We all go about our day harboring countless opinions about the world around us. Our thoughts may quickly jump from, "I think I'll get a burrito for lunch" to "Why am I always losing umbrellas?" to "I should see what movies are opening this weekend" in the span of seconds. Sometimes we even say them out loud.
All of this "big data" stored in our heads (some say about 2.5 petabytes) is gold to marketers, but only if they can get to it and find the relevant pieces. And they can't download our brains just yet. One of the biggest limits to consumer research is the finite number of questions we can ask people at any given time. Even if you could pose 100 or 200 questions (let's put aside the issue of survey fatigue -- and don't even think about trying to do this by phone), the respondents will only be answering questions about a narrow list of topics formed by even narrower human-based hypothesis.
Under typical survey constraints, how could you ever learn that the biggest fans of Ellen DeGeneres are two times more likely than non-fans to also like eating at IHOP? That's important data for IHOP (and maybe also Ellen) to know. Or that people who love Thai food are more likely than other cuisine fans to use Twitter and to be influenced by Internet ads vs. TV ads? Thai food restaurants and packaged food makers take heed. These are not questions that would intuitively go together.
The trick is to change the way we view surveys altogether. If you only ask a question or two at a time, more people will answer. Sample sizes of 500 or 1,000 can turn into 10,000 or even 50,000. And let's say you asked 1 Million people to answer two of a possible 1,000 questions. Each of the 1,000 questions would be answered 2,000 times. Then do that every day. Over time, you would have enough data to cross-tabulate all of those thousands of questions against each other and find correlations you would never expect. Advances in data mining technology could do all of that work for you.
In a new eBook called 101 Cool Consumer Insights, we pulled together a compilation of interesting and unexpected insights we found in our database of over 27 Million consumers and their 400+ million answers to 20,000 different questions. Here is my personal Top 5:
5: Despite similar formats, American cuisine casual dining chains have fan differences.
Most people may not expect to see major differences among the customers of casual restaurant "bar and grill" format brands like Chili's, Applebee's, TGI Fridays, and Ruby Tuesday. But as it turns out, there are real differences in their fan bases that set them apart. For example, TGI Fridays fans are more likely than the other brands' fans to be seriously into sports, and Chili's fans rank higher as market mavens, particularly for following music trends. Ruby Tuesday fans are more likely to try new products before others. All of them are more likely to be influenced by ads on TV than the general U.S. population.
Report date: March 19, 2014. Number of respondents analyzed: 188,055
4: The horror movie genre is least loved by hatchback drivers.
Auto makers love to cross-promote with Hollywood. When it comes to comparing car drivers to different movie genres, the data did reveal some obvious things, like people who drive a truck as their primary vehicle are much more likely to prefer Action flicks. Less obvious is that comedies rank highest among people who driver cross-over vehicles. Horror movies were globally the least favorite genre, with hatchback drivers disliking them the most (there's some guidance for an ad strategy that should NOT be pursued) -but horror films are actually most liked by people who don't have a car. So, naturally, think about plugging your next slasher movie on the bus.
Report date: April 29, 2014. Number of respondents analyzed: 9,634
3: The Starbucks happy hour may be ideal for job hunters.
As Starbucks rolls out its concept of an evening beer, wine and small plates menu in certain markets, consumers who say they are unhappy in their current jobs are nearly 3X more likely than the general population to want to check it out. Now, whether this is because they are more interested in all things alcohol in general, or see this as a relaxing environment for updating their resume, we don't know for sure. Nevertheless, the strong correlation here could come in handy to Starbucks or perhaps recruiters.
Report date: April 28, 2014. Number of respondents analyzed: 8,366
2. When it comes to sports marketing partnerships, who knew that a young female pop star would index so high?
Taylor Swift's popularity may rank highest among the younger female demo group (under age 24), but she's also quite liked among the overall population of sports fans. NBA fans are the most likely to also be fans of Ms. Swift, but NFL fans in general like her quite a bit too.
Report date: May 12, 2014. Number of respondents analyzed: 11,276 (NBA question) and 11,915 (NFL question)
1. Winning at "Rock, Paper, Scissors" may not just be about luck
Our data does show that certain consumer segments, based on their psychographic profiles, are more likely to pick either rock, paper, or scissors in this age-old game. Rock is the most popular selection in general, but cat people are more likely to pick scissors and Justin Bieber fans are more likely to pick paper. In fact, the list of rock/paper/scissors correlations is almost too long to name. But I need to keep a couple winning insights to myself.
Report date: April 29, 2014. Number of respondents analyzed: over 80,000 (to various questions cross-tabulated)
Connecting the dots of hundreds of millions of consumer opinions can uncover some revealing and often counter-intuitive insights. But to get there, we need stop treating people like guinea pigs that we bribe or hood-wink into completing lengthy, complex surveys. Instead, we just need to tap into some interesting opinions people want to share, combine them with millions of people like them, and let technology do the rest. How else could we ever know that Gary Busey fans are 288 percent more likely to love Kia cars? It's true.
Note: The survey results included in this article and the referenced eBook reflect the opinions of hundreds of thousands of people collected during the past 12 months. No clients, trademark owners, or celebrities have participated in conducting these surveys and they have not endorsed, approved, or otherwise provided any input regarding the survey results published here or in the eBook. The report date for the Thai cuisine fan data was April 28, 2014 with 23,448 respondents (Internet ads) and 14,767 respondents (Twitter data). The report date for the Ellen DeGeneres-IHOP data was May 12, 2014 with 8,928 respondents. All respondent data has been weighted for U.S. Census representativeness by gender and age, 13 years and older.