Viral Goes Scientific

In my class presentation this week, I spoke about the potential market share, brand resonance and, ultimately, profit to be gained by companies who are able to crank out cost-effective and successful viral ad campaigns. As I noted, the strategy is applicable to anyone, from large companies looking to create a viral success in combination with more traditional campaigns in order to stand out (note: happy “hump day”), to charities looking for a fundraising boost.

If done correctly there are plenty of stats to prove the post-campaign benefits of viral advertising. But a great question posed at the end of the presentation focused on how we know what’s going to go viral. Right now there isn’t really a perfect answer for that – but someday there will be.

Much like advanced analytics has taken over all of the major professional sports, statistical analysis has been brought to marketing. While this specialized area of study is still in its infancy, ad firms are investing in research around what connects with people via video or post or any available platform. The level of detail is incredible. Observe:

equation

This equation translated means that customer influence effect (CIE) of person j is equal to the degree of spread [aka the amount of messages posted] (w) plus the influence level of person j on person i times the customer influence effect of person i.

Follow?

Deconstructing the equation further allows us to arrive at this conclusion: a person’s ability to influence someone else to become a customer of a certain company relates to how much they post about that company/idea/product while also factoring how good they are at being an influencer based on what they say, how they say it and how well they know the person being influenced while also factoring in how good of an influencer the person being engaged is themselves.

Bottom line, it’s a lot to take in and process. Which is why this has become a full scale area of scholarly study. Field research is underway to explore consumer levels of engagement with advertising, investigate the staying power of a message and figure out what correlation there is between engagement, staying power and making a purchase decision. In a separate area, researchers are attempting to learn more about the networks behind word of mouth and how that can lead to improved ad messaging and a better understanding of how many additional “eyes” word of mouth brings to a message. The goal is to build better value behind traditional marketing terms, such as net promoter score and customer referral value.

Much like on base percentage is now considered a golden statistic (whereas it was once a throwaway number on the back of baseball cards), perhaps we will learn that a stat like net promoter score is invaluable to companies looking to understand how to engage the right “activist” customers. There are some companies who seem to think they’re on to something along those lines. Yet much of today’s research is actually suggesting that older marketing metrics don’t really tell us much about customers or their preferences.

As crazy as this sounds, a certain segment of agencies, while doing no independent research of their own, are predicting that viral campaigns are a wave about to crest. They’ve created the role of “online talent agent” and are signing viral video stars. Viral video sensations are licensed to TV shows and commercial producers just like more traditional copyrighted content. Online personalities are booked for paid appearances. It turns out, this may be the most successful market for viral yet – with many licensures netting anywhere between $10 – 100,000 per contract.

And there’s still a long way to go – there are dozens of different avenues to go down. Research proposals have covered topics such as:

  • The application of social marketing relative to industry (business-to-consumer vs. business-to-business)
  • Relationship of lifetime value score to a customer’s level of influence
  • Retention levels of word of mouth campaigns
  • Statistical analysis of the minimum level of influence required to convert one person to customer

For any math lovers out there, a whole host of great numbers and equations await you. For those who love seeing talking camels, talking babies or flash mobs – there’s plenty of things headed your way too.

Main Sources:

Kumar V., Bhaskaran, Vikram, Mirchandani, Rohan and Shah, Milap. “Creating a Measurable Social Media Marketing Strategy.” Marketing Science. March/April 2013. 32(2): 194 – 212.

Kinchen, Rosie. “Now Charlie Can Bite Your Finger Too.” The London Sunday Times. March 11, 2012. Pg. 17.

Teixeira, Thales. “How to Profit from Lean Advertising.” Harvard Business Review. June 2013. 23 – 25.

9 comments

  1. The business potential in this area is enormous which is probably why it is the focus of so much scientific research. University’s that are able to figure out the algorithm can sell that to a startup company who essentially hold the holy grail of marketing. The ability of viral videos to scale from a small campaign, to a national dialogue like the dress discussed by the nation a few weeks ago make these tactics extremely appealing. Great job of taking the equation and breaking it down for those of us who are unable to decipher complex-looking equations. I really enjoyed reading this follow up to your presentation!

