Crypto YouTube attention

Retail participation in crypto is cyclical and it leaves a trace. When a cycle heats up, people who were not previously interested start searching, watching and subscribing — and the clearest public record of that is the audience data of the channels they go to. Attention is not a valuation input, but it is a useful read on where the crowd is in the cycle.

The standard tools for this are imperfect. Search-interest indices are relative rather than absolute and get rebased without warning. Social-post counts are heavily contaminated by bots. Audience growth on large video channels is comparatively hard to fake at scale and is reported directly by the platform, which is why it is the measure used here.

Two series are tracked across a fixed set of major crypto channels: daily growth in total views, and daily growth in total subscribers. Each is normalized to a 0–10 scale against the full observed history, so a reading of 8 means growth is high relative to everything seen before, not high in absolute terms.

The two series say different things. View growth is fast-moving and reacts within a day or two to price action — it spikes on volatility in either direction. Subscriber growth is slower and reflects a higher-commitment decision, so it tends to lag but to mean more when it moves. Sustained divergence is the interesting case: heavy viewing without new subscribers suggests existing holders watching nervously rather than new entrants arriving.

This series is short. YouTube's API exposes only current totals with no historical endpoint, so the record began the day collection started and grows one day at a time — there is no way to backfill it. Until it spans a full cycle, the 0–10 normalization is calibrated against a limited window and should be read as provisional. The channel set is also fixed and English-language, so it misses regional audiences entirely. For those reasons this indicator is published as context and is not weighted in the risk model composite.