What is Virality?

Virality isn't a word technically, but among savvy internet users and social media marketers, it is used to describe how viral content on the internet is. Or in other words, how popular and how quickly it spreads across the internet from one user to another. Of course before you begin to get into an actual measure, you'll need to know what "viral" content is and how it's related to social media.

How Would You Descibe What "Viral" Media or Content is?

Surprisingly, no one has put their finger on what it means to be "viral" on the web. It's a fairly loose term and if you were to search for the exact requirements you might hear claims such as, "10,000 views per day makes something viral" or "100,000 views per day." Many people don't even apply the terminology to anything other than videos on YouTube (or perhaps other services as well at this point). The truth is, all content can be viral.

The truth is calling something "viral" solely based on a specific or constant number of views is incorrect. This is because the views metric will increase with time as more people use the internet and your favorite web site. Views also aren't even a measure of what we're talking about here. We're talking about the spread of information and a "view" is merely a byproduct of that. Further, views were never an accurate internet metric to begin with. Look up "pageviews" vs. "hits" vs. "views" vs. "unique visitors" and you'll quickly start to understand the issue. You have to take them with a grain of salt because of how different systems decide to count what a "view" is. As an example: if one user refreshes a page one hundred times, it could be counted as one hundred views. Someone else might confuse that with "unique" views. So views aren't what makes something viral; however, it is a good thing to look at and if something is truly viral then you should expect to see many views. There's a relationship, but it is the definition or measure of being viral.

The Real Definition

It is the spread of information that makes content on the internet viral. Any content shared with enough frequency is considered to be viral to a certain degree. So the Virality Index was created to measure that.

How do we measure the spread of information? The dots are all there, you just need to connect them. You need to look at any given URL and how often it is shared across Twitter, Facebook, and other social media networks (Google+, Pinterest, MySpace, LinkedIn, Twitpic, and the list goes on and on) then figure out how many new shares there are per hour, day, week, and so on. Likes, comments, views, etc. All these metrics play a role in spread information to a different degree. Combining and utilizing all of these values in a specific manner to create some sort of calculation is an algorithm. In this case, we are interested in a sort of "velocity" for how quickly information is spreading.

A Simple Analogy

The basis for the research in this viral phenomenon can also be compared to the stock market. There is certainly a deal of probability involved here, but let's keep it simple for a minute. The stock market has a bunch of stocks and, in our internet world, that translates to URLs or the pages of content. Internet content is traded, or shared, much like a stock. We have a measure for the number of shares and the value. So if we were to plot the number of shares on a graph for a given URL we would end up with something that resembles a stock ticker.

What if we want to know what's going on with the entire stock market at once (or the whole of social media...or even the whole of the internet)? Well, that's a difficult question given how fast and frequently things change (not to mention how vast everything is). To help us, we have what we call a stock market index. There's several indicies of course and you can read all about them here on Wikipedia or many other places including your basic economics classroom. Much like the stock market, social media has an index as well for all of the bits of information that are shared across it. That's what we call the Virality Index.

A New Metric is Born

The Virality Index gives us an industry average for sharing. With it, we can determine what is and what is not viral. Keep in mind that determining what is not viral is sometimes as important as determining what is. How many times have you seen something be called viral that you just really didn't think was? More troubling yet, how many times have we been sold on something that was not viral? Companies and people alike make purchases based on this information and it has been mostly inaccurate for years. How can we trust what people say without any kind of measure to back up their claims? I don't know, but we have.

Many people miss the moment something becomes or ceases to be viral. By the time we say, "It's viral." it actually has already been viral for a while and may not even be viral any longer. With this new "virality score" metric, we have a new and more appropriate metric to help us make decisions when it comes to social media and viral content.

The Value?

Well there's a lot of value. The bottom line is clarity and an actual measure for how viral something is for starters. So that means an immediate value and validation in any claim that anyone makes (including bloggers and news outlets). It's also good for content owners to understand how their content is being shared. They can better make various decisions based on this kind of data.

It's also a ranking algorithm, so there is a lot of value in advertising for understanding sharing performance (and ultimately views and potential future views) for any given piece of content. Advertising networks could benefit greatly from these kind of metrics.

Making predictions and assessing "risk" (or change) also becomes a reality. This is important to advertisers, marketers, and content creators. Even the general public if applied to information they are concerned about. Much like the stock market, we can apply many other techniques and algorithms here to use this index as a guide in many ways. This is only the tip of the iceberg.

Nerd Out!

Still curious? Let's nerd out real quick.

It is not really enough to just look at how many "views" any given piece of content receives. While the number of mentions or shares on Twitter is a closer representation, it is also not quite enough to look at any one metric even if it is a "share" or "like" or "mention" metric. It is a combination of several metrics, each with their own value, that make up a consolidated value. There are other sites that will tell you how many shares there have been for videos, even over time, but that's still not a complete measure nor a new metric. It's also not very useful unless you are intimately familiar with how many shares videos tend to receive and what is going on at any given time on the internet in terms of sharing activity...Related just to videos or just to YouTube. As if YouTube was the only place where videos were published.

We're, instead, comparing these consolidated values against each other. In other words, it is a metric that depends on other content. This way, you don't need to be an expert to understand what's going on. You may even consider this "Page Rank 2.0" or loosely compare it to Klout. Though we aren't talking about any particular user or URL's "influence" in this case. We're talking about how "viral" something is and you can't really measure that without factoring in other content...And yes, this changes over time.

We used the stock market as a metaphor above, but we really could use many others. For example, we could say the average heights of all males in the the world. Obviously, we're talking about a bell curve and standard deviation here. Now, our bell curve doesn't exactly look so nice and perfect like the one here. This is because shares for a given piece of content across the internet is a log-normal distribution. There can't be any negative values. However, this chart here still applies with regard to standard deviation and percentiles. Remember those standardized tests you took in grade school? Same kind of logic. The virality score of any given piece of content is simply a percentile. That means if some video out there has a virality score of 70, then it has outperformed 70% of all content in the index. Obviously not all of the content on the internet can be measured. The content that ends up being responsible for generating this index is part of the algorithm.

Scores are measured and change on an hourly basis. Why? Simply because it is very difficult to measure any shorter period of time. The metrics that we are given by most services out there are cached and don't update that frequently. To be frank, there's not much need to either. The other problem is gathering that amount of data and running calculations in real-time is also more demanding than worthwhile. There's nothing wrong with every hour. Can you react to changes that might occur in a matter of seconds? Probably not and it would certainly be advisable to not make any business decisions like that either...In this case. This isn't the stock market - we just draw a lot of inspiration and ideas from it.