Cheer Up, Grumpy Cat: Can Web Analytics work with Memes?

Thursday, January 24, 2013

Cheer Up, Grumpy Cat: Can Web Analytics work with Memes?

Like me! Pin me! Tweeeeeet meeeeeee!

Memes – like it or not, they are everywhere, spreading rapidly, and won’t be leaving anytime soon. Once confined to email inboxes, the omnipresence of social media platforms make it all too easy to share your latest comical discovery. Social media provides one other powerful element which fuels the spread of memes: feedback. ‘Likes’ from your friends on Facebook and the number of pins from fellow browsers of the Pinterest Meme Page will let you know very quickly if your picture of the “Hypno Dog” is funny, trying too hard, or about two weeks too late (more on this later).

Why bring this topic to the Web Analytics forum? Two reasons: one, it’s an area of the social networking phenomenon that stands to benefit from the predictive and investigative firepower of web analytics technologies; two, the results of a successful meme must sound like a dream to a web marketing firm. A successful meme has the power to spread organically through multiple social networks, converting users not through advertising keywords, but by empowering and encouraging people to be entertained by a piece of media, customize it, and then share it with everyone they know. It’s no wonder that the popular “lolcats” meme aggregator Cheezburger.com raised north of $30 million dollars in 2011[i].

So there’s some money to be had by collecting and propagating memes, but can analytics add any value to the equation? In other words, can web analytics predict and/or measure the impacts of memes in terms of how they drive traffic or promote a product?

Perhaps it’s best to first address the potential issues web analytics would face in the world of memes. With respect to predicting memes, the very nature of how a meme operates can be problematic. They are fickle creatures, prone to a meteoric rise on the Google Trends charts followed by an equally fast descent (see here and here for examples). Another difficult but equally important factor to consider is how one might measure the ‘it’s going viral’ factor – that is, how do you know a meme is spreading quickly enough to generate an acceptable level of interest and word of mouth? Unlike ads, there typically isn't a number associated with the clicks or impressions a meme generates. It's transmitting through twitter feeds, wall posts, and being posted on sub-forums. Are we doomed to making assumptions, and content saying the meme can be attributed to your increase in visitors to your landing page/Twitter feed/etc?

I hope not. While it is undoubtedly tough work to gauge the success of a meme, analysts are making progress in knowing when they have a hit on their hands. I mentioned the concept of instant feedback earlier. The feedback can be construed as a signal – telling you whether or not the meme has a life expectancy of a fruit fly. For example, take the success of “The Old Spice Guy” ads. We've all seen them, through one media outlet or another. But were you one of the viewers who contributed to the overnight success of the "Old Spice Responses"?

Visible Measures, an online video analytics firm, shows just how successful these Old Spice Responses were:

Old Spice: more popular than the President, source

6.9 million views in less than 24 hours. Doesn't take an analytics veteran to gauge the success of that campaign!

But we only know Old Spice Guy was successful due to a very primitive metric: view count. YouTube does that for us, and all Visible Measures had to do was start a 24-hour timer. While this proved to be a profound indicator Old Spice, I hope this is the tip of the iceberg in terms of what's possible with web analytics and its application to memes.