Methodology

Social Momentum Index (SMI)

A weighted combination of post count, engagement and share of voice, ranked across the scanned universe.

Scale
percentile Relative scale — the value states rank within the scanned universe, not magnitude.
Formula
0.4·socialVolume + 0.4·interactions + 0.2·(socialDominance × 10 000) → percentile
Inputs
  • socialVolume24h
  • interactions24h
  • socialDominance
"Momentum" here means level, not change

The index contains no time derivative. It compares an asset's social footprint today against other assets today. It does not compare an asset against its own past, and a high SMI does not mean social activity is rising — only that it is large relative to the rest of the universe. To see change over time you have to plot the daily values, which the application does; the number itself carries no trend.

Purpose

SMI answers: how much of the crypto conversation is this asset currently occupying? It is the platform's measure of raw social presence — deliberately unadjusted for market size.

That is what separates it from the Moonboard Score, which divides social share by market share and therefore rewards small assets. SMI does the opposite: it rewards absolute scale. The two are complementary, and the interesting cases are the assets where they disagree — high SMI with a low Moonboard Score describes a large, heavily discussed asset; the reverse describes a small asset punching above its weight.

Inputs

FieldWeightMeaningTypical magnitude
socialVolume24h0.4Number of posts mentioning the asset10⁰–10⁵
interactions24h0.4Views, likes, comments, shares, reposts10¹–10⁸
socialDominance0.2Share of all crypto social volume, percent0–40

Formula

composite = 0.4 × socialVolume24h + 0.4 × interactions24h + 0.2 × (socialDominance × 10 000) SMI = percentile of composite across the universe

The stated weights are not the effective weights

This is the most important thing to understand about SMI. The three terms are summed without being normalised first, so each one contributes in proportion to its raw magnitude, not its nominal weight.

Using realistic figures for a well-discussed asset:

socialVolume24h = 9 000 → 0.4 × 9 000 = 3 600 interactions24h = 2 700 000 → 0.4 × 2 700 000 = 1 080 000 socialDominance = 2.20 → 0.2 × 22 000 = 4 400 composite = 1 088 000 interactions contribute 99.3 % social dominance 0.4 % post count 0.3 %

Interaction counts are two to three orders of magnitude larger than post counts, so they dominate the sum almost entirely. SMI is, in practice, a ranking of interaction volume with a small correction. The × 10 000 scaling on social dominance is an attempt to bring that term onto a comparable magnitude, and it moves it from negligible to merely small.

A version that normalised each term before weighting — as COR does with z-scores — would behave quite differently. Neither construction is wrong, but only one of them matches the weights as written, and it is worth knowing which you are reading.

Normalisation

The composite is converted to a cross-sectional percentile: the share of assets with a strictly lower composite. Because the underlying distribution is extremely heavy-tailed, the percentile compresses enormous absolute differences at the top — the difference between rank 1 and rank 5 might be a factor of ten in interactions and only a few percentile points.

Interpretation

SMIReading
~20Minimal measured social presence. Common for new listings and assets outside the covered platforms — often an absence of data rather than an absence of community.
~50Median footprint for the scanned set.
~70Clearly above-average engagement volume.
~90Among the most-discussed assets tracked. Usually a major asset, or a smaller one during a specific event.

SMI has no sign. An exploit, a lawsuit or a delisting produces the same high reading as a successful launch. Pair it with SDE or the sentiment field if you need direction.

Limitations

  • Effectively single-input. Interactions carry over 99 % of the composite for most assets.
  • No sign. Attention is measured, not its valence.
  • Ticker collisions. Assets with short or common symbols pick up unrelated conversation. Short tickers and dictionary words are the usual offenders.
  • Platform coverage defines the universe. A community concentrated on a platform that is not collected is invisible here — the metric reports zero presence, which reads identically to no community.
  • No authenticity filter. Coordinated posting and automated engagement enter the composite exactly like organic activity. See AI agents and synthetic engagement.
  • Universe-dependent. The percentile moves when the scanned set changes.

Data quality

Missing social fields default to zero, so an asset with a collection failure is indistinguishable from a silent one. This matters more for SMI than for any other metric, because social data is its only input. A sudden SMI drop with no news is more often a collection gap than a real collapse in interest.

Related