Challenge
A large online retailer with its own mobile app and a web-based customer account faced the same problem as many e-commerce businesses: its most important messages went out by email, only to get lost in crowded inboxes. Push notifications helped to a point, but they were often seen as intrusive, and a growing number of users were opting out. As a result, abandoned-cart reminders, order-status updates and personalized offers reached customers when they weren't ready to buy — or didn't reach them at all. The retailer was looking for a channel that would let it talk to customers exactly where they make their decisions — inside its own app — without annoying them and without depending on email open rates.
Solution
- A dedicated inbox built into the app The retailer launched the In-Box channel and connected both its web and mobile apps to it. Messages looked like a native part of the product rather than a bolted-on ad. The entire visual layer stayed on the retailer's side, so the notifications fit seamlessly into the interface.
- Reaching only the right recipients Only logged-in customers were registered to the inbox — automatically, based on segments and via API during database synchronization. This meant every send reached an opted-in audience, and every registration and removal was recorded in the contact's activity history, giving the retailer full control over who receives what.
- Messages that drive purchases Every message was personalized with profile data. The CTA with a tracked link took web users straight to checkout, while on mobile a deep link opened the relevant screen. Abandoned-cart reminders were built into marketing automation scenarios and the abandoned-process recovery mechanism, so they sent themselves at the right moment in the customer journey.
- Unobtrusive delivery and measurement The pull-based model meant the message waited in the inbox and appeared in context, when the customer returned to the app — without the intrusiveness of push. Each message had an expiry date, so the inbox never filled up with outdated offers. A control group made it possible to measure the channel's real impact on sales, while delivery, read and click reports showed which messages worked and allowed them to be optimized quickly.
Results
- In-Box messages were read 41% more often than their email counterparts — because they appeared exactly where the customer was already engaged in shopping.
- Abandoned-cart reminders, built into scenarios and abandoned-process recovery, recovered 26% of interrupted transactions.
- Personalized codes and offers presented in the inbox increased add-on sales by 18%, and the comparison with the control group confirmed that the uplift came from the channel, not from seasonality.
