This HuffPost Canada page is maintained as part of an online archive.

TrendSpottr: Playing with Viral Possibilities

TrendSpottr: Playing with Viral Possibilities
This post was published on the now-closed HuffPost Contributor platform. Contributors control their own work and posted freely to our site. If you need to flag this entry as abusive, send us an email.

One of the most powerful videos of 2012, "Gangnam Style," was released in July 2012. On December 21, 2012, according to Wikipedia "Gangnam Style" became the first online video to record a billion hits. As of January 6, 2013, the music video has been viewed over 1.13 billion times on YouTube, and it is the site's most watched video.

Now imagine for a second having the ability to foresee the phenomenon of reaching 1 billion video hits, let alone 1 million. Not many can say they know a good thing when they see it.... At least, not this good! But if you did have that power, what would you do with the information?

This is the final article in a series that profiles four companies, who have developed technologies that are paving the way for businesses to rethink how they manage and measure information and communication in order to remain competitive.

This article profiles Mark Zohar, President and Founder of TrendSpottr. When I met with Mark, he told me about the Gangnam Style video and TrendSpottr's ability to predict, while views were well under 100K, how much this cultural pop video would catch fire when it hit the mainstream. This is my interview with Mark.

HJ: What fascinates me about your technology is its ability to consistently predict trends before it hits the mainstream audience. What factors come into play that detect a piece of content's potential to be picked up by TrendSpottr?

MZ: TrendSpottr's predictive algorithms analyze data streams in real-time to identify content that exhibits the attributes of viral growth. Such factors as the frequency, velocity, acceleration and amplification of content are used to predict trend data. By spotting emerging viral content within minutes of its origin, TrendSpottr is able to predict which information is most likely to trend, hours or even days before it has gained mainstream awareness.

HJ: When I think of trend data, it is about tracking the movement of information -- vast amounts of data -- over periods of time. Google Trends, for example, has amassed volumes of search data to accurately display what the world is searching for. There are varying opinions as to whether Google Trends' algorithm has the predictive gusto to accurately determine outcomes accurately using search data. Trendspottr differs because it only uses real-time data, which means the relevancy of information has a short lifespan. What is your stance in determining trends via user searches vs. content shares? Secondly, how can you consistently predict trends when the your algorithm is dealing with a relatively short period?

MZ: TrendSpottr's solution uses advanced algorithms to analyze real-time data for predictive trends. We are living at a time when consumer sentiment, political unrest and viral memes can spread like wildfire. TrendSpottr serves as an early warning system to alert businesses, governments and consumers to these impending trends before they have "tipped". This has proven to be of immense value to our partners and customers. TrendSpottr is currently being used for crisis and issue management, viral content discovery, predictive influencer analysis and ad yield optimization.

Google Trends offers a very different yet complementary solution for identifying trends. While Google Trends provides interesting historical trends using search data that may inform the future, there is nothing inherently predictive, algorithmic or forward-looking about this solution. Some of our customers use Google Trends in combination with TrendSpottr to provide them with both historical search trends and real-time predictive insights.

HJ: What recent trends was Trendspottr able to accurately depict? Take us through an example of that "moment" when content reaches that "tipping point".

MZ: A recent example was the photo the White House released where President Obama and McKayla Maroney posed together with their "not impressed" faces . The photo was taken on Thursday November 15, 2012 but was released early in the morning on Saturday November 17th. TrendSpottr picked up this photo within minutes of it being published on Saturday and issued an "Extreme Alert" to our customers indicating that this image was likely to go massively viral. In fact, TrendSpottr was able to predict that this photo would go viral within the first 35 tweets and accurately predicted that the photo would generate over 4,000 tweets within the next 5 hours.

Several of our customers used the predictive alert they received from TrendSpottr to launch customized content marketing and social ad buying campaigns that incorporated the Obama photo and that resulted in huge content sharing metrics and click-through rate (CTR) conversions.

HJ: Up until recently, the concept of "virality" has come to question. I've always been of the mindset that viral cannot be created. You cannot possibly create a viral video. Something has the potential to go viral based on the relationship between individuals. The value of that content and the context by which it is passed between users will determine its spread to beyond their immediate social graph. NOW, we not only can predict VIRAL content, it seems to me there is a formula to develop it. What's your stance on this?

MZ: I'm wary of claims that there is a formula one can follow to develop viral content. While there are some best practices we can learn from content that has gone viral (e.g., time of day publishing, media type, content messaging, etc.), almost all viral content is surprising, serendipitous and non-formulaic; think "Charlie Bit My Finger", "Golden Eagle Snatches Kid" and even "Gangnam Style". These are all examples of content that went viral because it resonated with people at the right time and without a specific formula or call to action beyond "this is worth sharing". I believe that manufacturing viral content is the equivalent of capturing lightning in a bottle, and let us hope it stays that way.

While no one knows how to make content go viral, what makes TrendSpottr invaluable is its ability to spot viral content at its earliest acceleration point and allow anyone to capitalize on this momentum.

HJ: You are providing a looking glass into the future. I see strong uses for journalists on the hunt for the "hot story", or for companies trying to stay ahead of the competition or keep up with their customers. What other groups do you see benefiting from TrendSpottr?

MZ: In addition to journalists and news organizations, TrendSpottr is being used today by large brands and PR agencies for real-time crisis and issue management. For example, one of the world's largest PR agencies was able to use TrendSpottr to identify and act upon emerging issues and trends related to its customers' sponsorship and ad campaigns during the London Olympics.

Financial services companies, including hedge funds, investment managers and financial news organizations, use TrendSpottr to gain early and predictive insights about individual stocks and macro-economic events that may impact financial markets.

TrendSpottr is also being used widely by content and social marketers to discover and share timely, trending and relevant content that will resonate with their audience. We have heard from many content marketers that TrendSpottr has helped them scale their content marketing initiatives and has resulted in significant increases in their social metrics and KPIs.

Another key customer group is advertisers who use TrendSpottr to predict the most effective keywords for their ad campaigns and optimize their ad spending based on emerging trends.

Other customer groups include governments, non-profits, media and entertainment companies and social analytics companies that license our API to integrate TrendSpottr with their existing suite of products.

HJ: I can see immense value for enterprise especially when combining your technology with customer transactional information. Being able to predict sales outcomes, reputational impacts from social data, customer service and sales data will allow organizations access to knowledge that will strategically impact the business overall. Is this an area you are developing solutions for?

MZ: One of our key partners is TrendSpottr is integrated with Salesforce Marketing Cloud and the Radian6 social analytics platform. This coming year we will be working on extending our integration to include other Salesforce applications, including Salesforce CRM and Salesforce Chatter. We are also starting to work with other enterprise customers and partners to integrate TrendSpottr with their business intelligence platforms, enterprise data networks and web analytics data.

With technologies such as these, perhaps we are much closer to not only predicting, but perhaps also developing formulas to create trends. Like Mark, I would hope that serendipity and surprise continues to be viral activity's main supporting cast, otherwise what a terribly uneventful "predictable" existence we would live.

Mark Zohar is co-founder and CEO of TrendSpottr, a Toronto-based predictive analytics company. Mark has over twenty years experience founding, funding, advising and operating technology, Internet and media companies. He previously co-founded View22 Technology, Inc., a 3D Web and rich media company, was a Partner at TD Capital Communications Partners in New York, led Forrester Research's Telecom and Media practice in Cambridge, MA, served as a technology consultant for the World Bank and worked in corporate development roles at several technology and media startups.

This HuffPost Canada page is maintained as part of an online archive. If you have questions or concerns, please check our FAQ or contact