The posts here represent the opinions of CMB employees and guests—not necessarily the company as a whole. 

Subscribe to Email Updates


see all

The New Math of Media Disruption

Posted by Lynne Castronuovo

Tue, Mar 13, 2018



During February’s Winter Olympics, NBC tried to measure exactly how many people tuned in to watch the games. This sounds like standard practice—and it is—but this time NBC strayed from the traditional Nielsen system by combining broadcast and cable channel viewership with streaming platform audiences to produce a single number—attempting to account for multichannel consumption

NBC isn’t the only one struggling to get a valid headcount. As I was reminded at last month’s Media Insights and Engagement Conference: no one in the media industry seems to be able to measure the crowd anymore.

The media and entertainment industry has been upended. Viewers are spread far and wide across devices, platforms, and time—no longer huddled around a single television set to watch primetime. We’ve said goodbye to the days of the standard 18-49 year old viewer group.

Further, the explosion of high quality, award-winning content from nontraditional producers like Hulu has fragmented audiences with niche tastes and demographics. There’s programming for nearly every interest.

Conversely, the meteoric rise of programs like Stranger Things underscores the emergence of programs that are beloved by a blend of demographics—from parents to young teens—making it increasingly challenging for advertisers to know what will resonate with such diverse audiences.

From splintering audiences to multichannel consumption, the disruption within the media industry is coming from all sides. It’s become harder for broadcasters to know who and how their content is being consumed and for advertisers to measure the ROI of ad spend. Data is coming in from a variety of sometimes incongruent sources, so it can be challenging to get the full picture.

The media industry needs researchers now more than ever to help uncover who, how, and why content is being consumed. Understanding the who, how, and why is critical for creating content and advertising that will resonate most with viewers and ensure advertisers are targeting  and reaching the right audience.

The changing media and entertainment landscape is daunting, but this is a tremendous opportunity for market researchers to innovate and rise to meet these new challenges. We can’t rely solely on traditional audience tracking methods—we need to dig deeper into the consumer psyche understand how media is being consumed.

The most successful researchers will be those who can balance the art and science of collecting insights—those who can parse vast amounts of data and stitch together a holistic story. I welcome these new challenges within the media and entertainment industry and encourage other researchers to help our clients face them head on.


Topics: digital media and entertainment research

Beer, Pot, Car Racing and More: A brief roundup of Quirks East 2018

Posted by Julie Kurd

Mon, Mar 05, 2018

quirks east poster3.png 

Here are a few learnings from last week's Quirk’s East Event in Brooklyn:

Measuring the success of Corona’s new nonlinear ads. Samrat Samran from AB InBev and Pranav Yadav from Neuro-Insight shared Corona’s advertisements that focus on the lime ritual. These “story fragments” equate the 'feeling' you get when a lime goes into the beer bottle with a surfer plunging into the water. To quantify success of a non-linear advertisement, they established guardrails on five key branding moments (brand memory, emotional intensity, and engagement etc.). Through this framework, they were able to see an increased recall on the second ad view because interestingly, a nonlinear story is challenging enough that the brain seizes on new aspects of the story in the second viewing.

Legal Cannabis is a Brand Innovation Game ChangerIn a time where many Fortune 500 companies are still asking for drug tests, it may not be intuitive to account for cannabis in your innovation pipeline—especially if you don’t work in the cannabis industry. But marijuana pairings are occurring beyond the music, snacks, etc., so it’s time to start paying attention to this growing category. In the BDS Analytics presentation, of those studied (28% were users, 34% were acceptors, and 38% were rejecters), almost everyone (including rejecters) universally accept some form of marijuana use as ‘acceptable’. And in this case, rejecters aren’t necessarily opponents, they just choose not to use. The cannabis market is diverse in generation, gender, and motivations—and likely will continue to grow in complexity.

NASCAR’s Passive Metering and Digital Media Tracking. NASCAR’s Norris Scott and Luth’s Candice Rab spoke about their behavior-based insights research. In the study, respondents downloaded an app that passively tracked behavior across devices—from PC, smartphone, and tablet. Integrating digital data with survey research helped contextualize participants’ behavior and shed light on the “why” behind attitudes and consumption of digital sports media among NASCAR fans and super-fans. Through this approach, NASCAR discovered that fans are also looking at a range of other sports content, such as ESPN, Yahoo Sports, etc., and half the fans use digital to enhance their race viewing experience.   

