Online hustles

What is viral? Base rates, explained.

"How many views is viral" has no single answer. A threshold can be an absolute count, a multiple of your own median, or a position in a ranking, and each gives a different base rate. Here is what that means for anyone planning around virality.

There is no platform-wide median and no agreed viral threshold. A reported rate only means something with its definition and sample size beside it. Use your own last posts, not a headline percentage.

Advertisement

Three common definitions

What the studies we found do and do not say

Academic work uses small, purpose-built samples. One TikTok study used 400 videos from 20 hashtags and labelled the top ten and bottom ten per hashtag as viral and non-viral (arXiv 2111.02452). Google researchers analysed 1.5 million YouTube videos in 2010 and found that only 21% of the most popular videos by 30-day views could be classified as viral in their sense (Google Research, "Catching a Viral Video"). Neither number is a chance that your next video goes viral; both depend on how the researchers defined the word and which videos they looked at.

A vendor analysis reported by SocialCrawl, described to us as 1,008 videos from 62 creators and dated 2026-10-01, is said to find that 18.2% of videos reached 2x the creator's own median and 4.6% reached 5x. We could not locate the primary publication when this page was written, so treat the figures as unverified and secondhand. Even taken at face value, it is one small vendor sample, it measures against each creator's own median rather than a shared scale, and it is not specific to one platform. It does not support a statement like "you have a 4.6% chance of a hit".

Why a platform-wide median does not exist as a usable number

The platforms do not publish a median views per video. Third-party averages by follower count mix accounts of different ages, posting rates and niches, and an average of a heavy-tailed set is dominated by its largest values. A median is less affected by outliers than a mean, which is why the scenario tool asks for your median and lets you set the rest.

How to use this

  1. Pull your last 20 to 30 posts and note their views.
  2. Take the median and write down how many were at least 2x and 5x that median.
  3. Enter those counts as probabilities in the scenario calculator.
  4. Look at the typical post and the mean side by side. Plan around the typical one.

Methodology

This page is explanatory and states no formula of its own. The multiple-of-median definition is arithmetic: a post is a hit if views / median is at or above the multiple you choose. The count of hits divided by the number of posts is an observed frequency for your account and period, and it can be noisy when the sample is small: with 20 posts, one extra hit moves it by 5 percentage points.

What this does not cover

Results vary. This page is arithmetic on assumptions you enter. It does not know what any platform, buyer or audience will pay you, it is not a forecast, and it is not an offer or a promise of income. Tax content is general information, not tax advice; check the IRS or CRA source or ask a qualified preparer.

Everything runs in your browser. Nothing you type is sent or stored.

Sources

Retrieved 2026-10. The SocialCrawl figures are reported secondhand and not independently verified here.

FAQ

How many views is viral?

There is no agreed number. Some people mean a million views, some mean a multiple of their usual views, and researchers use rankings within a sample. The answer changes with the definition.

What percent of videos go viral?

No reliable platform-wide percentage exists. Reported figures come from small samples with their own definitions, so they do not transfer to your account.

Is there a median number of views for TikTok or Shorts?

The platforms do not publish one. Third-party averages are not medians and mix very different accounts.

How do I find my own base rate?

Count how many of your recent posts reached 2x or 5x your median views and divide by the number of posts. Keep in mind that a small sample makes the figure noisy.

Advertisement
Advertisement