Net Gain 2016: A Conference of Qualitative and Quantitative Disrupters in MR

By Arundati Dandapani

Technology as a disrupter in the market research industry was the theme for this year’s MRIA Net Gain Conference, an annual event that brings together buyers and sellers of research, decision-makers, practitioners and students in search of the industry’s future. Home and overseas presenters shared strong insights into the future of the North American market research landscape. Unsurprisingly, the US elections drew repeated references.

Highlights from the event are described below:

The Internet of all your things

Greg Dashwood of Microsoft Canada, began the day with “The Internet of Things: Crafting insight in a connected world.” The Internet of Things (IOT) is a consolidation of everyday objects equipped to transmit and receive data. Doing business in such a deeply connected world requires a new kind of savvy and attention central to our lives today.

Data show the majority of CEOs are making digital transformation a priority because of how it is disrupting business. According to Dashwood, digital transformation is the 4th industrial revolution after the invention of the steam engine, electricity, and the microprocessor. It hinges on four imperatives: engaging customers, empowering employees, optimizing operations, and transforming products. Dashwood listed the impact of such digital transformation in manufacturing, the public sector, service industries, and others.

IOT is everywhere, in vehicles, assets, buildings, cities, even currency; there’s a connected world of solutions on devices and assets using cloud and infrastructure intelligence. Dashwood engaged with case studies of Microsoft and Rolls Royce as companies that have harnessed IOT’s sensor-based technologies profitably. Rolls Royce’s Trent XWB is the world’s most efficient aero engine because of its sophisticated sensor technology, which allows pilots and other stakeholders to track fleet status and fuel utilization through advanced analytics and communicate repercussions internally. Smart garbage cans that report when they are full and need emptying, washing machines that track revenue per usage and change dynamic pricing according, smart-shelves, smart buildings, and even smart cows are non-automotive examples of the proliferation of IOT.

The restaurant industry as an IOT-enabled marketplace holds much promise too for mining real-time feedback at restaurants through table-talk analytics or tools allowing customers an instant rating of their server, food and restaurant, serving as a constructive feedback mechanism that incentivizes staff to be at their best at all times.

A typical loyal restaurant customer offers a lifetime value of $2000 to $5000 for any business. Customer retention is thus the most prized commodity today and IOT can improve that too. An average business hears back from only 4% of its dissatisfied customers. In order to avoid customer attrition, a business must reach out more to its unhappiest customers who offer negative feedback, because they are a great source of learning to the service industry.

Arundati Dandapani@itadnura

IOT will saturate the marketplace but there will be winners too @MRIAARIM Germany and Asia Pacific are IOT leaders

The list of IOT enabled devices is long. Germany and Asia Pacific are market leaders in consolidating and leveraging their IOT. All businesses will need the internet of things sooner than later to gain insight and agility, build competitive edge, and redefine their customer service experiences.

Emotional analytics can predict what people do and why

The next speaker was qualitative disrupter tech entrepreneur and global innovator Lana Novikova, CEO of Heartbeat Ai Technologies Inc. Lana talked about the predictive powers of emotion analytics. Fifteen years into the industry, while managing quant studies at Nestle Ice Cream, Lana decided to shake things up by pursuing psychotherapy and affective neuroscience to better understand consumers’ desires because, “quantitative research could only tell you so much.” She spent hundreds of hours talking with people about their emotions. All social data is biased, but with her new venture Heartbeat Ai technologies, Lana applied algorithms to the text of open ended answers and employed a clustering system of psychology. Using the translanguage map of human emotions with three levels of emotion (primary, secondary and tertiary) she segmented emotions and used language and facial analytics to better understand customer loyalty.