Gen Z (Tweens, Teens) and their Secret (Visual) Languages. Kids have always loved having secret languages that bonds and empowers them. Writing has given way to typing to tapping to snapping, per Stephanie Retblatt of SmartyPants. Text has morphed into videos, and videos to emojis, GIFs, memes, filters, and stickers. This evolution marks the significant shift in how kids and tweens experience emotions in ways that text hasn’t kept up with. “Animoji” and “Bitmoji” are part of a new visual curation brought to us by Snapchat.

Changing role of Artificial Intelligence in research. Whether you’re a F500 company or a Consumer Insights firm, Peter Mackey of Wizer says we need to take AI seriously. Peter showed how a chasm is starting to build between client reality (speed over quality, tighter budgets) and traditional consumer insights. In traditional insights, qualified thinkers brainstorm with you, frame the challenge, and craft the research design—all of which is time intensive. These days, budget-conscious marketers have resorted to DIY surveys, templated survey automation, human supervised/semi-automated coding and transcription. Today, we’re all experimenting with AI in market research world (#MRX):

  • Input – Automating the survey
  • Data Processing – quantitative and qualitative quality control such as respondent fraud detection algorithms
  • Output – positive findings in green and negative in red.  Natural language interpretation.
Co-Creation and the Future of Loyalty & Rewards with the Hilton. To break out of the “sea of sameness”, executives and consumers can come together to ideate innovative solutions using a co-creation approach. My company, CMB, presented with our client, Jessica Boothe of Hilton, on a recent co-creation session that explored the future of Global Loyalty and Rewards programs. Hilton has renovated its rewards program, simplifying both the earn and the burn aspects of rewards. This co-creation initiative discovered and re-discovered potential new emotional and functional rewards for both elite and non-elite members of the Honors program.      

Whether you always have a travel bag to unpack, or you ‘never get to go anywhere,’ learning is always available to everyone. Sign up for our upcoming webinar on Wednesday, 3/7 at 12pm ET.

Register Now

Topics: conference recap

Employee Appreciation: The Importance of Providing Emotional Benefits

Posted by Heather Magaw

Wed, Feb 28, 2018


An organization’s people are its most valuable assets.

And in today’s job competitive market, companies finding and retaining top talent need to go beyond “benefits” like ping-pong tables and yoga classes. These tangible perks look great, but on their own they won’t foster employee loyalty and motivate productivity.

A key to a corporate culture that inspires and motivates employees is ongoing appreciation—showing gratitude each and every day. Providing emotional benefits (e.g., feeling appreciated and valued) is one of the most important things a company can do for its employees—in addition to providing functional benefits (e.g., free lunches).

But identifying what employees truly value and what makes them feel appreciated can be challenging. It requires a thoughtful approach to understanding human behavior and acknowledging our intrinsic desire to be recognized, celebrated, and appreciated every day.

At CMB, we found that our employees feel more appreciated by intangible, personal gestures like:

  • Receiving an email of appreciation from a client
  • Finding a “thank you” post-it from a colleague stuck to the desk
  • Getting a handwritten thank you note in your company mailbox
  • Seeing an email of acknowledgement to a manager about an employee’s unique contribution

These small acts of kindness and appreciation can speak louder than a free lunch or Summer Fridays. They are thoughtful, meaningful, and make employees truly feel valued for the work they do.

Springing for a midafternoon ice cream party is a lot easier than encouraging busy colleagues to take the time to write personal notes. But, I challenge leadership teams to foster workplace environments that practice ongoing appreciation. As Stephen R. Covey once said, “Always treat your employees exactly as you want them to treat your best customers.”

In an upcoming webinar, join Erica Carranza, PhD., and learn how building meaningful connections promotes workplace satisfaction and productivity.

Register Now

Heather Magaw is VP, Client Services at CMB. She challenges each reader to write 5 emails or notes of appreciation on Friday, March 2, Employee Appreciation Day.


