Ensuring Housing Access Through Accurate and Reliable Mobile App Measurement

Challenge

To improve access to their services and provide the most engaging experience for customers, Property Finder increased their focus on mobile apps for iOS and Android users. Their paid media team was investing in user acquisition campaigns across multiple media partners, endeavoring to optimize their omnichannel measurement strategy. However, they were hindered by several pain points:

• Getting a single source of truth across media partners
  With active campaigns across multiple media partners, the team lacked a unified view of cross-partner performance and an unbiased source of truth for final attribution on app installs and in-app events. Logging into each partner’s dashboard was inefficient and showed only siloed performance.
• Capturing reliable and accurate app engagement data
  The team faced internal infrastructure scaling issues in attempting to reliably ingest app installs and in-app events, then syndicate them to other elements of their tech stack. Server outages and system errors caused data gaps that impacted data integrity and real-time campaign and user experience optimization.
• Combating costly mobile ad fraud
  Expanding their paid spend to new ad networks and partners coincided with a rise in suspicious campaign activity. Abnormal traffic patterns along with unrealistic click volumes and conversion rates set off alarms that mobile ad fraud was stealing ad dollars. Authenticating app installs was also an issue, as sudden traffic spikes indicated possible spamming of spoofed payloads.

Solution

Property Finder connected with Kochava, a mobile measurement partner (MMP), to gain new measurement and attribution tooling for their mobile apps across both Google Play Store and Apple App Store.

Leveraging Kochava, the Property Finder team

• Gained a single source of truth across their media mix
  Kochava’s global footprint of thousands of integrated partners enabled Property Finder’s mobile team to plug and play with all growth partners across social, display, video, connected TV, and other channels. Omnichannel, cross-partner attribution data was unified within Kochava, so the team no longer needed to bounce among partner dashboards to piece together performance data. This holistic attribution enabled more seamless optimization of campaigns for maximum growth and return on ad spend.
• Stabilized their app measurement data and ETL processes
  The mobile team implemented the Kochava software development kit (SDK) in both iOS and Android apps. This reduced the infrastructural burden on their internal systems that was causing outages and data gaps. The Kochava SDK features built-in redundancy safeguards—ensuring always-on connection to a global inboarder network for 99.9% uptime. This standardized and stabilized the measurement of Property Finder app installs and in-app events, as well as the extract, transform, and load (ETL) process to syndicate this data across other partners and their marketing tech stack.
• Activated real-time fraud detection and prevention
  Recognizing the anomalous signs of mobile ad fraud in their campaigns and app signals, the Property Finder team wasted no time in activating Kochava’s fraud prevention tools. This enabled always-on, real-time detection and prevention of ad fraud using advanced machine learning. Anomalous traffic was flagged and blocked from attributing to real conversions—preventing Property Finder from paying fraudulent parties for bogus traffic. Server-side authentication technology was also able to detect fake install spoofing, blocking fraudulent payloads.

Impact

Gaining a single source of truth across their media partners, reliable app measurement, and built-in fraud prevention was a game-changer for Property Finder’s team.

Over 90 days, Property Finder observed:

• Cross-partner attribution of over 700K new app installs
• SDK measurement & authentication of 1.3M+ app installs and 170M+ in-app events
• Real-time blocking of 2.15M fraudulent clicks and 400K+ fake install payloads