Skip to main content
Split traffic between two call treatment paths to measure which experience drives better conversions, before rolling a change out to everyone.
With A/B testing, you can experiment with different touchpoints in your IVR to find the balance that keeps callers happy and conversions high. By testing variations of your call flows in real time, you can see how changes affect caller behavior before fully implementing a new strategy.

​
Benefits of using A/B testing

  • Optimize touchpoints: find the experience that increases conversions or reduces call handling time.
  • Data-driven decisions: use built-in reporting to compare performance, with no extra tools required.
  • Minimize risk: validate a new flow on a small group before rolling it out to everyone.
  • Prove value: demonstrate the ROI of new features, like AI agents or priority routing.

​
Common use cases

Testing a new operational call flow. If you’re rearranging menu options or updating messaging, don’t switch all at once — compare your current flow against the new one and gradually shift more traffic to the winner. Proving the effectiveness of PreSense Priority Routing. If you use PreSense to identify high-value callers, run a test comparing priority routing to your standard IVR to measure actual business impact, like revenue lift, rather than just call center metrics. Testing your AI agent. Split traffic to compare interactions handled by your AI agent against your human workflows, to see how it performs on conversion rates and qualified leads. Before testing, make sure you have:
  1. Two clear paths: know exactly what Path A and Path B will look like.
  2. A success metric: decide how you’ll pick a winner (for example, total sales or appointments booked).
  3. Confirmed Signals: if you’re using a Signal to track success, make sure it’s already set up in your account.

​
Initial setup

Step 1: Open your campaign
  1. In the left navigation menu, click Campaigns.
  2. Select the campaign you want to test.
  3. Scroll to Call Treatment and click Edit.
Step 2: Set up the split
NoteBefore making any changes, take a screenshot of your existing call treatment so you have a record of its original configuration.
  1. Log in to your Invoca account and go to your call treatment settings.
  2. Locate the “At the Start of Call” section.
  3. Select the A/B test option from the dropdown. You can place this node anywhere in your call treatment to test specific segments, not just at the start, and you can use multiple A/B test nodes in a single treatment for more complex testing.
  4. Enter a clear name for your test in the Name fields — one for Path A, one for Path B. This name appears in your reports under the “A-B Test Path” field. Use intentional naming, like “AI_Agent_Test_V1,” to make your data easier to read later.
  5. Set the split by choosing the percentage of callers for each path. A 50/50 split is the standard starting point.
Step 3: Build your paths
  1. Configure the routing and prompts for Path A and Path B.
  2. Click Save. Your test isn’t live yet at this point.
Step 4: Go live
  1. Return to your Campaign Overview page.
  2. Click Activate changes to push your test live to callers.

​
Reviewing your results

Compare your A/B Test Paths against your Success Signals to see which experience is driving the most value. There are two primary ways to view your data: Option 1: The side-by-side dashboard (best for quick comparison). Create a Dashboard Table for a high-level overview:
  1. Add a new tile on your Manage Dashboards page, and select the Builder Tile.
  2. Select your desired visualization type.
  3. Set the Leaderboard Category to the A-B Test Path data field.
  4. In the Ranked By dropdown, add Total Calls or your specific Success Signals.
You’ll see a side-by-side comparison of which path has the higher conversion rate. Option 2: Filtered tables (best for granular insights). If you want to know why a path is winning, create separate tables filtered by a specific path — this lets you see secondary dimensions, like which marketing sources are driving calls to each path:
  1. Create a table and filter where A-B Test Path = Path A.
  2. Set the Dimension to Source or Campaign.
  3. Create a second table and repeat for Path B, to compare how different marketing channels interact with each test experience.
Beyond the dashboard: using reports. Dashboards are great for visual summaries, but you can use the same comparisons in your standard reporting tools for more robust data manipulation:
  • Calls Report: filter to isolate calls that went through a specific path — useful for listening to call recordings and hearing the caller’s reaction to Path A vs. Path B.
  • Custom Summary Report: use Path as a primary or secondary dimension to see a breakdown across your entire account, and export the data for further analysis.

​
FAQs

Can I add more than one A/B test node to my call treatment? Yes. Invoca only shows the last node the caller reached, so name your A/B test nodes to represent the specific part of the IVR the caller reached, especially in more complex call treatments. Fewer variables are usually more effective.

​
Where to go next

Last modified on September 23, 2026