Topics: our people, emotional measurement, emotion

Predicting Olympic Gold

Posted by Jen Golden

Wed, Feb 21, 2018


From dangerous winds and curling scandals to wardrobe malfunctions, there’s been no shortage of attention-grabbing headlines at the 2018 Winter Olympics.

And for ardent supporters of Team USA, the big story is America’s lagging medal count. We’re over halfway through the games, and currently the US sits in fifth place behind Norway, Germany, Canada, and the Netherlands.

Based on last week’s performance (and Mikaela Shiffrin’s recent withdrawal from the women’s downhill event), it’s hard to know for sure how America will place. However, we can use predictive analytics to determine the main predictors of medal count to anticipate which countries will generally be on the podium.

We’ll use TreeNet modeling to identify the main drivers of medal count based on previous Winter Olympics outcomes. For the sake of simplicity, we’ll focus on the 2014 Sochi winter games (excluding all Russia data which would skew the model!) From there, we can infer similarities between medal drivers for Sochi and PyeongChang.

Please note all these results are hypothetical, and not reflective of actual data!

To successfully run a TreeNet analysis, you need both a dependent variable (e.g., the outcome you are trying to predict) and independent variables (e.g., the input that could be possible predictors of the dependent variable).

In this case…

Dependent variable: Total 2014 Sochi Winter Games medal count
Independent variables (including data both directly related to the Olympics and otherwise):

  • Medal count at the Vancouver Olympic games
  • Medal count at previous Winter Games (all time)
  • Number of athletes participating
  • Number of events participating in
  • Number of outdoor events participating in (e.g., downhill skiing, bobsled)
  • Number of indoor events participating in (e.g., figure skating, curling)
  • Average country temperature
  • Average country yearly snowfall
  • Country population
  • Country GDP per capita

The Results!

Our model shows the relative importance of each variable calibrated to a 100-point scale. The most important variable is assigned a score of 100 while all other variables are scaled relative to that:

Olympic Medal Predictors.png

Meaning, in this sample output, previous medal history is the top predictor of Olympic medal outcome with a score of 100 while # in outdoor events and indoor events participating in are the least predictive.

This is a fun and simple example of how we could use TreeNet to forecast the Winter Olympic medal count. But, we also leverage this same technique to help clients predict the outcomes of some of their most complex and challenging questions. We can help predict things like consideration, satisfaction or purchase intent for example, and use the model to point to which levers can be pulled to help improve the outcome.  

Jen is a Sr. Project Manager at CMB who was a spectator at the Sochi winter games and wishes she was in PyeongChang right now.

Topics: advanced analytics, predictive analytics

Relatability, Desirability and Finding the Perfect Match

Posted by Dr. Jay Weiner

Tue, Feb 13, 2018


Dear Dr. Jay:

It’s Valentine’s Day, how do I find the one ones that are right for me?

-Allison N.

Dear Allison,

In our pursuit of love, we’re often reminded to keep an open mind and that looks aren’t everything.

This axiom also applies to lovestruck marketers looking for the perfect customer. Often, we focus on consumer demographics, but let’s see what happens when we dig below the surface.

For example, let’s consider two men who:

  • Were born in 1948
  • Grew up in England
  • Are on their Second Marriage
  • Have 2 children
  • Are Successful in Business
  • Are Wealthy
  • Live in a Castle
  • Winter in the Alps
  • Like Dogs

On paper these men sound like they’d have very similar tastes in products and services–they are the same age, nationality, and have common interests. But when you learn who these men are, you might think differently.

The men I profiled are the Prince of Darkness, Ozzy Osbourne, and Prince Charles of Wales. While both men sport regal titles and an affinity for canines, they are very different individuals.

Now let’s consider two restaurants. Based on proprietary self-funded research, we discovered that both restaurants’ typical customers are considered Sporty, Athletic, Confident, Self-assured, Social, Outgoing, Funny, Entertaining, Relaxed, Easy-going, Fun-loving, and Joyful. Their top interests include: Entertainment (e.g., movies, TV) and dining out. Demographically their customers are predominately single, middle-aged men.

One is Buffalo Wild Wings, the other, Hooters. Both seem to appeal to the same group of consumers and would potentially be good candidates for cross-promotions—maybe even an acquisition.

