(888) 815-0802Sign In
revenue - Home page(888) 815-0802

Why You Need to A/B Test Voicemails

4 min readAugust 16, 2018

Leaving voicemails is part of the sales process. It’s like every other essential sales activity that can be measured, analyzed, and improved. Depending on the size of the sales team, reps could be leaving thousands of voicemail messages every week. But thanks to features like voicemail drop, leaving these messages is no longer an arduous, manual process. Reps can record highly contextual, pre-recorded messages and then “drop” them into prospect’s voicemail inbox with a single click. With this function alone, reps can save up to 25 hours a month.

Since reps can leave consistent voicemails at scale for potentially every call, it opens up the door to test messaging. You may have heard of A/B testing, or split testing, for marketing (i.e. subject lines, headlines, advertising creatives, etc.) but this methodology can also be applied to voicemails.

Why You Need to A/B Test Voicemails

Studies show that 97% of business calls go to voicemail. Of those, only 11% of voicemails are ever returned. For example, if a single sales rep makes 50 calls per day, they will leave approximately 48 voicemails. A team of five sales development reps could leave up to 1,200 voicemails per week. With this volume, it’s no surprise that reps should work on optimizing and testing voicemail messages. And with voicemail drop, reps can leave consistent voicemails every time and experiment with components of the voicemail message to set up A/B tests.

Revenue.io Sales Voicemail TestHow to A/B Test Voicemails

If you use voicemail drop, you can A/B test your voicemails to increase your team’s performance. Here’s how:

1.     Larger Sample, Better Results

A general rule of A/B testing is the larger the sample size, the more accurate the results. So, to properly measure voicemail effectiveness, you need as many messages as possible.

In this case, your voicemails are the sample, so you can increase the number of voicemails by employing a large number of reps, running the test over an extended period of time, or some combination of both. To get an idea of how many voicemails you need to generate, you can use a sample size calculator.

Once you have decided how many reps you will test, and for how long, divide your reps into two groups that are as even as possible. Ideally, they should make the same number of calls and leave the same number of messages. The more balanced you can make your test groups, the more reliable your results will be.

2.     Determine The Message Purpose

Next, determine what stage of the sales cycle your message will be used in, and what it’s specific purpose will be. For instance, during the awareness stage with reps doing outbound prospecting, the message could be used for first time cold-calls or email follow-ups. It’s important to specificity how your message will be used so you can control as many of the variables as possible.

3.     Select a Single Factor

The key to a successful A/B test is that only one variable is tested at a time. Select a single aspect of your messages to vary between the two groups. When it comes to the contents of the voicemail, you can test components like:

  • Keyword or competitor mention
  • Feature mention
  • Long vs. short message
  • Casual vs. urgent tone

After you have determined which variant is the most successful, you can run another test to perfect another aspect of the voicemail message.

4.     Create a Script and Record Your Messages

Now that you have a purpose and factor to test, create two scripts, one for each voicemail version. The voicemail script should be easily repeatable for every rep.  Deliver them to the appropriate groups, and have reps record. With something like an Intelligent Dialer, reps are able to pre-record an unlimited number of messages they can use for voicemail drops. When reps record messages, ensure they all save them under the same name, with either “A” or “B” to designate which variation they are using. This will make your reporting much easier.

5.     Run the Test

Have your test groups begin using the messages at the same time, and allow it to run for your chosen period. As the test progresses, monitor your reports to ensure each version is being used evenly.

6.     Continually iterate and improve

Monitor the test results as it runs. If you see one version is performing far better, it may help to pivot to the next variable and begin a new test. If the results are closer, wait until the end before you draw any conclusions.

If you are using a platform like Revenue.io, you can view your team’s voicemail callback report right in your Salesforce dashboard. Otherwise, you can create a custom report based on your reps’ activities and callbacks. Then, select your winner. You can make the usage of the winning phrase, mention, or tone standard, and move to test another variable.