Axlebolt leverages AI to boost ad spend efficiency and user acquisition with Adjust

Axlebolt aimed to optimize the allocation and management of its advertising budget, using a data-driven approach to gain traction, expand its user base, and stay competitive in the mobile gaming market. However, they faced challenges regarding the timeliness and accuracy of data. While Axlebolt relied on real-time data analytics to strategize and fine-tune their advertising campaigns, delays in receiving raw data from certain advertising partners hindered their ability to make quick, informed decisions.

To monitor and tailor their campaigns, Axlebolt employs a complex approach that involves more than 60 metrics in their business intelligence (BI) system. These metrics play a crucial role in assessing campaign performance through Axlebolt's user acquisition (UA) dashboards, helping them evaluate the effectiveness of their advertising spend. However, managing and interpreting such a vast amount of data, especially to maintain a focus on cohorted return on ad spend (ROAS), added another layer of complexity to their strategy.

The introduction of SKAdNetwork (SKAN) presented additional challenges. The privacy-centric nature of SKAN led to a loss of insights into performance and user value, making it difficult for Axlebolt to accurately measure and understand the profitability and quality of their iOS users. Moreover, their low App Tracking Transparency (ATT) conversion rate hindered efforts to scale their iOS user base and revenue. With a budget predominantly focused on Android, Axlebolt's overall revenue growth was adversely affected.

To address these issues and achieve their return on investment (ROI) and lifetime (LTV) goals, Axlebolt aimed to develop a predictive revenue model for a more accurate understanding of product profitability. This is why Axlebolt sought the support of a mobile attribution partner like Adjust.

Adjust's SpendWorks ad spend solution provided Axlebolt with a comprehensive platform to consolidate all their advertising spend data from various sources. This unified system offered the level of detail Axlebolt needed to optimize campaign performance and improve their ROAS.

Initially, Axlebolt collected some partner spend data through Adjust but faced challenges with partners who were not integrated with Adjust. As they expanded and partnered with more networks, they encountered the time-consuming and costly task of manually obtaining spend data, which had different formats and required extensive extract, transform, and load (ETL) processes.

At first, Axlebolt used SpendWorks to track only partners that required direct interaction. However, as they aimed to ensure consistency between the data received in their BI system and the data obtained from Adjust, they expanded its use. This decision was partly driven by delayed data delivery from some partners, which Axlebolt hoped to address through SpendWorks to minimize discrepancies in data transfer timing.

To tackle the challenges posed by SKAN, Axlebolt adopted Adjust’s new AI-based Predictions solution. The Adjust team provided a thorough explanation of the predictive model's process and structure and conducted a two-stage training of the models before delivering the final product. The predictions’ accuracy was further enhanced after fine-tuning the SKAN conversion value schema to better distribute predictions across various buckets.

Axlebolt leveraged Adjust’s machine learning (ML) model to predict user value and behavior 14 days after app install. By analyzing user signals from the first 24 hours post-install, the ML model has a prediction accuracy of over 90%. Predicting day 14 revenue within the first 24 hours allows Axlebolt to assess the quality of advertising channels and increase purchase volumes on the iOS platform through channels like Google Ads and TikTok.

By utilizing SpendWorks and the Predictions solution, Axlebolt achieved the following:

• Optimized their advertising budget using data-driven insights, resulting in a 34% higher ROI by focusing on the most profitable channels.

• Gained insights into the most engaging channels, achieving a 9% increase in user retention and enabling the tailoring of strategies to maintain player interest.

• Identified channels where users are 2.5x more likely to make in-game purchases, leading to significantly higher conversion rates for both new and returning customers.

• Achieved  35% revenue growth after they started targeting iOS users with a budget increase of 20% for iOS campaigns.

• Increased their spend 10x with Google Ads, now that they have the intel needed to justify the spend.