What could we have done to help distinguish between them? Perhaps a more robust attitudinal battery of items or interests would have helped. 

Or, we could look through a social identity lens.

We found that in addition to assessing customer clarity, measuring relatability and desirability can help differentiate brands:

  • Relatability: How much do you have in common with the kind of person who typically uses Brand X?
  • Social Desirability: How interested would you be in making friends with the kind of person who typically uses Brand X?

When we looked at the scores on these two dimensions, we saw that Buffalo Wild Wings scores higher than Hooters:BWW vs hooters1.png

Meaning, while the typical Buffalo Wild Wings customer is demographically like a typical Hooters customer, the typical Hooters customer is less relatable and socially desirable.  This isn’t necessarily bad news for Hooters–it simply means that it has a more targeted niche appeal than Buffalo Wild Wings. 

The main point is that it helps to look beyond demographics and understand identity—who finds you relatable and desirable. As we see in the Buffalo Wild Wings and Hooters example, digging deeper into the dimensions of social identity can uncover more nuanced niches within a target audience—potentially uncovering your “perfect match”. 

Topics: Dear Dr. Jay, Identity, consumer psychology

The Anchoring Effect—Avoiding Bias in Market Research

Posted by Hannah Russell

Thu, Feb 08, 2018


Consider the following questions:

  1. Did Thomas Edison patent more or fewer than 7,000 inventions?
  2. To the best of your ability, estimate the number of inventions patented by Edison.

Unless you retained your 7th grade social studies knowledge, you’d probably have a tough time answering. But based on the context given in question 1, you may guess somewhere in the several thousand range. Why is that?

As Nobel Prize winning author Daniel Kahneman explains, this is an example of the anchoring effect—a cognitive bias in which humans tend to rely on the first piece of information offered (the “anchor”) when making decisions. The detail (e.g. number) we automatically “anchor” to then influences subsequent decisions.

So, in the Edison example, according to the anchoring effect, the “7,000” in the first question impacted your answer to the second question.

Consider then, if instead the first question had been: “Did Thomas Edison patent more or fewer than 100 inventions?” Your answer to the second question would likely be a lot less than what it had been in the first scenario.

Our perceptions are often influenced by the stimuli we are exposed to—both consciously and subconsciously. Sometimes, though, this information is useless in helping us make correct judgements (such as the number 7,000 in the example above). Even if we’re aware of an external influence, it can be hard to discount.

As market researchers, we have an obligation to manage and mitigate this type of bias to preserve the integrity of our data.

When creating a survey, we try to avoid anchoring respondents in a particular number (or other pieces of information) and are careful in the way that we order questions. If we’re exposing respondents to various numbers (which is often the case in pricing research), we rely heavily on analytical techniques that ensure randomization and exposure to multiple scenarios.

Ultimately, we can’t avoid priming effects altogether—there is no such thing has 100% unbiased data. But, we need to keep these psychological biases in mind when designing, implementing and presenting data. By recognizing the downstream consequences of something like the anchoring effect, we’re better positioned to find truthful and actionable insights for clients.

Hannah Russell is a Project Manager at CMB who indeed retained her seventh grade social studies knowledge. Thomas Edison accumulated 2,232 patents worldwide, 1,093 of which were in the US.”

Topics: consumer insights, research design, consumer psychology

Begin with the End—Lessons Learned

Posted by Caitlin Dailey

Fri, Feb 02, 2018


A former colleague of mine had a post-it note on his wall that read: You have as many hours in the day as Beyoncé. Inspiring words, but for those of us in professional services (rather than entertainment) it can feel like the to-do lists never end.

I recently watched a webinar on productivity given by MIT Sloan School of Management Senior Lecturer Robert Pozen. There was a lot of useful information in the webinar, but one piece of advice really resonated: begin with the end.

“Beginning with the end” means letting your desired outcome drive the planning and execution of your task. If you are cognizant of what your end-goal is, it will make tackling projects of any scope a lot easier—whether that’s writing an email or the final report of a multi-phased segmentation study.

At CMB, we always begin with the end in mind. When kicking off a project, we meet with key client stakeholders to align on business and research objectives. We leverage our proprietary Business Decision tools to identify what the desired business objectives are, and use that information to inform our research objectives and design. This preliminary decision-focused conversation ensures the research solution, story, and results are actionable and will deliver meaningful outcomes with true business impact.

