Setting up Data Infrastructure to Optimize Marketing Efforts: Simple App Case

The Challenge

Simple, a popular intermittent-fasting app available on both iOS and Android, was investing heavily in marketing across the US market. However, their analytical capabilities couldn’t keep pace with their marketing spend. The team lacked the infrastructure to properly evaluate ROI by channel, understand user acquisition costs at a granular level, or identify which marketing initiatives were actually driving profitable growth.

Without a dedicated data analytics team, marketing decisions were based on platform-level reporting from Google Ads and Meta — which, as every experienced marketer knows, gives an inflated view of each platform’s contribution. Simple needed an independent, unified analytics system that could provide honest answers about marketing effectiveness.

The mobile app space presents unique attribution challenges that Simple was struggling with. iOS App Tracking Transparency had significantly reduced the visibility of user-level attribution from Facebook and other ad networks. The team was making budget decisions based on platform-reported metrics that were increasingly inaccurate — Google claiming credit for conversions that Meta also claimed, with the sum far exceeding actual installs. Without an independent measurement system, Simple was effectively flying blind with a multi-million dollar annual marketing budget.

Our Approach

We built a comprehensive marketing analytics infrastructure designed specifically for mobile app user acquisition:

• Data Collection Architecture: We set up automated pipelines to extract data from multiple sources: ad platforms (Google Ads, Meta, Apple Search Ads), mobile attribution (AppsFlyer/Adjust), in-app subscription events, and backend user activity data. Each source was normalized into a consistent schema.
• Marketing Analytics Data Model: We built a data model that connected marketing spend to downstream user behavior. This wasn’t just about installs — it tracked the full funnel from ad impression to install to trial to paid subscription to retention. This allowed calculating true CAC (not just cost-per-install) and LTV-based ROAS.
• KPI Dashboard Suite: We designed dashboards focused on the metrics that matter for mobile app growth: channel-level CAC and ROAS, cohort retention curves, subscription conversion funnels, geographic performance breakdown, and creative performance analysis.

The key insight driving our design was that mobile app marketing analytics must account for the full subscription lifecycle. A channel that delivers cheap installs but poor trial-to-paid conversion is worse than a channel with expensive installs but high lifetime value. Our data model captured this nuance.

We also implemented a creative performance analysis module that connected ad creative variations (video vs. static, messaging themes, duration) to downstream subscription behavior. This went beyond standard platform reporting — which only shows install rates — to reveal which creative approaches attracted users who actually subscribed and retained. These insights directly informed Simple’s creative production strategy, helping them produce more of what works and less of what generates cheap installs with poor conversion. Additionally, we built automated anomaly detection that flagged unusual changes in channel performance within 24 hours, enabling rapid response to market shifts.

Results

• Optimized marketing strategy with clear visibility into true channel-level efficiency beyond platform-reported metrics.
• Full-funnel analytics connecting ad spend to subscription revenue, enabling accurate LTV:CAC calculation by channel.
• Identification of marketing channels with high install volume but poor downstream conversion, enabling budget reallocation.
• Cohort retention analysis revealing product-level opportunities to improve trial-to-paid conversion rates.
• Scalable infrastructure supporting expansion to new geographies and marketing channels without rearchitecting.

Technologies Used

Python, SQL, BigQuery, Tableau, mobile attribution platforms (AppsFlyer), automated data pipelines.

Project Screenshots