Your behavioural profile is not a personality test. It is the shape of your attention.

What I was trying to avoid

When I started designing the behavioral profile feature, I spent a week looking at how other apps in this space handle user categorization. The pattern was consistent and, I thought, wrong: personality archetypes. "The Deep Diver." "The Casual Scroller." "The Creator Enthusiast." Quiz-style labels that flatten complex behavior into a thumbnail description.

These archetypes are not useless. They are legible, shareable, and they give users a quick frame for understanding their own habits. But they misrepresent what behavioral data actually contains. Your YouTube behavior is not a personality type. It is a set of patterns that change week to week, shift with context, vary by time of day, and evolve as your interests and circumstances change. Compressing that into a fixed label discards most of the information.

I built the Gazenest behavioral profile to do something different: model the shape of your attention, not categorize it.

What signals feed the profile

The profile draws on everything Gazenest tracks:

Session structure: How you enter sessions (search, subscription feed, homepage, direct link), how long sessions run, whether you tend toward single-video sessions or extended browsing chains, how often sessions end with deliberate exit versus browser close.

Completion patterns: What percentage of videos you watch to completion, where you typically drop off in videos you abandon, whether completion rate varies by video length or topic.

Temporal patterns: When you watch, how your Clarity Score varies across times of day and days of week, whether you have identifiable drift windows.

Content breadth: Diversity Score history, channel concentration, topic distribution across the rolling 90-day window.

Intentionality signals: Ratio of search-initiated to recommendation-initiated watching, rewatch frequency, re-upload engagement rate.

Each of these signals is a dimension. The profile is the combination of dimensions, not compressed into a label, but kept as a multi-dimensional model that changes as your behavior changes.

What the categories actually mean

The profile presents several categories, each reflecting a dimension of behavior:

Intention mode is the aggregate reading of whether your YouTube use tends to be purposive (you know what you want before you open it) or ambient (you open it and follow what appears). Most people are somewhere in the middle, and the reading shifts week to week. This is the most directly useful dimension because it reflects the gap between what you intend and what you do.

Drift susceptibility is a measure of how strongly your behavior responds to algorithmic nudges. High drift susceptibility means that when the algorithm serves you something adjacent to what you were watching, you tend to follow it. Low susceptibility means you make more deliberate breaks, you stop a chain, return to search, start fresh. Neither is inherently good or bad, but knowing your susceptibility helps you understand when structural controls (like hiding the sidebar) will make the most difference.

Depth vs. breadth reflects whether you tend toward sustained engagement with specific topics and channels, or broad sampling across many topics and creators. Both can be intentional. The dimension is useful when you're trying to understand changes in your Diversity Score, a natural depth preference reads differently from algorithmic narrowing.

Temporal consistency is a measure of how stable your watch patterns are across days and weeks. High consistency means your behavior is predictable and structured. Low consistency often correlates with reactive use, watching more when stressed, watching differently on weekends versus weekdays, sessions that are highly variable in length and intent.

The honest limits

The profile is a model, not a measurement. It infers behavioral tendencies from a set of signals that correlate with those tendencies. It can be wrong. If your YouTube use is genuinely atypical, if, for instance, you use YouTube primarily for research in a professional context, the profile may not reflect that accurately, because the signals that indicate intentional use in typical contexts may pattern-match to something different in your case.

I've been deliberate about not over-claiming. The profile page in Gazenest presents the dimensions with confidence intervals rather than precise numbers where confidence is genuinely uncertain. If the data is sparse, you've been using Gazenest for two weeks, the profile will tell you that the reading is provisional.

There is also a dimension the profile cannot capture: your goals. Whether your current watch patterns are aligned with what you want your YouTube use to look like is a question the profile can inform but cannot answer. It can tell you that your Diversity Score is low and your drift susceptibility is high. It cannot tell you whether that matters to you or what you should do about it.

Why the profile changes

The profile is a live model, not a fixed result. It recalculates continuously as new data comes in. A behavioral shift, using the Customise Feed controls, changing your starting point from homepage to subscriptions, reducing your watching in a specific drift window, shows up in the profile within a few weeks.

This is intentional. The goal is not to diagnose a permanent type. The goal is to give you an accurate model of your current patterns, one that updates as those patterns change. The profile is useful when it's accurate, and it stays accurate by staying current.

The behavioral profile is available on Gazenest Plus.