Once the project is kicked off, no time should be wasted—consider building out a narrative and recording tentative conclusions as soon as data starts coming in. It can be tremendously helpful to have a mid-field check of the data to revise those conclusions, and then do a final revision once you have all the data. The story might not change much during this time, but writing and revising your conclusions prior to the close of an initiative can make delivering the final report less stressful.

Particularly in market research, there’s pressure to deliver results faster than ever. When you start with the end in mind, you can be building out the story in an iterative process, rather than scrambling to at the end. Since unearthing a clear and meaningful story is one of the most important pieces of a project, you’re only helping yourself (and your colleagues) by beginning with the end.

Other ways to improve productivity

As I mentioned, there were loads of other useful tips from Pozen’s webinar on how to increase productivity:

  • Write down your daily goals: Rome wasn’t built in a day, so jot down objectives you can realistically accomplish today.
  • Don’t exhaust your schedule: Avoid scheduling every minute of your day. Having a calendar filled with meetings may look productive, but it’s important to include “thinking time” for yourself.
  • Include work and non-work tasks: Your list should include routine essentials like going to the gym or having dinner with your family. This will help maintain a healthy work/life balance and will give you time to “recharge”.
  • Manage your inbox: If you’re in the zone, don’t feel pressured to stop and respond to each email immediately (unless it’s urgent, of course). Instead, set aside time a little later to respond to all emails.
  • Let go of perfectionism: Do you reread an email 5 times before you hit send? Scan through a deck repeatedly? Chances are, it were ready to go after the second review, so save your mental energy for something else and move on.
  • Quit procrastinating: One of the biggest hurdles to getting things done is simply starting them.

I’m now being mindful of how I can incorporate these practices into my life to maximize my productivity, and in turn, hope to tip the scale of my work/life balance in favor of a more stress-free work week. I hope you can too!

Caitlin Dailey is a Senior Project Manager on the Financial Services, Insurance, and Healthcare team at CMB and is looking forward to trying out these tactics to help get her out of the office a little earlier in 2018.

Topics: methodology, business decisions, research design

Competing on Image Isn't Enough: Why and How to Make Your Brand an Expression of Identity

Posted by Dr. Erica Carranza

Wed, Jan 24, 2018

girl with coffee.jpeg

Brand image matters.

For marketers, that’s a truism—and for good reason. Brand image does matter. I see evidence of it every day, in the work we do at CMB, uncovering insights that help brands craft winning strategies. We spend a lot of time helping our clients decide how to fine-tune their brand image, market it effectively, support it with products and customer experiences, and track their progress. Some brands have been wildly successful in their pursuit of a brand image that has helped them maintain a competitive edge (e.g., Disney is “magical,” Apple is “innovative,” Walmart is “affordable”).

But the traditional focus on brand image hasn’t kept-up with people’s lives.

We live in a world where people are inundated with options. Are you looking for something to eat? Something to watch? Something to wear? Whatever it is, rest assured you’ll have lots of possibilities. Even something as mundane as shampoo yields over 100,000 hits on Amazon. It’s gotten to the point where scientists are studying the effects of “too much choice” on our wellbeing.

In a market this saturated, competing on brand image is no longer enough.

Most brands already strive to communicate a positive brand image and a well-defined set of brand benefits. In every industry, many brands are vying for the same customers and claiming the same (or similar) attributes. People are quick to say that Apple is “innovative”—but they say the same thing about Samsung. So, when they’re choosing their next smartphone, “innovative” won’t be a deciding factor.

Furthermore, competing on brand benefits (like service, cost, and convenience) isn’t always practical. I witnessed that firsthand in my time at American Express. Great customer service and Membership Rewards were once part of a unique value proposition. But, nowadays, card benefits offered by one brand are quickly copied by others, and the industry is stuck in a “race to the bottom.” In their efforts to beat competitors and increase share, brands are undercutting profitability to offer ever richer card rewards.

What’s a brand to do in a world where it’s gotten this hard to compete on brand image and benefits?

The answer: Compete on brand tribe.

