How Trend Hunter uses big data to forecast the next big thing

And which big brands are using the site to find out what the cool kids are doing

Murad Hemmadi for Canadian Business January 12, 2015

It’s a little after 9 a.m. on the Friday before Thanksgiving, and Jeremy Gutsche is holding court in his Toronto office, talking about cupcakes. Bacon-flavoured cupcakes, to be specific, and how restaurants can use them to generate interest in their brands. The meaty-sweet treats are among the thousands of wacky ideas, products and fads catalogued on Trend Hunter, the website Gutsche launched in 2006.

At 36, Gutsche is quoted regularly as an expert on such topics by the likes of CNN, the Guardian and the Globe and Mail. He’s the author of two books, including the forthcoming Better and Faster. His first, Exploiting Chaos, was named one of the Best Books for Business Owners by Inc. magazine. “I like to say it’s half pictures, so you know it’s good!” he quips, fanning the pages to demonstrate.

On this day, Gutsche has just returned from a speaking tour that took him to Atlantic City, Dallas and New York — a mere fraction of the 60-odd keynotes he gives each year. One CEO described him as “an intellectual can of Red Bull.” And there’s not a hint of fatigue in his voice as he preaches his vision for changing the way companies find new ideas. Trend Hunter, he explains, is more than a website: It’s a market research firm for the digital age, with a data-driven model that contrasts with more traditional “cool gurus,” who identify trends and dispense insightful pronouncements based solely on observation and “gut instinct.”

The company’s Soho Street office is located a short distance from Toronto’s hip Queen West shopping strip and has the exposed brick and informal vibe of a startup. Candy apple – red walls and couches abound. At long wooden tables, eight millennials are clicking away at laptops, well on their way to crafting the 85 or so posts added to the website each day. Browsers open, earphones in, they sweep the corners of the internet for concepts and products that might appeal to the roughly three million web surfers who land on Trend Hunter’s articles each month. Once a team member finds something worth sharing — Ron Burgundy–branded holiday tree baubles, perhaps, or a photo gallery of a light festival in Prague — the concept is quickly written up in a 120-word article to add to the 250,000-plus that already populate the website. It gets a formula-following title (“Famous Reporter Christmas Ornaments” or “Spectacular Luminous Events”) and 10 hand-selected links to push visitors further down the site’s rabbit hole of weird, buzzy content. A global network of contributors and tipsters ensures that few ideas, however outlandish, escape Trend Hunter’s notice.

A monitor in the centre of the room acts as a digital scoreboard, ranking staff in real time on the quantity of articles they’ve written, the amount of online traction their work is getting and how much it’s being shared on social media. At 9 a.m., only a couple of names appear on the display; by noon the number of contributor avatars and their associated progress bars has ballooned.

Trend Hunter’s readers are a captivated bunch, looking at an average of 20 articles per visit — a staggeringly high number for any website. To those eyeballs, the site appears not unlike fellow click-mills Buzzfeed and Mashable. But every article view and click on Trend Hunter’s litany of listicles and galleries is fuel for its data machine. “When you’re running thousands of people through these articles and tracking what they’re choosing next, you start to get a sense of preference,” says Gutsche. The company uses the data it gleans to compile what it calls Pro Trends, sets of related concepts and ideas that Gutsche believes expose gaps in the consumer market. That Ron Burgundy Christmas ornament isn’t just a novelty purchase — it’s an example of “modernized tradition,” which, according to Trend Hunter’s analysis, is an indication that younger consumers don’t much care for the rigid formalities of the holiday season.

It’s these kinds of insights that Trend Hunter’s clients — a roster of major brands that includes Samsung, Kellogg’s, Crayola and Nestlé — are paying for. For a starting price of $24,000 a year, subscribers receive a customized monthly report from an analyst detailing Trend Hunter’s findings. Brands can use their reports to answer specific questions (“What’s new in holiday products this week?”) or find inspiration (“We’re launching a line of winter clothing, so show me examples of loud sweaters”).

