API Automations
Start a project →
Research · In submission 2025

Invisible Signals

Mapping cross-device surveillance and API telemetry in Connected TV ecosystems — a controlled, reproducible study of how streaming devices track and share data.

In submission · 2025 · proposed under a Ludlow Institute research grant.

Overview

Everyday Connected TV (CTV) devices — Roku, Fire TV, Android TV, and Smart TVs — operate within opaque data ecosystems that extend well beyond simple viewing analytics. These environments continuously collect and transmit telemetry about user behavior, location, and device identifiers through third-party SDKs, hidden APIs, and ad-serving protocols.

Invisible Signals systematically uncovers how consumer streaming devices track and share data across platforms, analyzing real-world traffic flows and API behavior. Using ethical, replicable methods, the project exposes the hidden data exchanges between streaming apps, advertising partners, and analytics services — work intended to promote transparency, shape privacy-labeling standards, and help consumers make informed decisions about their digital environments.

Why It Matters

By combining real-world ad-tech knowledge with rigorous ethical analysis, this research bridges the gap between consumer privacy and the opaque systems that shape it. It stands to become one of the first comprehensive, reproducible studies of CTV telemetry — an area of growing regulatory interest worldwide.

Scope

  • Network Traffic Analysis — what data leaves the device, and where it goes.
  • API Behavior Analysis — how apps and SDKs call hidden endpoints and ad-serving protocols.

Research Approach

The study uses a controlled, ethical test environment to capture and analyze network activity across major Connected TV ecosystems and related mobile apps.

1. Environment Setup

A controlled LAN environment with mirrored network ports for packet capture (Wireshark / tcpdump). Test devices span a Roku Streaming Stick, Amazon Fire TV, a Google device, and a Smart TV OS (Tizen or webOS). All sessions use clean, consented accounts and privacy-isolated logins so no real user data is exposed.

2. Network Traffic Capture & Analysis

Capture publicly visible traffic and metadata during app sessions across multiple CTV apps. Identify the domains, endpoints, and server destinations used for telemetry, analytics, and ad requests — and map the frequency and persistence of data calls, including those that occur without any user interaction.

3. API Behavior Inspection

Inspect the APIs and ad-serving protocols these apps rely on — endpoints, request and response patterns, and server-to-server data flows — to characterize what data is shared, with whom, and how persistently.

Deliverables

  • A 20–25 page open-access research paper (arXiv).
  • Visual mapping of telemetry flows between CTV apps and networks.
  • Comparative analysis of observed app behavior versus published privacy policies.
  • An optional video explainer, co-branded with the Ludlow Institute.
  • Progress reports at each milestone.

Timeline

A four-month program in four phases:

  1. Environment setup — network configuration and test planning.
  2. Data capture — network traffic and API calls across devices.
  3. Analysis — data analysis, policy comparison, and visualization.
  4. Dissemination — final paper writing, peer review, and publication.

Research Team

  • Kurt Sather — Co-Investigator. CEO of Show Me Television and API Automations, and CDO of Leanback Digital. A digital advertising strategist specializing in programmatic video, data, systems architecture, and consumer-privacy integrations, with nine years leading CTV analytics and privacy-compliance work across streaming ecosystems.
  • Shivani Patel — Primary Investigator. CTO and Lead Systems Engineer at Show Me Television and API Automations, specializing in API architecture, server-to-server data mapping, and privacy-centric telemetry analysis.

The team merges advertising-system expertise with privacy-first research principles, so findings are both technically accurate and socially impactful. All experiments are performed in controlled, consented data environments and documented for peer review and open publication.

Collaborate

We collaborate with institutes, journalists, and industry partners advancing data ethics, privacy innovation, and API standardization. To propose a collaboration or learn more about this research, contact our research team at contact@apiautomations.com.