People love brands that help them express their identities. And, thanks to the explosion of options for consumers, every choice is now a chance to express who we are.

Yet decades of scientific research have shown that our identities are social—they are shaped by our social groups, norms, and connections. Who we are depends on our real and aspirational relationships with other people. So truly strategic brands lead people to equate using the brand with joining a tribe that expresses an identity. And the secret to creating that connection is a clear, compelling brand customer image. After all, brands aren’t people. But brand customers are.

Your brand’s customer image is the mental picture people have of the kind of person who typically buys or uses your brand. It’s related to brand image, but it’s not the same. To take one of my favorite examples, consider Subaru. When we ask people to describe the brand Subaru, they say “safe” and “reliable.” But when we ask them to describe the typical Subaru owner, they say “middleclass,” “family-focused,” and “outdoorsy.” They picture someone with kids and a dog, who likes to hike, and who supported Bernie Sanders in the 2016 presidential primaries. There’s a lot of nuance to their image of the typical Subaru customer—including attributes a person can embody, but a brand cannot.

Of course no image of your brand’s typical customer will truly capture your actual customer base. Your brand’s customer image is more like a stereotype: A set of overgeneralized assumptions about typical members of your brand tribe. But people tend to rely on stereotypes—often unconsciously—in order to navigate our complex world. Accordingly, brand customer image has powerful effects on consumer behavior.

For example, at CMB we’ve found that:

  • When people identify with their image of a brand customer, they are 14-times more likely to choose that brand, and 15-times more likely to recommend it.
  • As predictors of brand engagement, our measures of identification with the perceived customer routinely beat perceptions of the brand—even on dimensions as important as quality, price, value, service, convenience, authenticity, reputability, and innovation.

Taking all this into account, it’s no surprise that many of the most iconic ad campaigns have invoked a clear, compelling customer image. Remember “I’m a Mac / I’m a PC”? Dove’s “Real Beauty”? Or the insidiously cliquey “Choosy moms choose Jif”?

To effectively compete on brand tribe, make sure that you have answers to these three questions:

  1. What is your brand’s current customer image? Does your target audience already have an image of the kind of person who uses the brand? If so, how clear is it? What attributes define that image (e.g., what demographics, motives, and values)? And what (if anything) makes it unique compared to competitors’ customer images?
  1. How compelling is that image? Is it an image of a person your target audience can relate to? Is it a kind of person they know and like, or would like to know? Does it represent an “ingroup” or an “outgroup” tribe—and how appealing is it compared to their images of competitor brand tribes?
  1. How can you optimize that image? What’s working about the image, and what isn’t? Which assumptions should you reinforce—and which should you work to change—to own a customer image that is compelling and unique for your audience, and realistically attainable for your brand?

If we want to influence consumer behavior, we must remember that consumers are people, and that people are social animals. Show them a group that they want to belong to, and they’ll adopt the attitudes and behaviors they believe to be normative (i.e., typical) for that group—including choosing the same brand.

Yet most brands today are not leveraging this powerful insight in a truly disciplined, quantitatively-validated, systematic way. 

And in the current competitive context—across industries—it’s more important than ever. Brands assume that consumers are asking themselves, “What brand do I want to use?” But, at a deeper and more decisive level, they are really asking: “Who do I want to be? Do I want to be the kind of person who uses this brand?”

Interested in learning more about how CMB leverages consumer psychology, advanced analytics, and market strategy to help clients build customer-centric brands? Watch out latest webinar on BrandFx and the three critical pieces to the brand engagement puzzle:

Watch now!

Topics: brand health and positioning, Identity, AffinID

CES 2018: Virtual Assistant Battle Royale

Posted by Savannah House

Wed, Jan 17, 2018


Last week the 2018 Consumer Electronics Show (CES) wrapped up in Las Vegas and left us feeling excited and invigorated about what’s to come in tech. From talking toilets to snuggle robots, CES 2018 was yet another reminder of how deeply technology has infiltrated every aspect of our lives.

This year, once the world’s largest tech show found its way out of the dark, CES was all about virtual assistants.