Heartbeat AI Technologies enables a full automation of text data analysis from surveys, political polls, social media and other data sources. Lana’s team used such emotion analytics to ask Canadians what they felt about chocolate, their period, the elections and Prime Minister Justin Trudeau. But her most “presented” example remains the case study of Australia’s elections this year. Candidate Malcolm Turnbull beat Bill Shorten on neither emotion nor anger but on trust. Trust became an important differentiator in comparing models of likeability and leadership. Similarly, in the recent race to the White House, emotion analytics revealed a surprising range and surge of emotions for Trump and Clinton that acted as predictors of the real life outcomes: people regarded the former with more “trust” and “joy” and the latter with more “anger”. Emotions also drive the NPS score for banks. NPS increases as joy, trust and love increase.

Arundati Dandapani@itadnura

The translanguage map of humanemotion bubbles of anger joy void etc Heartbeat uses emotionclustering tech @MRIAARIM

Using a combination of AI, machine learning and emotion analytics, it is possible to build intelligent predictive models that improve customer loyalty and drive profits. Given the power and complexity of these award-winning technologies and algorithms, Lana concluded with the hope that if we combined all the tools we have access to, we can bring more empathy and understanding to research.

The Presidential Likeability vs. Unlikeabilty Panel

The US-based duo of Andrew Konya from Ramesh, an artificial intelligence company, and Merrill Raman from Penn Schoen Berland, collaborated to launch highly scalable focus groups (“supergroups”) that quantified the results of qualitative inputs on Ramesh, with the moderation, analysis and content organized at Penn Schoen Berland. They talked about neural networks and language embedding technologies that they adopted when trying to determine Presidential likeability, a key driver/predictor of voter intent historically, whether during the Bush vs. Kerry or Clinton vs. Trump contests. “Likeability” was determined as a group of qualities that triggered a favourable reaction. Empathy, ability, authenticity and personality were its core components. They constructed a largely qualitative questionnaire. Results revealed that both Hilary Clinton and Donald Trump fared poorly on empathy. Then they got creative and asked questions like: “which candidate do you think is funnier?”, “which would you be likely to invite to a BBQ?” – answers to both favoured Trump. However, the answer to “which candidate would be more likely to be your friend?” was “neither.” Neither candidate fared high on “trust” and “dislike” for both candidates was high as well! With vast amounts of unstructured data that can sometimes seem conflicting and disparate, Raman expressed an appreciation for solid platforms and text-analytical tools that serve researchers well in this “golden age of qualitative research.”

The Death and Rebirth of Marketing

President of Fresh Squeezed Ideas, John McGarr held forth on marketing ideology in his talk, “Reincarnation: the death and rebirth of marketing”. Brands that are able to build themselves around ideas and ecosystems by understanding culture’s current values and grasping the myriad media platforms and digital tools available are the most successful today. “People are drawn to what reflects their values,” he said, citing the case of Lululemon, a company that did not “invent” yoga but became wildly popular for building an ecosystem of products, services and channels to yoga at a time when buyers wanted it most.  “Every choice a customer makes is an act of culture,” he said. Brands can learn from pop culture that has successfully mirrored global economic uncertainty and outlook over time.

Pop culture has always sufficiently responded to cultural needs. For example, the social value of “being able to take care of yourself” was echoed with Survival TV series offering ideas on “how to survive.” Uber and Netflix are examples of brand responses to a cultural need for fluidity, flexibility and low-commitment options. Parenting values have also changed dramatically in a turbulent world as parents seem to value their children’s achievements over their authenticity (natural inclinations). Lego’s Empower kids is an example of brand messaging that reflects that value. Marketing is about understanding which values are important to your users and in using such insights to change their lives for the better.

Arundati Dandapani@itadnura

It is every marketers responsibility to be a force of good in consumers lives. 1 cdget it right like Lululemon or not like Scion @MRIAARIM

The New Age of Advertising Effectiveness Measurement

Melanie Drouin of Research Now talked about factors that should be considered in order to maximize your media buy. Tracking ad exposure is now about tracking people and their devices. This can be done through a combination of single-source cookie online panel, passive monitoring of cross device digital ad exposure, cross-media exposure data from digital (desktop, mobile, in-app, native) and traditional (print, TV, radio and out of home) channels, integrated data sets and graph ids or APIs. Multi-touch attribution modeling or frequency of exposure to digital media and to combinations helps effectively analyze your advertising ROI. One should analyze single-source exposure to brand KPIs to determine individual attitudes and understand which media channels are driving the metrics. The key to maximizing any media buy is in understanding all media together and not in isolation. In these early stages of tracking in-app ads and geo-targeting (digital audience tracking validation), it is important to understand that “one size does not fit all’.

When Behavioral Science Turns the Classical Marketing Model on its Head

Alex Hunt, president of the NYC-based firm Brain Juicer, shared breakthroughs in behavioral sciences and convention-defying insights about today’s consumer. The rules of classical persuasion-based marketing no longer apply, he said, because people think a lot less than they think they think! If thinking is painful, is marketing and market research keeping up with psychology?

Nobel Prize-winning author of Thinking Fast and Slow, Dan Kahneman first said that System 1, the impulsive-reactive fast-brain, accounts for 90% of all our thinking. Hunt advised that as specialists/experts, we have to fight our System 1 brains all the time (as opposed to the deliberate, slower, conclusive and rational System 2) in order to be good at our jobs. He said that Trump-mania is an example of System 1 thinking and the thinking that we must feel more to buy more, because buying today is about seduction, not persuasion. It makes us rethink marketing research strategy. Rational thought leads to indecision, emotional thought leads to decision. “While emotion leads to action, reason leads to conclusions,” said neurologist Donald B Caine.

Arundati Dandapani@itadnura

Behaviorial science according to Alex Hunt (personal prediction) will see a Trump win in US @MRIAARIM @BrainJuicer

For market researchers, behavioral science is not/should not be rocket science. The challenge is that experts and specialists (e.g., judges, doctors) are just as susceptible to making System 1 decisions as others, and the challenge to researchers is thus in understanding that force to advance collective intelligence.

Combining Social Media Analysis and Traditional Survey Research

Margot Acton and Vanessa Killeen from TNS talked about the power of social media listening and combining it with traditional survey research. The brain has 11 million entry points. Success in social media research lies in how well you can web scrape and clean. Deciding on which traditional methodology could be combined with social media evaluation and model connections between social modelling and online behavior remains a key objective of social insight generation.

TNS has developed its fully commercialized predictive modelling of brand equity from social data. This model predicts a brand’s equity 8 to 12 weeks into the future and correlates at over 90% to results of brand equity surveys. Success in this area has led several large global clients to use social-derived data as a key, foundational element in their brand tracking, allowing them to ‘turn off’ other cumbersome, less predictive brand KPI’s from their traditional research programs. For instance, the beer market in Canada is highly fragmented. Craft brands flip the major brands in social media power, because of its more defined share of voice there than in traditional media—it’s not just about sample size anymore. Harnessing this combination of traditional and social research and investing in iterative modelling has thus helped TNS come up with brand equity insights that were otherwise not possible across categories.

The Automated Analyst: Why Jobs in MR are Under Threat from Smart Machines

Briana Brownell of Pure Strategy Inc., reflected on the promise of artificial intelligence and automation despite its perceived threat to human labour.

There have been three stages of technology in the workplace: human only, human-augmented technology and human-monitored technology. The five eras of automation have been: the eras of mechanics, force, interpreting (e.g., Watson and Jeopardy), practice (e.g., Sodol v. Alphago) and adapting. Self-driving cars represent this era of adapting. What does such evolution mean for market research? Machines are good at speed and diligence. However, there are jobs at which humans still excel, such as social intelligence, open-ended problem solving, creative and technical fields. AI journalism is the closest parallel to the research report, causing ripples of uncertainty among reporters. However, AI research (or any other research) will not go out of demand, it will just be done differently as researchers become partners who contribute to the bottom line than just “regular employees.” And, if automation takes away some jobs, it will also create new ones!