Figuring out what customers want is a $21-billion industry in the U.S., according to IBISWorld, and many major brands solicit advice from polling firms such as Ipsos Reid or the legacy research arms of advertising agencies like JWTIntelligence to help them spot the next big thing before it arrives. More and more, brands are turning to big data to find market insights, in the hope that it will yield more scientific and more granular views into consumer behaviour. But just as it takes a master craftsperson to cut, polish and give a rough diamond its shape, raw data ultimately requires a human being to make it shine. Data might add scientific sparkle to the business of finding the next big thing, but there’s still an art to telling us what the numbers mean. And that too is what companies pay people like Gutsche to do.

Trend Hunter is privately held, and Gutsche won’t disclose revenues, but he claims he’s had “eight-figure” takeover offers. “A lot of people were interested in Trend Hunter for the eyeballs — as a media site,” he says. “But I believe the power of the research model is much more interesting long term.” He claims that replicating his massive experiment in consumer preferences would cost about $30 million. Gutsche is betting on a revolution in the way brands find out what the cool kids are doing.
The business of trying to divine customers’ desires has been around almost as long as there have been customers, but market research as an industry began with the advent of modern magazine ads and radio commercials in the early 20th century. George Gallup was the first director of market research, appointed by Young & Rubicam in 1932. Though he’s better known for his innovations in political polling, which showed that a relatively small sample of respondents could reflect public opinion just as well as the million-person surveys conducted at the time, he also developed methods to reliably measure the size and makeup of radio audiences. By mid-century, firms like Yankelovich, Skelly & White had begun sending out annual questionnaires to a representative set of Americans in order to track shifting opinions and fashions. Comparing one year’s responses to the next was a statistically sound approach in a world where western consumer markets were the only ones that mattered and one-third of Americans tuned into Bonanza on a Saturday night.

But by the 1990s, the fringes of popular culture began to assert themselves, and with them came a new method of tracking consumer behaviour. The idea of trend hunting started out as an insurgency on the edge of the staid, respectable market research industry. Immortalized in a 1997 New Yorker article by Malcolm Gladwell, “cool hunters” searched the streets of large urban centres for young people on the borders of the mainstream whose choices would inspire their peers — like the Reebok exec Gladwell writes about who used a group of teenagers in the Bronx as a flash focus group to review the footwear brand’s latest creations.

The persisting uniformity of popular culture at the time Gladwell was writing helped cool hunters find their subjects. “It was a lot easier to tell who the influencers were, because they were so distinctly different than the rest of the population,” explains Kristin Jones, senior account executive at Los Angeles–based Trendera, an agency that, like Trend Hunter, employs data in its market research. “Now it’s a lot harder, because the general cultural narrative is more expressive and inclusive.” Being unique, in other words, has become mainstream.

Ironically, this cacophony of niches, countercultures and innovations has been magnified by the very same force that Trend Hunter is using to make sense of the chaos: the internet. Social media in particular has democratized influence, allowing any one idea to quickly gain momentum online. Thousands of fashion, food and technology trends pop up seemingly overnight and disappear just as quickly. A fad like the flash-in-the-greasy-pan cronut craze of 2013 might have lasted years in another era.

Researchers can’t simply compare annual survey data as they might have in the 1960s and ’70s, because no trend or market stays the same for anywhere near that long. Being fast and nimble is paramount in the new cool hunt. “It’s important to recognize which trends are fleeting things that are cool right now but might not be worth talking about next month,” says Jones, “and which things have the potential to be a movement of how consumers are thinking and acting.”

In a way, the fragmentation of the mainstream represents not so much a qualitative shift in pop culture as a quantitative one. We may not all buy the same Michael Jackson albums or sit down at the same time to watch Cheers, but ultimately there are more consumers listening to music and watching TV shows than ever before. Big data represents an evolution in the scale, not the scope, of market research.

“The term just refers to the quantity of information and the number of users,” notes David Soberman, the Canadian national chair of strategic marketing and a professor at the University of Toronto’s Rotman School of Management. “There are all sorts of places that are sources of big data.” Large retailers are likely to collect their own information, and many are digitizing existing rewards systems (such as the revamped Canadian Tire “Money”) to more easily do just that. Internet titans like Google and Facebook already sell search and interactivity data to other corporations.

But having the spreadsheets and the statistics has never been enough. “Big data can be very helpful in testing ideas that you have or getting basic information, but it’s not predictive,” notes Jones, adding that “so much of trends is about creativity.” In other words, data might tell us what’s worked before and what’s working now, but it has a hard time inventing something new.

