Canvas Branching for Cart Abandonment with Channel Fatigue
An e-commerce brand has a Canvas for cart abandonment with three steps: push at 1 hour, email at 24 hours, and push at 72 hours. The overall conversion rate is 8%, but the team notices that users who receive all three messages have a 15% unsubscribe rate from push, compared to 5% for those who receive only one. The Canvas currently sends to all users who abandon, regardless of past engagement. The brand has 500,000 monthly active users and wants to increase conversions without increasing unsubscribes. They have access to Braze's Intelligent Channel and can add audience paths.
1. How should the Canvas be restructured to reduce unsubscribes while maintaining conversions?
Read the worked answer
Introduce a decision split after the first push: if the user has opened a push in the last 30 days (engaged), continue with the original sequence but add a frequency cap (max 2 pushes per week). If the user is unengaged (no push opens in 30 days), skip the second push and send only email at 24 hours, then a single push at 72 hours only if they opened the email. Alternatively, use Intelligent Channel to let Braze choose the best channel per user, but that may still over-message. Better: create two paths: high-engagement users get push+email+push with a cap; low-engagement users get email only, then a push only if they click. Also, add an exit event: if user purchases, exit. And add a suppression: if user unsubscribed from push, don't send push. This reduces unsubscribes by not pushing unengaged users. Measure conversion per path and unsubscribe rates. The key is to use engagement data to tailor the sequence.
2. What metrics would you track to evaluate the new Canvas structure, and how would you set up a control group?
Read the worked answer
Track conversion rate (purchase within 7 days), push unsubscribe rate, email unsubscribe rate, and revenue per user. Set up a control group: randomly assign 10% of users to receive the original Canvas (all three messages) and 90% to the new structure. Compare conversion and unsubscribe rates between groups. Also, within the new structure, compare the two paths (high vs low engagement) to see if the tailored approach works. Use Braze's Canvas report to view step-level performance. Ensure the control group is statistically significant: with 500k MAU, 10% is 50k, enough to detect differences. Run for at least 4 weeks to account for variability. If the new structure reduces unsubscribes without hurting conversion, roll out fully. If conversion drops, adjust the timing or add a push for low-engagement users after email click.