If you had the superpower to predict the future, what would you do? Buy a lottery ticket? Bet on a sports team?
Consider broader horizons. To crush it in the retail marketplace, think about forecasting the success of the next must-have product. Or invest, foreseeing the growth of potential financial markets to become the next Warren Buffet. Turning to politics, how about predicting the winners of the midterm elections, earning guest appearances on all the news networks? If you’re concerned about public health, announce the next medical breakthrough at the best time to be accepted by doctors, patients and the general public alike. For national security, you could anticipate peaceful protests or acts of domestic violence and be able to distinguish the difference.
These are not pipe dreams for Cognovi Labs, a predictive analytics company that applies advanced technological and psychological developments to gaze into possible futures for government, social welfare and private sector interests. Predictive analysis itself is not new. For decades, forecasters have gathered historical data to assess potential outcomes, typically by using quantitative statistical analyses of occurrences, prices, scores, etc. But Cognovi offers more than a backward look at old, cold numbers: it anticipates how impassioned humans will make decisions and act accordingly.
Cognovi’s methods are based on several insightful premises. First, they realize that with the growth of personal computing and social media, many significant sociopolitical attitudes and events are dictated by whole communities or large populations, e.g., our collective reactions to the Covid pandemic or political disinformation. Second, their cognitive and behavioral science expertise confirms that the decisions and behavior determined by these attitudes are primarily driven by emotion, not facts or logic. How a person feels about a topic may be subconscious or arbitrary, but it always influences the decisions they make. Third, to measure these emotions, researchers must discern sentiment patterns expressed in social communication, accomplished by applying complex analytical processes.
This is where Cognovi’s technologies excel.
The result of eight years of research at Wright State University – funded by the US Air Force, the Department of Energy, and the National Science Foundation – Cognovi’s proprietary and patented methods merge artificial intelligence with human psychology. On the AI side, their software engineers and data scientists employ machine learning, natural language processing and emotion AI. On the human side, their clinical psychologists apply prestigious research in behavioral, affective and cognitive psychology, neuroscience, sociology, and the emerging field of neuroeconomics.
What do they do with this wealth of knowledge and expertise?
Step One: Cognovi Emotion AI has learned to identify the most salient emotions expressed in eighteen natural languages, even semantically difficult implicit meaning (when “that’s a great look” means just the opposite). It data-mines words from numerous sources, such as short-form social media texts, internal industry sources, transcribed conversations or custom data generated by their Dynamic Diagnostic Interview. Specifically, it searches for ten topic-specific expressions of emotions, which can be labelled joy, anger, disgust, fear, sadness, surprise, amusement, contempt, hope and trust. Cognovi then displays this information over time in customized dashboards or ongoing reports, categorized by demographic, geographic or economic variables.
Step Two: After they determine what people say and how they say it, Cognovi employs human decision-making psychology to predict motivation and behavior: what they intend to do and what they will do next. Often, results are not foreseen. In one case analyzing response to announced Covid vaccination eligibility, Cognovi discovered that including obese BMI as an underlying condition actually made overweight people feel targeted and shamed, and outraged those who were not included. Obviously, neither response was anticipated by the state health agencies that issued the guidance, nor favorable toward vaccination campaigns. Thankfully, in addition to initial descriptive reports Cognovi offers ongoing analysis and updates, customized for their client’s needs.
There’s also Step Three, since predicting the future is not the same as influencing it – if you know the next Powerball numbers but don’t buy a ticket, you won’t wake up a multi-millionaire tomorrow. Cognovi puts its descriptions and predictions to work by identifying the words and narratives that evoke precise emotions and lead to successful behavioral outcomes. Recently, Cognovi analyzed food service employee motivations to resign, and the results are illuminating. In January 2021, employees felt joy and hope in posts about quitting that reflected optimistic aspirations, but actual intent to resign was 3%, measured separately by Cognovi’s Intent score. Later in September, employees displayed anger, disgust and sadness, revealing hostile reactions that were accompanied by 37% intent to resign. The Cognovi takeaway: companies should encourage positive thoughts about quitting because joy and hope reduces the motivation to do so. In such cases, Cognovi provides consultation tailored to address negative results and produce a desired effect.
Cognovi also blunts the effect of unexpected events with its real-time triage tool, which assesses whether an event can change customer or public sentiment or behavior and details proactive efforts to mitigate that change. For both corporate and public sector clients, they proactively contribute to reputation enhancement and branding campaigns.
As effective as they have been, Cognovi cannot predict everything, like random lotteries or the winner of the Super Bowl. However, they can predict where a controversial NFL quarterback pick is welcome (see Draft day fan analysis: Tua Tagovailoa – Cognovi Labs). And for anyone who can run with that information, it’s all they need.