Alexa vs. Google Assistant

Amazon’s Alexa has dominated the virtual assistant category—claiming 70% of the market share in 2017 and then ending the year with strong holiday sales as the most downloaded app for Apple and Android on Christmas Day. But this year, Google (who typically keeps a low profile at CES), made its presence loud and clear.

From wrapping the Las Vegas monorail with the words “Hey Google” to erecting a massive playground in the CES conference center parking lot (complete with a giant gumball machine), Google is making it clear that it intends for Google Assistant to be a legitimate contender in the virtual assistant space.

It’s about integration, not separation

Both Google and Amazon used CES 2018 as a platform to announce new partnerships for their virtual assistants. Alexa will soon be found in Toyota cars, Vuzix smart glasses, and Kohler smart toilets. Meanwhile, Google is integrating its smart technology with a slew of products from leading brands like Sony, Lenovo, and Huawei.

If there’s one takeaway from these partnership announcements, it’s that voice assistant technology will not be confined to the realm of their makers’ product lines. Instead, voice assistants intend to be everywhere—plugging into smart glasses, smart earbuds, and smart toilets—underscoring the tech industry’s expectation that voice assistants will continue to play a much bigger role in our digital lives.

Crossing the chasm

It appears Google’s goal at CES wasn’t necessarily to woo tech lovers with its Google Assistant. Rather, it was to show regular people what is possible with virtual assistant technology. This is important because it demonstrates the (potential) ubiquity of this category once thought of as only for early tech adopters.

However, despite pushes to show “regular" people that virtual assistants are meant for everyone, our research indicates that social identity is playing a role in preventing widespread virtual assistant adoption.

As the chart indicates below, peoples' ability to relate to the typical user is the biggest driver in virtual assistant usage:

VA drivers (branded)-1.jpg

However, currently, consumers can’t relate to the typical virtual assistant user, which is keeping them from “crossing the chasm” and becoming regular users themselves.

The virtual assistant category will only grow in complexity as more companies enter the game (let’s not forget about Siri and Cortana). But, while flashy conference displays, exciting partnership announcements, and product demos are all helpful in attracting more consumers, if virtual assistant brands want to achieve more mainstream adoption, the brand and creative teams need to tackle the virtual assistant image problem head on.

Savannah House is the Marketing Manager at CMB, and as a light sleeper, is most excited about the robotic pillow.

Topics: technology research, internet of things, Identity, AffinID, Artificial Intelligence

AI's Image Problem: Who's the "Typical" Virtual Assistant User?

Posted by Chris Neal

Tue, Jan 09, 2018


Every nascent technology and every tech start-up faces the same marketing challenge of “crossing the chasm” into mainstream adoption.  Geoffrey Moore framed this very well in his 1991 classic, “Crossing the Chasm”:

adoption curve.pngWord of mouth can play a huge role in motivating certain segments to sip the Kool-Aid and make the leap.

With CES 2018—the world's largest gadget tradeshow—happening in Vegas this week, I can't help but wonder if mainstream consumers don’t relate to the early adopters of a new technology? What if they think it’s used by people who aren’t part of “their tribe”? Will it prevent them even considering the new tech? There are countless technology categories that have faced this challenge, for example:

  • certain gaming categories trying to expand beyond 15-24-year-old males
  • consumer robot products to this day
  • social media when it was first introduced
  • Second Life and other virtual worlds

I hypothesized that the virtual assistant (VA) category—and specific brands within it—faces this challenge. Yes, many people have tried and used Siri, but few mainstream consumers are truly using virtual assistants for anything beyond basic hands-free web-queries. To further complicate things, an increasing number of “smart home” products that connect to intelligent wireless speakers in the home (e.g., Amazon Alexa, Google Home, Apple’s forthcoming HomePod) are proving divisive. Some people love the experience or the idea of commanding a smart device while others categorically reject the concept. 

My team and I had the chance to test out a few hypothesis through our Consumer Pulse program and —voila!—we’ve got some tasty (and useful) morsels to share with you about how social identity is influencing consumer adoption in the virtual assistant space using our proprietary AffinIDSM solution.

Here’s what we found:

Social identity matters in the virtual assistant space. We studied US consumers (18+)—covering usage, adoption, and perceptions of the virtual assistant category and a deep-dive on four major brands within it: Apple’s Siri, Amazon Alexa, Google Assistant, and Cortana by Microsoft. We covered rational perceptions of the category, emotional reactions to experiences using virtual assistants, and perceptions of the “typical” user of Siri, Alexa, Google Assistant, and Cortana.

