SensorTowerSensorTower

Speeding Up the Back End with Graph Theory

Nov 26, 2019·4 min read

Summary

Sensor Tower improved a sluggish endpoint by analyzing its profiler data and benchmarking. The first optimization added a dirty flag to nodes, skipping re-encoding for unchanged parts, cutting time by 30%. The second optimization replaced linear tree traversal with precomputed paths, enabling logarithmic-time node access.

This required occasional background updates to maintain accuracy. Combined, these changes reduced latency from 300-500ms to about 80ms, a fourfold speedup, with negligible impact on user experience.

You Might Also Like

AppsFlyerAppsFlyer

Cross-channel marketing analytics: how it works in 2026

Cross-channel marketing analytics isn't about putting Meta, Google, and TikTok numbers side by side—they often double-count the same customer journey. Fragmented identity is the real culprit; without a first-party Customer User ID, attribution measures platform credit, not customer value. The article explains that deduplicating conversions across mobile, web, and CTV can lift attributed revenue by 30–60% and improve ROAS by 20%. It walks through attribution models, warns against platform-native analytics, and advises using an independent MMP for true cross-channel measurement. Ad ops takeaway: fix identity resolution first, because AI-driven optimization and budget allocation depend on trustworthy, deduplicated data.

Rachel Siegman·Sep 3, 2026·14 min readRead article →
MolocoMoloco

Inside Moloco's CARA: the AI that compounds

The open internet presents unique challenges for performance advertising: fragmented identity, closed first-price auctions, and non-stationary supply. Moloco's CARA compound architecture tackles this with six integrated technical domains—Campaign Automation, Supply, Ad Recommendations, Bidding, Creative, and Signals—running on a unified ML infrastructure. Key insights for ad ops: the system continuously learns from every interaction, uses knowledge distillation to serve real-time predictions under 10ms latency, and validates improvements through rigorous live experiments. In 2025, 65 validated model updates reduced CPA by 17% and improved ROAS by 27%. The key takeaway: compound AI architectures that connect prediction, bidding, creative, and data can unlock measurable performance gains beyond walled gardens.

Tal Shaked·Aug 7, 2026·14 min readRead article →
AppsFlyerAppsFlyer

Pitfalls of modeling LTV and how to overcome them

LTV forecasting is challenging due to stale ML models and user heterogeneity. Effective systems combine ML, Bayesian methods, and secondary models to reduce bias and variance for accurate predictions.

Shani Rosenfelder·Jan 29, 2020·6 min readRead article →
MetaMeta

Performance Spotlight: How Laura Geller Turned Creative Volume and AI Into a Competitive Edge

Will Frank of Laura Geller describes their Meta ad account as an ecosystem where creative, media, offers, and channels interact. Key insights for ad ops leaders: build a constant creative pipeline to feed AI-driven delivery, test with patience as value optimization took three weeks to show conversions, use AI to automate reporting and pressure-test hypotheses, and focus on system architecture rather than manual targeting. Their multi-touch attribution, MMM, and incrementality testing form a 'suite of truth' to validate performance. The marketer's role is shifting to strategy and AI stewardship.

Aug 13, 2026·4 min readRead article →
MetaMeta

Performance Spotlight: How Instant Hydration Built a System for AI to Scale

Instant Hydration scaled Meta spend in under 18 months by letting creative variety, not manual targeting, drive audience discovery. The brand diversified creators, ran Partnership Ads under brand and creator handles, and used AI to tailor briefs to creator audience themes. It runs Advantage+ broad targeting and automated placements, intervening manually only for lifecycle exclusions and brand-safe creative. Incremental attribution delivered ~35% more net-new visits and revived “burned out” creative. Takeaway for ad ops: automate delivery, own creative strategy, use incremental measurement, and extend creator-led systems across DTC and retail.

Meta·Sep 10, 2026·6 min readRead article →
AdjustAdjust

How mobile app marketers should approach CTV | Adjust

CTV has become performance-ready for app marketers. Recent acquisitions (Fox/Roku, Walmart/Vibe) signal a shift to self-serve, measurable channels. Marketers can reuse existing UA creative instead of producing TV ads. QR codes drive direct response, but halo effects often matter more. Start with small, additive test budgets and measure assists/incrementality to understand true impact. CTV offers a way to find incremental users and diversify beyond paid social.

Nicoline Str·Aug 4, 2026·7 min readRead article →
AppsFlyerAppsFlyer

Five marketing lessons from football’s biggest tournament

Mobile engagement during the football tournament was fragmented, not continuous, with spikes lasting ~3 minutes around goals and pauses. Purchases peaked at halftime, not during play. Emotional stakes drove higher engagement than audience size—the third-place match outperformed the final (+21.7% vs +6.3% lift). Local factors (regulation, payment infrastructure, routines) caused market-specific behaviors. The customer journey continues post-match, requiring measurement beyond live events. Ad ops should align campaigns with attention patterns, optimize for local nuances, and track the full funnel.

Stephanie Chevalier Naranjo·Jul 28, 2026·10 min readRead article →
AppsFlyerAppsFlyer

How an eCommerce Loyalty Platform Cut Time-to-Purchase by 66%

A leading eCommerce loyalty platform integrated AppsFlyer's deep linking and audience segmentation with Braze's engagement platform to unify personalization, measurement, and lifecycle orchestration. This solved fragmented data and manual campaign production, driving a 66% faster time to first purchase, 500% uplift in push revenue, and 50–80% revenue lifts in email/content cards. The key insight for ad ops: accurate deep linking and behavioral data are foundational—when they work as one system, personalization scales and ROI improves.

Amar Balu Rajan·Aug 26, 2026·9 min readRead article →

More from SensorTower