30-day window
The report centers on supported viewing events dated within the latest 30 days of the parent-selected history update.
How ScreenSignals Analyzes YouTube Watch History
A transparent explanation of the ScreenSignals 30-day analysis window, evidence sources, scoring dimensions, privacy boundaries, and important report limitations.
Start my child’s reportVideo-level evidence
Each usable public video is analyzed before results are combined into viewing patterns, channel summaries, and the overall context score.
Age-aware context
The child’s parent-provided age helps interpret language, suggestiveness, violence, intensity, and developmental fit. It is not used to create a child account.
Signals, not verdicts
Scores organize evidence for the adult parent. They do not diagnose a child, certify a video as safe, or replace review of the underlying context.
1. The adult parent chooses the source data
The parent or legal guardian creates a supported YouTube History export through Google Takeout and selects the downloaded ZIP in ScreenSignals. Children do not create ScreenSignals accounts, purchase subscriptions, submit information, upload viewing history, or operate the service.
The browser inspects the Google Takeout ZIP on the parent’s device. The original ZIP is not uploaded or stored by ScreenSignals as a file. Supported extracted viewing entries and the records needed to create and maintain the private report are sent to and stored by the service.
Read the parent’s Google Takeout guide
2. ScreenSignals validates the history
The import checks for a supported JSON Watch History file, parses valid watched-video records, and removes exact duplicate rows while preserving legitimate repeat watches at different times. Records outside the current analysis window, without usable timestamps, or dated implausibly in the future stay outside the report.
History may still be incomplete. Paused or deleted activity, signed-out viewing, shared accounts, private or unavailable videos, and account changes can all affect what the export contains.
3. The report resolves public video context
For usable video links, ScreenSignals attempts to retrieve public information such as title, channel, description, duration, category, thumbnail, and publication details. It also looks for available captions or transcripts. When transcript evidence is unavailable, the system may use a cautious metadata-only analysis and marks the evidence as lower confidence.
Unavailable, private, deleted, malformed, or otherwise unreliable videos remain visible where appropriate but do not distort aggregate scores through guessing.
4. Each usable video is evaluated across separate dimensions
- Educational value: learning, curiosity, creativity, explanation, and useful skill-building.
- Repetitive or low-value pull: passive repetition, low-context novelty, recurring sensationalism, and attention-grabbing formats. Lower is better.
- Language: harmful or age-inappropriate language considered in context.
- Suggestiveness: sexual or suggestive material considered against age and context.
- Violence and intensity: violence, fear, dangerous stunts, and imitation risk.
ScreenSignals treats isolated words cautiously. A news report, documentary, fictional story, joke, or critique can contain the same vocabulary as material that endorses harmful behavior. Transcript and video context matter.
5. Video signals become a parent-readable report
The private Screen Report combines analyzed videos into a 0–100 context score, category mix, enrichment balance, repeated formats, channels, trends, and items that may deserve a closer look. It also suggests conversation starters, watch-together ideas, and quality-screened channel starting points related to interests visible in recent viewing.
Approximate watch-time views use available public video duration and history events. YouTube Watch History does not reliably state how many minutes of each video were actually watched, so these estimates should not be treated as precise attention measurements.
6. Serious signals and low-value viewing stay separate
Repetition or low educational value alone does not create a serious safety warning. Closer Look items are reserved for stronger age-aware signals involving language, suggestiveness, violence, fear, dangerous behavior, or related intensity. The parent can open the public video, review why it surfaced, and record the family’s decision.
What the report cannot establish
- Who was physically watching a shared device.
- Whether the full video played or received the viewer’s attention.
- A child’s intent, belief, mood, health, or developmental outcome.
- Whether every video from a channel is suitable.
- Whether deleted, private, signed-out, or paused-history viewing occurred.
- Whether an automated classification captured humor, irony, cultural context, or family circumstances correctly.
How parents should use the result
Review the evidence, apply knowledge of the child and family, and use the report as a starting point for conversation or a narrower platform setting. Do not use a score to shame a child, label a creator, or make a medical conclusion. For wellbeing or immediate safety concerns, consult an appropriate qualified professional or emergency resource.
Privacy and corrections
The full data categories, service providers, retention practices, advertising boundaries, and parent rights are described in the Privacy Notice. Questions about methodology or a factual correction to a public guide can be sent through the private support form.