  2. What a great way to follow up on your presentation topic! I’ll admit that I was intimidated by the equation when I first saw it, but you did an excellent job of explaining it. What I’m currently most interested in is the “influence level” of a person. Brands are currently experimenting with paid, earned, and owned media, and the kinds of paid media are constantly evolving. In the past, paid media was often limited to a commercial or print ad, but now many brands are paying “influencers” to promote their brand on their blogs or Youtube channels. These influencers have huge followings and have earned the trust of these followers, which is why brands are willing to pay upwards of $100,000 per contract like you mentioned. I’m interested in learning more about how these influencers have become so popular!

  3. Really interesting insight into viral videos — I posted on the same topic this week but we both took opposite approaches. I looked more and the psychological and emotional drive behind viral videos and their reception so reading about the statistical analytics is really interesting! I find it hard to believe that a math equation can account for all of the variables in an area like (maybe because I am not a math person) but it will be interesting to see if companies adopt such an equation and how successful it is.

  4. meganvtom · ·

    One of the most interesting posts I have read all semester! I find it so interesting that scientists have been able to quantify traditionally qualitative data. It seems to be a fairly simple concept however I wonder how scientists will quantify influence level, because social media is so interconnected and each person might share with others, etc. I predict that the best marketers in the future will understand how to best use the equation and other statistical tolls alongside the psychological and emotional aspects of marketing. I am excited to see where this heads in the future, especially if researchers are able to quantify the minimum level of influence needed to convert a prospect into a customer.

  5. When I first started my own business, I went online and made a few videos in the very hope of them catching on and going viral. I had always heard that there is “no formula” when it came to achieving that type of popularity and I learned first hand how true that was. It really is crazy what scientists or researchers are able to achieve now. This formula blows my mind because I had no idea you could quantify things like influence or company message. I thought you did a really great job at breaking down the science behind the variables. This is definitely something I’m going to keep up with because the implications of creating viral marketing campaigns could really be make or break with companies. Great post!!

  6. Very well-research post and a nice follow-up to your presentation Todd. I understood that advertising is an area of scholarly study but I had little idea that such study in the area of viral advertising was so advanced. I appreciated the great job you did in deconstructing the customer influence effect equation as well as the reporting on how companies are acquiring “online talent agents” and signing up viral video stars to enhance their viral video building capacity. I love many of the well known viral videos out there but I’m unaware that any has influenced any purchasing decision that I’ve made. As the field grows, it’ll be interesting to see correlations on how the hype and popularity of viral videos relates to purchase behavior and revenue generation.

  7. I’m impressed. You’re going hardcore when citing “Marketing Science.” (I’ve published once there, and I only understood about half of my own paper). Seriously, the question I have about these types of “metrics” is that we can come up with all sorts of fancy ways to quantify things, but how well it matches reality is another story….particularly as reality may evolve over time with SM.

  8. Great post! I really like how you followed up on the questions from your presentation, for I think many people wonder what actually makes an advertisement go viral and how companies are able to pull this off. I had no idea that there is an actual scientific equation that has been formulated around this concept, but with the rise of viral marketing it definitely makes sense! I will be working at an advertising agency next year, so these posts have really caught my interest, as I am sure that this will be a hot topic of discussion going forward. Really nice job!

  9. So glad you wrote this post as a follow-up to your presentation! You are brave to take on this topic. Especially with Big Data, math and data analysis seems to be a huge opportunity for marketing. Although one does not want to be purely quantitative, one probably does not want to be purely qualitative either. I really appreciate your research and explanation on the influence equation. It will be interesting to see how this subject evolves in the next three to five years. While I am sure marketing data collection advances will be made in that time, I hope interpretation of this data will evolve equally. I always find it easier to pull the data versus making conclusions regarding what the data means. Thanks again!

%d bloggers like this: