Moonlytics.io
Data / Social

Social data

Social activity turned into series you can compare. Raw platform chatter is counted, attributed to a canonical asset, expressed as a share of total tracked activity, and stored on the same hourly grid as market data.

What this data is

Two distinct things travel under the name "social data", and conflating them is the most common way to get a wrong answer.

The first is activity about an asset: how many public posts mention it, how much engagement those posts attract, and what share of all tracked crypto conversation it represents. This is hourly, and it covers the same 1,000 assets as the market snapshot.

The second is the reach of a project's own channels: follower counts, subreddit subscribers, post counts. This is daily, it covers a curated set of 69 assets, and it measures something different — audience size rather than conversation volume. It is documented under ecosystem data, and both are listed here so the boundary is explicit.

Raw activity to signal

A raw mention count is close to useless on its own. Ten thousand mentions means nothing without knowing whether the whole market is loud that day, how large the asset's usual baseline is, and whether those mentions came from distinct accounts. Normalization is what turns the count into something comparable.

Raw social dataupstream collection
public postsengagement eventsaccount activity
Normalization
attribute post → asset topicdiscard unattributablecount over 24h window
Aggregation
volume per assetinteractions per assetshare of total activity
Signal
socialVolume24hinteractions24hsocialDominancesentiment

socialDominance is the step that matters most. Because it is a share rather than a count, it removes the market-wide noise floor: an asset whose dominance rises while the whole market is quiet is doing something its raw mention count would not show.

Where it comes from

Post-level collection, topic attribution and the initial counting are performed by a commercial social and market data aggregator. Moonlytics does not scrape social platforms directly and does not hold raw post content.

What we do not claim

This is not coverage of "all social media". The tracked surface is whatever the upstream aggregator covers for a given asset topic — predominantly public posts on X, plus community sources. Private channels, Telegram and Discord are not part of it. An asset with an active Discord and a quiet X presence will look less social here than it is.

Update frequency

Hourly for the per-asset activity series, on the same schedule and the same timestamps as market data. Daily for the channel reach series, collected at a fixed time of day. See freshness for measured values.

Available fields

Hourly activity series

socialVolume24hdecimalNumber of public posts mentioning the asset over the trailing 24 hours, as counted by the upstream aggregator.
interactions24hdecimalEngagement on those posts over the trailing 24 hours — the reaction, reply and share volume they attracted.
socialDominancedecimalThis asset's share of all tracked crypto social activity, in percent. Sums to roughly 100 across the covered set.
observedAttimestampUTC snapshot timestamp, aligned to the top of the hour, identical to the market row for the same asset.

Daily channel series

Curated coverage: 69 assets. Null-heavy for projects whose community source is unmapped or private.

followersintegerFollower count of the project's primary X (Twitter) account.
followingintegerAccounts followed by that account.
tweetsintegerCumulative post count of that account.
subscribersintegerSubscriber count of the project's primary subreddit. Nullable.
activeUsersintegerConcurrently active users on that subreddit at collection time. Nullable.
likes / sharesintegerAggregate reaction counts on the tracked subreddit. Nullable.
mentionsintegerDaily count of tracked mentions for the asset's topic.
createdDatetimestampUTC collection timestamp of the daily record.
canonical social observation
{
  "asset": { "id": 1, "symbol": "BTC", "name": "Bitcoin" },
  "observedAt": "2026-08-31T09: 00: 00Z",
  "socialVolume24h": 41208,
  "interactions24h": 9137442,
  "socialDominance": 18.4
}

Limitations

  • Attribution is imperfect. Posts are matched to an asset by topic, and topics collide — an asset whose name is also a common word accumulates mentions it did not earn. Treat absolute volumes for such assets with suspicion; relative change over time is far more robust.
  • Bots are not fully separated. Coordinated posting inflates volume and interactions. The layer does not certify a mention as organic. Moonboard's bot-check surface exists precisely because this problem is real.
  • Sparse channel data. Reddit fields are frequently null, including for major assets, when a community source is not mapped or does not expose the figure. Null means unknown, not zero.
  • Social activity is not a price forecast. It is one input among several. Nothing here predicts a future price movement.

Which applications use it

Social volume, interactions and dominance feed the social momentum and discovery views on moonboard.ai. The weighting of each input into a published score is documented on moonboard.io, not here — this page describes the data, not the methodology built on it.

Related