What is social momentum?
How much an asset is being talked about - and the widely repeated claim that this arrives before price moves.
The components
| Component | Measures | Answers |
|---|---|---|
| Social volume | Posts mentioning the asset | How many people brought it up |
| Interactions | Views, likes, comments, shares | How far it travelled |
| Engagement rate | Interactions per post | Whether posts landed |
| Social dominance | Share of all crypto conversation | How much of the market's attention it holds |
Different combinations of these appear under the name "social momentum" across the industry. There is no standard definition, so the term means whatever a given product computes — which makes checking the formula worthwhile before comparing numbers between sources.
Level versus change
"Momentum" normally implies rate of change. In practice most social momentum metrics — including Moonboard's — measure level: how large the footprint is right now, compared against other assets right now.
These are different questions and can point opposite ways. A major asset always has a high level and may be in decline. A small asset can be growing fast and still rank low.
SMI is a level metric. It contains no time derivative and does not compare an asset against its own past. A high SMI does not mean social activity is rising. Seeing change requires plotting the daily values, which the application does.
The lead claim
The reason anyone collects social data is the belief that conversation precedes price movement — that a community notices something before the market prices it.
The mechanism is plausible: people discuss what they are researching, and research precedes buying. But the claim is asserted far more often than it is tested, and testing it properly is harder than it sounds:
- Both series respond to the same news. A protocol announcement moves conversation and price simultaneously. Correlation is expected regardless of any lead.
- Social data is not real-time. Posts must be published, collected upstream, aggregated into a 24-hour window and pulled by the platform. A measured "lead" can be an artefact of comparing a delayed series against a live one — see refresh and latency.
- Base rates are usually omitted. "Social volume rose before 70 % of large moves" is uninformative without knowing how often it rose before everything else. Social volume rises constantly.
- Look-ahead bias is easy to introduce. Using revised or backfilled data produces results that look excellent and mean nothing.
Moonboard does not publish a finding on this. It is the first entry in the research notes queue, with the conditions any answer would have to meet.
How Moonboard computes SMI
With one important caveat documented on the SMI page: the terms are summed without being normalised first, and interaction counts are orders of magnitude larger than post counts. For most assets interactions carry over 99 % of the composite, so SMI is effectively an interactions ranking. The stated weights are not the effective weights.
Reading social momentum well
- Attention has no sign. An exploit produces the same high reading as a launch. Pair it with sentiment or news.
- Low can mean absent, not quiet. Communities on uncollected platforms report as zero.
- Check for ticker collisions. Assets with generic symbols accumulate unrelated conversation.
- Cross-check against turnover. Social activity rising with no movement in trading is the pattern most worth examining — for authenticity as much as for opportunity.
Related
- SMI methodologyThe formula and its weighting problem
- Social dominanceShare of voice as a measurement
- Social intelligenceWhat social counts capture and miss
- AI agentsWhen the conversation is not human