How a Leading Marketplace Drove +84% LTV Increase with AI-Driven Journeys

Customer Challenge

A leading international marketplace with billions in GMV sought to enhance their user experience by better personalizing interactions to drive lifetime value (LTV). While they had identified a set of high-value actions (HVAs) that correlated with LTV growth, they faced several key challenges:

• Suboptimal Decisions: The company used Braze for in-app promotional campaigns but struggled to determine the optimal next best action for each user. A/B testing at the segment level led to increased usage, but that impact did not translate into significant LTV growth.
• Limited Personalization: The team hypothesized that different users had different preferences for different features and needed a more tailored approach to maximize engagement and conversions.
• A/B Testing Constraints: Traditional A/B testing methods made it difficult to add, modify, and remove content at scale.
• Lack of Control Over UI/UX: The company wanted greater control over how the user experience was rendered within their app, which was difficult to achieve using their existing SDK-based setup.

To address these challenges, the company partnered with Auxia to implement a modern AI-driven decision agent that could dynamically personalize each user’s journey.

Solution

The company selected Auxia’s AI platform as the optimal tool to leverage their first-party data and enhance user engagement. Auxia’s decisioning capabilities provided:

• Real-Time Data Consolidation: A no-code integration allowed for seamless processing of events, user attributes, and transaction data into a real-time feature store.
• Automated Machine Learning Models: The platform enabled the company to train and deploy multiple model architectures (e.g., reinforcement learning, causal uplift, contextual bandits) without requiring their internal teams’ direct involvement.
• Content Variation & Personalization: Using Auxia’s UI console, the company created and tested over 200 content variations—20x more than what was possible with their existing Braze campaigns.
• Multi-Objective Optimization: The platform allowed the company to optimize multiple KPIs simultaneously, dynamically adjusting weights to match evolving business needs.

Results

Within four months of implementation, the company drove:

• +84% Uplift in Cross-Category LTV: Auxia’s AI models nearly doubled the effectiveness of determining the right treatments for maximizing LTV.
• 5x Increase in CTR: Personalized distribution significantly boosted click-through rates.
• +4% Increase in Cross-Category Purchases: Dynamic personalization led to previous purchasers making more cross-category transactions, a significant focus area for the team.

Additional benefits included:

• Enhanced UI/UX Control: Greater autonomy over app elements and user experiences.
• Seamless Data Integration: Simplified data-sharing process, moving from integration to launch in under six weeks.
• Advanced Personalization Capabilities: The ability to embed dynamic variables from existing recommendation models (e.g., recommended categories) directly into treatment content.
• Low Latency: The system delivered real-time recommendations in a low-latency environment (<100ms).

By leveraging Auxia’s AI-driven decision engine, the company successfully transitioned from a campaign-centric approach to a user-centric model, driving substantial improvements in user engagement and LTV.