How social media trends spread, from first Reel to everywhere

What does it actually mean for a trend to “spread”?

On Reels or Shorts, a trend is usually a format that other people can reuse: a piece of audio, a dance, a template, a joke structure, a transition, a caption pattern. A single post with a huge view count isn’t a trend by itself. A trend exists when many people make their own version.

That distinction matters because it means trends spread through behaviour, not just attention. Watching is cheap; making a version yourself takes effort and a little social risk. The research below helps explain when people are willing to take that step.

Why do people share some posts and ignore others?

A landmark study analysed every New York Times article over three months and tracked which ones readers emailed to others. The key factor was not simply whether content was positive or negative, but how much it activated people. Content evoking high-arousal emotions such as awe, anger or anxiety was shared more; content evoking low-arousal emotions such as sadness was shared less. Overall, positive content was more viral than negative content (Berger & Milkman, 2012).

Moral emotion adds another push. In more than half a million social media posts about polarising issues, each additional moral-emotional word (a word carrying both moral and emotional weight) was associated with roughly 20% more spread, mostly among people who already shared the poster’s views (Brady et al., 2017).

Novelty matters too. A study of around 126,000 rumour cascades on Twitter found that false stories reached more people, and faster, than true ones. The authors suggest that novelty and the emotional reactions false stories provoke may explain part of the gap (Vosoughi et al., 2018). This is one reason alarming posts travel so well, and why doomscrolling feeds can fill up with them.

Is going viral the same as being popular?

Not quite. Researchers studying a very large set of online cascades introduced the idea of structural virality to separate two shapes of spread (Goel et al., 2016):

• Broadcast: one account with a huge audience posts, and most views come directly from that single source. Wide, but shallow.
• Viral chains: person A shares with B, who shares with C, and so on through many generations. Each step is small, but the chain is long.

Two posts can reach the same number of people through completely different shapes. Their analysis showed that popularity and this kind of virality are distinct properties (Goel et al., 2016). In Reel terms: a celebrity using an audio clip and thousands of small creators remixing it are different routes to the same view count.

Why do some trends need several nudges before they catch on?

A clever experiment created artificial online communities and watched a health behaviour spread through them. People were much more likely to adopt it when they received signals from several contacts rather than one. And the behaviour spread farther and faster in tightly clustered networks, where friends share friends, than in random networks with more long-distance shortcuts (Centola, 2010).

This idea, often called “complex contagion”, helps explain why Reel trends often grow inside one niche first (a city, a fandom, a profession) before jumping elsewhere. Seeing three people you follow use the same audio feels like permission to try it yourself.

What does a typical trend life cycle look like?

Putting these findings together, many online trends follow a rough arc. This is a conceptual model to help you think, not a law, and real trends often skip or repeat stages.

1. Seed. Someone posts a format that is easy to reuse. Most seeds go nowhere.
2. Remix. A cluster of creators who see each other’s work start copying it, which is the reinforcement Centola’s experiment describes.
3. Bridge. A few people who belong to several communities carry it across, for example from dance creators to comedy creators.
4. Mainstream. Large accounts adopt it, so broadcast-style reach takes over.
5. Fatigue. Once it’s everywhere, it is no longer novel, and the emotional spark that drove sharing fades.

People who do this professionally try to spot trends at the remix or bridge stage, while there is still time to act. Our sister guide sets out a practical method for identifying viral Reels (our sister guide, Reel Trend Research Guide), and doomscrolling vs professional trend research explains how that work differs from casual scrolling.

What role do recommendation algorithms play?

Feeds don’t show posts in the order they were made; recommendation systems choose what each person sees, largely based on what similar content and similar users have engaged with. That means the human tendencies above, such as responding to arousal, moral emotion, novelty and negativity, can be amplified at scale. Randomised headline tests have shown that negative words raise click-through rates (Robertson et al., 2023), and any system that learns from clicks will learn that.

What we can’t tell you is exactly how a given platform’s algorithm weighs these signals. Independent researchers have pointed out that the data needed to study such design questions is largely held by the companies themselves (Montag et al., 2019). Be wary of anyone who claims to know the “secret formula”.

What does this mean for you as a viewer?

• Notice the emotion first. If a post makes you feel a jolt of anger or anxiety, that is exactly the feeling that makes content travel. Pause before sharing.
• Check novelty against accuracy. Surprising claims spread faster than accurate ones, so the most shocking version of a story is often the least reliable.
• Remember the feed is a sample. What reaches you is shaped by your network and the recommendation system, not a neutral view of the world.
• Use the life cycle. If a format is at “mainstream”, you are not missing anything by stepping away.

For why these emotional hooks make feeds so hard to leave, read the science and psychology of endless scrolling.