StockTwits, Sentiment, and Message Volume: What the Numbers Actually Mean

StockTwits is a social platform for traders. Think Twitter, but every post is tagged to a stock symbol and labeled bullish or bearish. The product has been around since 2008 and has become one of the standard places where retail sentiment shows up before it gets priced in. Two metrics get cited from the platform constantly: the sentiment score and the message volume.

The way I think about StockTwits is that it's a noisy proxy for retail attention, not a forecast of price. The sentiment score tells you what the people posting feel right now. The message volume tells you how many of them are paying attention. Neither is a buy signal. Together, in context, they're a useful read on the social layer of a stock.

Plain English

Every post on StockTwits about a specific stock can be tagged bullish or bearish by the user (it's optional). The sentiment score for a stock is the percent of tagged posts that were bullish, over a rolling window. Message volume is just the count of posts mentioning the ticker.

What Sentiment Actually Captures

StockTwits sentiment is self-reported. A user clicks bullish or bearish before posting. That's the input. The aggregate is influenced by who happens to be active in a given hour, which means the score on a small-cap micro stock can be moved by 10 motivated posters. On a high-volume name like Tesla or Apple, the sentiment is averaged across thousands of users and is more stable.

The score doesn't capture intensity, position size, or quality. A retail trader with $500 in a position votes the same as someone with $500K. A first-day account votes the same as a 10-year veteran. The platform has algorithmic weighting that adjusts for some of this, but the raw sentiment is still a one-vote-per-post metric.

What Message Volume Captures

Message volume is more interesting than sentiment, in my view. It's a measure of how much retail attention is flowing toward a stock right now. When message volume on a small-cap suddenly spikes 10x, something has happened: a Reddit thread, an earnings beat, a viral tweet, a short squeeze setup. The spike often comes before the price move, especially on micro-cap names where retail flow can actually move price.

Volume doesn't tell you direction by itself. A 10x spike in posts could be people piling into a rally or people screaming as a stock dumps. You have to look at the sentiment in the same window to interpret it. High volume plus rising sentiment is bullish flow. High volume plus crashing sentiment is panic.

When the Signal Has Worked

The cleanest cases I've seen are short-squeeze setups on small-cap stocks. The pattern: low message volume for months, then a sudden spike, accompanied by high bullish sentiment. If the underlying stock has high short interest, that combination has predicted multi-day rips fairly often. The famous 2021 GameStop run was visible on StockTwits message volume two weeks before the squeeze peaked.

It also works in reverse. Stocks at all-time highs with sustained 95%+ bullish sentiment and elevated message volume are statistically vulnerable. When everyone agrees, there's no marginal buyer left. The crowdedness shows up in the numbers.

When the Signal Has Failed

On large caps, both metrics are mostly noise. Apple has 40 million shares trading hands a day. The most extreme StockTwits sentiment swing barely moves a name like that. The platform has more predictive power on small caps where retail flow is a meaningful share of volume.

It also fails when the platform itself is being gamed. Coordinated pumps on penny stocks have been a thing since at least 2017, and StockTwits is a target. Sudden bullish sentiment with a flood of new accounts is a red flag, not an opportunity. The platform has moderation tools, but pump campaigns still get through, especially on stocks with no institutional coverage.

Reading Sentiment In Context

I look at three things together:

Sentiment direction. Bullish or bearish, and is it changing?
Message volume relative to its own baseline. 2x normal? 10x normal?
The actual posts. Skim the top 20 messages. Are they substantive analysis or rocket emojis? The qualitative read is faster than any score.

That trio gives you a real read in maybe 60 seconds per ticker. A score by itself is mostly meaningless because the same number means different things on different stocks.

Takeaway

StockTwits is a retail attention signal, not a price forecast. Sentiment captures direction, message volume captures interest, and the qualitative read of recent posts captures intensity. Use it as one input on small-cap or volatile names. Mostly ignore it on large caps where institutional flow dominates.

The Take

StockTwits is a window into the part of the market that doesn't show up in 13F filings or earnings calls. That window is most useful when the retail flow is the marginal force, which is mostly small caps, meme stocks, and short-squeeze setups. For long-term investors in large companies, it's entertainment. For short-term traders in micro-cap names, it's genuine signal mixed with genuine noise. Knowing which case you're in is half the battle.