We then ran fancy math™ on our data to create a model to predict the likelihood of a virtual assistant “category rejecter” (i.e., someone who has never tried a VA before) to try any one of those assistants in the future. Our analysis indicates that how much a current VA category rejecter relates to their image of the type of person who uses a virtual assistant is the number one predictor of whether they are likely to try the technology in the future:


Unfortunately for the industry, category rejecters do not find the typical VA user very relatable. 
AffinID metric by brand.png

As the chart indicates, relatability (biggest predictor of likelihood to try as shown previously) scores the lowest of the three components of AffinID: relatability, clarity, and desirability. You may ask yourself: “are scores of 12 to 14 ‘good’ or ‘bad’?  They’re bad: trust me. We’ve now run AffinID on hundreds of brands across dozens of industries, so we have a formidable normative database against which to compare brands. The VA category does not fare well on “relatability,” and it matters.

Some brands’ VA ads, while amusing, are not very relatable to “normal” mainstream consumers. For example as my colleague Erica Carranza points out in her recent blog, Siri’s ad featuring Dwayne “The Rock” Johnson doing impossibly awesome things in one day (including taking a selfie from outer-space) with the help of Siri isn’t exactly a “normal” person’s day. A-grade for amusement on this one, but it is playing into an existing perception problem.

Stereotypes about users’ age and income are currently keeping “rejecters” away from the virtual assistant category.

The age gap between rejecters and “typical” virtual assistant users is a social identity construct keeping rejecters out of the category. Current rejecters, not surprisingly, skew older while current heavy VA users, also not surprisingly, skew young.

We uncovered this disconnect with a big predictive model using “match analysis” on a variety of demographic, personality, and interest attributes. For every attribute, we examined whether there was a “match” or a “disconnect” between how a rejecter described themselves vs. how they perceived the typical user of a virtual assistant brand.

The two specific perceptions that had the greatest ability to predict a rejecter’s likelihood to consider using a brand in the future was an age-range match and an income-range match. For example, if I’m over 35 years old (hypothetically!), and I perceive the “typical” user to be under 35 years old and higher-income than me…so what? Well, it does matter. For new technologies to achieve mainstream adoption, they must debunk the widespread perceptions that the early adopter is “young” and highly affluent, and that their product can be used by everyone (think: Facebook). SNL pokes fun at this generational discrepancy.

But in all seriousness, if a virtual assistant brand wants to achieve more mainstream adoption among older demographics, the brand gurus and creative teams working on campaigns need to tackle this head on.

And they must try to do this—ideally—without alienating the original early adopter group that made them their first million (think: Facebook, again…how many Gen Zers do you know who actually use it actively?). I—prototypical 45-year-old suburban dad—can’t imagine using Snapchat, for instance. If Snapchat wanted to get me and my tribe to buy in as avid users*, it needs to convince me that Snapchat isn’t just for teens and early twenty-somethings. Or it needs to launch a different brand/product targeted specifically at my tribe, and market it appropriately.

It’s worth noting there are other social identity constructs that help predict whether a non-user of a virtual assistant is likely to try a product in the future. For instance, the few VA category rejecters who perceive the typical (young, affluent) user as being as “responsible/reliable” as themselves are more open to trying a VA in future than those who do not perceive VA users this way. So, we’re seeing this stereotype that virtual assistant products are for young, affluent professionals living in a major coastal city with no kids to contend with yet, and this is turning some consumer segments off from trying out the category in earnest.  

Stay tuned to this channel for more on our study of the virtual assistant category. I’ll be covering some key insights we got by applying our emotional impact analysis—EMPACT℠to the same issue of what virtual assistant brands should be doing to achieve further adoption and more mainstream usage of their products. 

*I am more than 95% confident that the Snapchat brand gurus do not want me as an avid user…and my ‘tween daughter would definitely die of embarrassment if I ever joined that particular platform and tried to communicate with her that way.


Topics: technology research, EMPACT, Consumer Pulse, AffinID, Artificial Intelligence