This Is How You Make Something Go Viral: An Impractical Guide

In effort to free up the longtime staff writers from a daily content quota and give them more breathing room, we instilled the whole traffic-whore model for about a month with varying results. The ultimate goal of this exercise was to show how, often times, the stories thought to be guaranteed traffic-drivers never materialized and how some of the longer stories outperformed them. The message: good is good, and you don't have to anchor your success to the oftentimes flukey nature of internet readers' tastes.

Internet publishers — successful ones — have concocted numerous sub rosa playbooks around traffic-goosing techniques: from the early days of annoying email blasts and Google-juicing headline techniques to, you know, more boobs. Then HuffPo came along with its lefty megaplex Drudge Report amalgamation which has morphed into the modern day Buzzfeed curating algorithm of business in the front, party in the back (or vice-versa), now considered standard wisdom for some of the bigger publishers. Admittedly, I hate this trend. I hate cats in hot tubs, cats sitting on babies, Keyboard Cats playing off babies who suck at singing, etc. Looking at them is fine. I mostly hate that, at some point, a viral video becomes the default hit-switch for a slow news day. But when your job is to grow a site's traffic, it's tough to ignore — and for the sake of the other writers, it's a necessary cog.

To acknowledge this trend of internet strip-mining, we hired Neetzan Zimmerman, founder of The Daily What, to essentially "cover" The Internet for Gawker. His coverage isn't done conventionally, nor is he gravitating to posts he particularly likes: he's tracking what the Internet's hive mind will react to by his own system, which is to me more of a sabermetrician's approach to viral content. I've asked him to outline his formula for both me, but also for readers to show how his method of traffic-generating content is actually employed and why you're reading what you're reading from him on a daily basis — A.J. Daulerio

The Most Valuable Content On The Internet: A Manifesto

Step 1: Background.

In the Summer of 2008 I started a small Tumblr called "The Daily What." It was meant to expand on a feature I had been running on a separate site dedicated to a zine I was self-publishing at the time called Man With Crisis. That feature, "While You Were Working," was just that: A compendium of the most "newsworthy" items of the day, as determined by the Internet.

I became increasingly intrigued by what I like to call "the Internet as Value Barometer" — deciding not only what there was to know, but what was worth knowing. Can the wisdom of the crowd be used to determine what warrants attention and what can be willfully ignored? If enough people think some story deserves to be shared, does that automatically make that story more valuable than all the stories not being shared? To put it another way: Do stories that are not being shared even matter?

To answer these questions, I first needed to come up with a structured formula for figuring out the Most Valuable Content (MVC) on any given day.

Step 2: Experimentation. I started by tracking down the Internet's main sources for MVC.

In order to seperate the "listeners" from the "storytellers" I put together two lists: One with all the websites from which content was being culled for dissemination, and another with all the sites doing the disseminating.

Examples of the former include big deals like Fark, Reddit, Digg (this was in 2008, mind you), Slashdot, MetaFilter, b3ta, and Oh No They Didn't, as well as lesser-known sites like Super Punch, Cynical-C, Arbroath, and 3 Quarks Daily, that seemed to be a regular source of content for larger, more popular sites. Examples of the latter include Boing Boing, the Gawker Media sites, BuzzFeed, Neatorama, Laughing Squid, and Urlesque.

The point was to try and suss out not only where the content was actually coming from (as opposed to where it was ending up), but also to keep track of an item's movements across the web in an effort to pinpoint the exact moment at which it could be defined as having "gone viral."

Using an RSS reader to organize my findings, I established a set of categories for each site-type (news, vids, pics, link-sites, general interest, tech, geek, gossip, entertainment, design, art, fashion, and food), and further arranged each site within each category by order of influence. The top tiers were reserved for "mainstream sites" — sites where most of the sharing was occurring (i.e., the content disseminating sites listed above). The lower tiers were reserved for sites that supplied content to top tier sites, but were themselves low on visits.

This allowed me to have a streamlined, bottom-up view of content progression: From the lower tiers to the top, where the viral magic happens.

In short order I was able to track content from the point of inception to pre-mainstream saturation. I learned to recognize when items were reaching that critical stage of going from radar blip to full-scale red alert. Some of The Daily What's most successful "gets" could be attributed to the employment of this system: Sad Keanu, Nyan Cat, Double Dream Hands, and Gotye's "Somebody That I Used to Know" — which The Daily What posted months before many other major sites became aware of its existence.

Step 3: Maintenance.

The Internet is mercurial to say the least, and while certain conversation-mongers like Reddit, BuzzFeed, and The Huffington Post are likely here to stay, lower tier sites — the ones that tell them what there is to discuss — change rapidly.

In order to stay as current as possible, I make sure to run a spot-check of the most visible sites at least once a week. Refreshing the index with the most fruitful lower tier content sources is only half of it: Losing the dead weight is crucial as well. My rule is simple: If a site hasn't produced at least on item of value during the week, it drops down a tier. If it bottoms out and still hasn't proven useful, it's gone.

Step 4: Predictability.

At this point, you should be well on your way to gaining a firm grasp on the inner workings of the Internet. So much so, that you don't even need to wait for content to be deemed valuable by a top tier site in order to know it will eventually end up with that designation.

Let's review what we've learned so far:

  • The Internet starts below the surface, so get to know the unsung sites supplying the big shots with fodder for dialogue.
  • Pay particular attention to the items that get published on lower tier sites. Chart their progress across similar tiered sites. If they shows up on enough lower tier sites, they're likely in the process of going viral.
  • Learn what the top tier sites like and look for. This will help to cut down on time spent waiting for an item to make its way across the trenches.
  • Ditch the sites that don't improve predictability: Lower tier sites are only as good as their efficacy in determining the output of top tier sites.
  • Note that it may not always be readily apparent why something is more shareable than something else, but having that intangible quality necessarily makes that thing more valuable. That may seem like a truism, but the fact is that, once you've established the viral probability of a video or an image or an idea using the methods outlined above, the "why" is no longer important. It has been commodified by the Internet, and lifted above the noise. It is significant because it was significant enough to get noticed in the first place.

    (A word of caution: When the internet is particularly starved for content, it tends to reach for low-hanging fruit. And that fruit is low-hanging for a reason: It's ripe to the point of being rotten. Insipid, pointless, patently unintesting and unfunny items are brought to the fore when they would otherwise remain unmissed in obscurity. Be on alert: Just as in other areas of life, obnoxious elements can, occasionally, gain undeserved attention when there is nothing good on.)

    Step 5: Results.

    So do stories that are not being shared even matter? The answer, undoubtedly, is no. How could they, when they are seemingly invisible? This isn't a comment on their bona fide merit, but if the purpose of the Internet is to engender exchange, then anything not being shared must therefore, in this context, be worthless.

    As above, understanding why something works is an exercise in futility. More than that, it is completely unnecessary when the point is to be part of a larger community, not an outlier. Being able to determine what will be discussed next is, therefore, far more valuable. Advancing the conversation will always be looked upon more favorably than trying, and likely failing, to start it.

    In the end, the Internet will always tell you what you need to know because it is a digital extension of the world writ large, and out there, as in here, the greatest story will always be retold.

    Image by Jim Cooke.