Ep 228: How motion data from phones can inform your growth strategy
Dieter Rappold, co-founder and CEO of ContextSDK, joins Apptivate to explore how motion sensor data and on-device AI are reshaping mobile marketing strategies. He explains how contextual signals from accelerometers, gyroscopes, and other device inputs can help marketers identify the right moment to engage users, optimize conversion timing, and improve monetization without relying on personal data or tracking permissions. The conversation covers the evolution from event-based to moment-based optimization, practical implementation considerations, privacy implications, fraud detection use cases, and how contextual intelligence could power the next generation of agentic AI and decision-making in mobile growth.
Listen and subscribe:
Listen on Spotify
Listen on Apple Podcasts
Questions addressed in this episode:
- What types of smartphone sensor signals can marketers use today?
- How does ContextSDK turn motion data into actionable growth insights?
- What is the difference between event-based and moment-based optimization?
- Which app categories benefit most from contextual engagement timing?
- How does on-device AI change personalization in a privacy-first world?
- What implementation effort is required to test contextual optimization?
- How can motion signals support fraud detection and ad performance?
- Where could contextual intelligence influence agentic AI and future UX?
- What strategic priorities should mobile marketers focus on next?
Timestamps
- 0:04 — Motion sensor data in mobile marketing
- 0:56 — Accelerometers and gyroscopes explained
- 2:06 — Founding story and origins of ContextSDK
- 3:10 — Awareness gap among mobile marketers about sensor data
- 4:24 — Airbnb example illustrating real-world user intent
- 5:44 — Physics-based data vs opinion-driven marketing models
- 6:55 — Session duration differences and conversion timing opportunities
- 8:22 — Event-based vs moment-based optimization strategy
- 10:13 — App verticals and use cases for contextual timing
- 11:15 — Additional signals beyond motion sensors
- 12:02 — Model training requirements and data scale needed
- 14:02 — Privacy compliance, ATT and permissionless personalization
- 15:11 — Fraud detection applications using motion behavior
- 15:43 — Ad network integration and performance uplift example
- 17:28 — Context data as signal layer for agentic AI
- 19:18 — Strategic priorities and competitive positioning for marketers
- 20:23 — Limits of sociodemographic targeting frameworks
- 22:06 — How to connect with ContextSDK team
- 22:29 — Rapid-fire questions
- 24:58 — Episode wrap
Quotes
- (4:45) “If you're walking down the street and open Airbnb, you're probably looking for the key code of the apartment that you have booked. But if you're comfy on the sofa and open Airbnb, you are probably planning your next vacation.”
- (5:03) “Humans constantly move in a three dimensional space while they're using their smartphones and the apps on it… we have different needs, different pain points, different session durations and different likelihoods to convert in different actions.”
- (8:47) “I don't think we should get rid of event-based targeting, but we should combine it with moment-based targeting because both things can tell us something. Event-based gives us behavioral context (meaning the behavior in the app) - but the question is, when is the timing right?”
- (13:35) “Based on how we built this, our architecture and our approach, we don't need ATT, we don't need any permissions, and we are out of the box GDPR compliant because we don't collect any PII, we don't collect a unique user ID and we don't collect unique device ID.”
