SMS customers in your flow or split your flow so that your SMS subscribers go down a different path than customers who have not opted-in to your SMS messages.
Adding a Flow Filter
Add a flow filter to create an SMS-only flow, and add a flow filter that will only allow customers who have consented to receive SMS to enter the flow.
Adding a Conditional Split
To split off customers who have consented to receive SMS from those who haven’t, drag in a conditional split and add the condition that someone has consented to receive SMS at least once. You would use this option if you want email-only profiles in the same flow as your SMS A/B test.
Random Sample Conditional Splits for Testing
To build out any of the following ideas, you should create a random sample conditional split in your flow where you want to test your SMS messages. To start, drag in a conditional split and add in a random sample split of greater than 10% of your customers going through the flow.
Timing of Messages
Because it’s important to contact customers at a time when they are most likely to convert, you can play with the timing of your flow sends. In the case of an abandoned cart flow, for instance, you can build out a split where you test a one-hour and a four-hour delay (or any time iteration that makes sense for your business).
Content of Messages
Your SMS message content may also impact your conversion rates. Test out using personalization, a discount, or the first word of your SMS, and see if it leads to a lift in your click rates.
Number of Messages
You can test to find the number of SMS messages that receive optimal engagement. Try adding an additional message to your current SMS sending strategy and see how it performs.
SMS vs. Email Message
Lastly, you can see if your SMS or email performs better at a certain point within the customer cycle. This could be especially useful in your welcome series and abandoned cart flows. Split your flow with a random sample and see if sending an email or SMS leads to higher revenue.
