BlabPredicts is a platform that can predict which online conversations will trend up or down, by what velocity, and on what channel up to 72 hours in advance and with 70 percent accuracy system-wide. This platform is offered by a company called Blab, which is run by public relations industry veteran Randy Browning.
We spoke with Browning about this intriguing endeavor.
Q: What was the inspiration for creating Blab?
Randy Browning: As CEO of Publicis in the West, I took a local shop and built it into a $650 million agency powerhouse representing some of the biggest brands in the world. However I couldn’t shake the nagging feeling that marketing intelligence (historical data and backwards looking metrics) had lost its edge and no longer could keep up with how people were making decisions in real-time.
For brands to engage their agency resources and paid media, they have to control the message deployment, and so real-time marketing isn’t viable as it forces a state of reaction.
And so came my “aha” moment – I left Publicis to found Blab based on the idea of shifting time for marketers. To predict what, where and when their target audience will be talking about days in advance, so that brands can contextualize the creative and optimize the media to deliver higher real-time engagement.
We raised a few million dollars and began development on something that had never been done before: building a software engine that can dynamically classify the world’s conversations in real-time and predict the paths a topic’s conversations will take.
Fast forward to today and Blab is the first and only predictive analytics tool that optimizes how brands engage with their audiences in real-time by predicting which relevant online conversations will gain volume, by channel, by time – up to 72 hours in advance.
Q: Predicting conversational trends has a bit of a sci-fi element to it. How exactly does this technology work?
Randy Browning: To do this right for 100 million posts per day in real-time we needed to shake conventional thinking up. It turns out that challenge is also the mother of invention, and our team with data guru CTO Stefan Papp invented a patented, proprietary statistical approach that reveals the true breadth of digital conversations and predicts where they will go as much as 72 hours into the future.
We realized early on that taking the tried and true approach was the dragon we would have to slay:
• Searching reveals what you ask for, but we wanted to show you all of the conversation even if you don’t know you should be asking for it
• Linguistic Natural Language Processing (NLP) only knows what it’s been told – we might miss the most interesting new trends this way.
• Predicting which conversations will blossom or fade using conventional mathematical means can require rooms full of computers. We want to scale.
Instead, our engine renders conversations into a finite set of “patterns” (like DNA), and we discovered that conversations go viral in similar ways regardless of topic. So whether it’s the World Cup or war crimes in Syria, the way something grows or shrinks or goes from this channel to that channel, is very similar across the entire conversation canvas.
Put another way, we classify or “index” conversations much like the way Google “indexes” data search. Conversations are automatically organized by contextual relevance, size, and velocity for any area of interest using a seed term to fuel discovery rather than traditional search terms.
The result is that we analyze and organize 100 million conversations per day across all languages (including Klingon!), images, and video from over 50,000 data sources, and predict when these will grow or shrink and where.
Q: How can businesses take advantage of Blab in regard to their sales and marketing endeavors?
Randy Browning: The predictive power of Blab allows sales and marketing teams to understand tomorrow’s conversation today to make proactive decisions and take visionary actions, such as:
• Optimizing campaigns for theme, content, channel, and rotation
• Driving higher return on investment of paid media activities by adjusting on-the-fly for maximum impact
• Deploying more relevant, targeted and timely content across digital channels
• Assessing how advertising or promotional programs are resonating in real-time
• Improving click-through traffic by embedding predicted conversational topics and terms
• Predicting and mitigating potential brand risks by getting ahead-of and shaping conversations proactively
• Capitalizing on the spirit of the moment, and moments to come, through real-time engagement with influencers of a specific conversation
• Understanding the behavioral characteristics, digital gathering places and drivers of engagement for target customer segments
• Gaining a real-time view of the competitive landscape through the lens of the customer to develop strategies to acquire new and/or adjacent markets