Multivariate testing lets you test multiple offer variables at once—like discounts, timers, and product bundles—to find the best-performing combination. It’s one of the most powerful ways to optimize your post-purchase upsell strategy.
Step-by-Step Instructions:
1. Define Your Goal
Before setting up your multivariate test, take a moment to define what success looks like for your store. Setting a clear objective helps you choose the right variables to test and measure what truly matters.
Some popular goals include:
Increasing conversion rates on upsell offers
Boosting average order value (AOV)
Identifying the most effective product bundle or incentive
💡 Need inspiration? Check out our Best Practices for Upselling article for ideas that drive real results.
2. Choose the Variables to Test
Select 1–5 high-impact variables that could affect buyer behavior. For example:
Discount Type: Percentage vs. fixed dollar
Urgency Timer: 5 min vs. 15 min
Product Type: Single item vs. bundle
Shipping: Free vs. flat-rate
✨ Tip: Too many variables = harder analysis. Keep it focused!
3. Create Your Variations
Set up 0-5 options for each variable:
Discounts: 10%, 20%, 30%
Timers: 5, 10, or 15 minutes
Bundles: 1 product vs. 2-product combo
AfterSell automatically combines your variations into multiple offer paths—no manual setup required.
4. Set Up Your Multivariate Test
Open the AfterSell Dashboard
Go to Funnels
Create a new funnel or open an existing one
Create test
Select your desired variations
Start the test
Notes:
Each variation is automatically labeled as Page A, Page B, and so on for easy reference in your analytics.
Be sure to press Enter after typing in each variable to save it properly:
5. Review the Results
Once your test has been running for a sufficient amount of time—ideally 2 to 4 weeks—you’ll have enough data to make informed decisions. Shorter test durations may not yield statistically reliable insights, especially if traffic is low.
Here’s how to analyze the outcome:
Go to the Analytics tab in your AfterSell dashboard and select 'Split Tests' to view details about your experiments.
Compare key metrics like:
Conversion rates
Average order value (AOV)
Revenue per visitor
Identify the top-performing variation set based on which combination delivers the strongest results
📈 Your test should list the impressions needed to be statistically viable—this helps you know when the data is ready to act on. See screenshot below for reference:
💡 Tip: Look for consistent performance across multiple metrics—not just a spike in one area.
6. Apply the Winning Variation
Once a winner emerges:
Apply that variation across your funnel
Monitor performance to ensure it continues delivering strong results
Consider testing again with a new set of variables
Best Practices
✅ Start Simple: Don’t overload with too many variables
⏳ Be Patient: Let tests run long enough to collect data
🔁 Keep Testing: Use winners to inspire your next round of tests
Need More Help?
Chat with our support team using the live chat at the bottom right of the app.