Putting the People back in Quantitative Research

Mike MacLeod of Lightspeed GMI chatted about the importance of voice and video in an age where video creation and consumption has exploded. “Kids don’t watch TV anymore, they’re watching PewDiePie and Miranda Simms,” he groaned. Research reveals that 64% of consumers are more likely to buy a product after watching a video about it. Moreover, every month approximately 5 million years’ worth of video is shared and created online. In an environment where surveys are changing, getting shorter and being completed on mobile devices and open-ended questions on mobile are fading, video is becoming the norm.

Benefits of video response include incorporating qualitative techniques into quant research, offering a more direct link between product users and creators and hearing respondents’ emotions, reactions and opinions first-hand. Many different platforms for video management and analytics exist, including Living Lens, 24tru, Vox Pop me, 2020 and bigsofa. Video responses are different from text responses, generating more and richer data.

MacLeod offered some guidelines for those who would like to include video/voice questions in their quant research:

  • Ask more personal questions
  • Specify what should be in the video
  • Tell them what you want to know
  • Make questions individualised
  • Straightforward/clear guidelines maximizes respondents’ quality of video (by 85%!)

A quarter of all respondents today will be willing to provide video, although higher incentives will be required. Video increases engagement, and within 5 years, a quarter of all quantitative projects will incorporate video capture, said MacLeod.

Facial Biometric Ad Testing – Research Without Asking a Single Question

Bernie Malinoff of element 54 and Norman Chang of the Ontario Lottery and Gaming Corporation led a panel on facial biometric ad testing and tracking emotions in consumer research through Real Eyes Facial Coding. Facial Coding is about going beyond the basic six emotions. Real Eyes R&D found that people are rarely angry with advertising and are more likely to be confused. Attraction— retention—engagement – impact was the range of emotional impact from ads tracked through facial coding or marketing research without asking questions. The methodology relies on the maxim that people who feel more, do more.

Norm Chang of the Ontario Lottery and Gaming Corp said that lottery is all about emotions. Functionally he just sells bits of paper that mean the world to consumers. The Lottery and Gaming Corporation strives to hit a unique emotional cord in its advertising. They invested in 10 ads using facial biometrics. It was found that a 30-second ad did better than the 15-second ad because people were more engrossed with the longer content. Chang argues that we’re still in the early days of ad testing through facial biometrics but are already gaining insights on its benefits. Facial biometric ad testing relies on people’s System 1 biases and is hard to argue with because people act/buy on emotional triggers and short thought. Facial biometrics is easy to launch, takes only a couple of days to collect the data and contrary to expectations, is not costly!

From Good to Great: How to get there

Evan Wood, SVP of Marketing and Custom Services at Environics Analytics was the final keynote speaker. His company was built around the idea of connecting data to small area geography and micro-targeting under the leadership of Jan Kestle. Their core tenet remains, “Get the right people on the bus, figure out where they sit later.” 80% of what they do today is based on geodemographics and segmentation. For any organization that believes in its prowess and aspires to scale up, Wood recommends the book Good to Great: Why Some Companies Make the leap and Others Don’t by Jim Collins. Wood shared 26 ingredients of Environics Analytics’ success—insights relevant to any corporate data-house. In this analytics age where data is a core differentiator, most organizations are already building analytical road maps. With newer cloud based intelligence solutions, thousands of variables (and data anxiety) and market research you have to be able to access different sources of information at given points of time for a true understanding of your consumers’ psychographically, behaviorally and demographically.

Big data has been distracting, but it has also propelled the industry forward, moving it from hype to productivity. Wood shared his company’s graph of innovation triggers, inflated expectations, trough of disillusionment, slope of enlightenment, plateau of productivity and narrated a growth strategy for every aspiring organization.

Net Gain 2016’s range of speakers, technologies and content were impressive. Behavioral science, big data, facial biometrics and automation are only the tip of the iceberg in new knowledge and methodologies of the future. Reflecting on these along with milestones passed offers a more holistic understanding of our business journey today.

Leave a Reply