Even a data evangelist like Gutsche admits his clients are looking for more than numbers. Until last year, what Trend Hunter sold was access to its analytics platform, providing users with a personal dashboard that allowed them to explore patterns and see how individual ideas were faring. But clients were requesting someone to tell them what the data and patterns meant.

Back then, the massive number of website hits Trend Hunter logged translated into significant advertising revenues, but the bottom fell out of the banner ad market in the fourth quarter of 2013. “We had an ad agency that once paid us $1 million a year, and we’ll make $50,000 from them this year,” Gutsche explains. “Trend Hunter was looking at its first loss, which would have been $700,000.” Facing questions about the sustainability of its business model, Trend Hunter added an advisory arm, helmed by president Shelby Walsh. Named one of Marketing Magazine’s 2013 Top 30 Under 30, Walsh had joined the company in 2009 fresh from university; she was previously a scriptwriter and editor for the website. Just 26 years old, Walsh is the person clients call when they have research questions they want Trend Hunter to answer.

After a period without any new sales, Trend Hunter signed up two brands in one month, five the following month and a dozen the next. As of October of this year, the company had about a hundred clients.

If the platform identifies what consumers are searching for, then Walsh’s team of advisers tries to answer the question at the heart of all market research: Why? That’s where things get less scientific.

Soberman believes real-world smarts are a vital skill for anyone trying to explain and interpret customer behaviour. “The number one thing you need is experience in the actual universe,” he notes. “And you also need to have some understanding of how business works.” Gutsche certainly has experience in the corporate world, having spent time as a management consultant at Monitor Group after completing his MBA at Queen’s University and having built a billion-dollar portfolio for Capital One Canada. (“Which sounds great on your resumé, but if I was talking to my 12-year-old self, I’d have to say, ‘Hey, you grew up to be a banker,’” he jokes.) He says Trend Hunter’s client advisers are company veterans, editors who have published thousands of articles and, perhaps most important, are young enough to identify with the audiences that brands are hoping to reach.

“When we write our reports, we try to think about why,” Gutsche explains. “We use our patterns to think about the social influence causing them.” But Trend Hunter’s advisers are not trained data scientists or psychologists, so while their conclusions might be based on a sound data foundation, they’re ultimately the product of deduction and intuition. That sounds awfully like the kind of “gut instinct” Gutsche is fond of railing against.

The Globe and Mail once called Gutsche an “oracle,” and he has all the hallmarks of a public intellectual. But Gutsche waves aside suggestions that he’s become something of a guru himself. “The guru, that’s our competition, and we believe in the power of the crowd,” he says.

Who the crowd consists of may present another problem for data believers. The cool hunters of the 1990s focused on the increasingly individualistic generations X and Y, and the audience from which internet-based research dredges its data skews young as well. “When you start looking at the people who are active online, they tend to be demographically younger, much more electronically active than the average in the population, more educated,” notes Soberman. Companies may be keen to target those kinds of customers, but the concern remains that consumer insights are being based on a self-selecting audience of young people. It’s not quite the know-everything-about-everyone promise of big data.

In this way, we might not have come all that far from those late-century cool hunters after all. “There’s definitely an art to it—you need to have an eye for recognizing what’s different and popping,” says Jones, whose firm, unlike Trend Hunter, still does the kind of street interviews Gladwell was writing about. “You’ve got to know how to spot the cool kids.”

Trendera and Trend Hunter share some corporate clients, including Target. That’s par for the course in the market research industry, because every firm offers something slightly different. At $24,000 a year, Trend Hunter absorbs a miniscule fraction of the marketing and product-development budget at most major corporations. “For a company like Samsung, this is a no-brainer,” says Soberman. “This is definitely something you’d buy, because the price point is pretty low. If I’m a small entrepreneur, a two-person startup in Toronto, it may be a somewhat different question.”

Perhaps companies are motivated by a trend that is just as rooted in the Internet age as big data: fear of missing out. No brand wants to be the one that failed to ride a consumer wave because it wasn’t willing to spend the equivalent of a rounding error in research costs. “If your approach is to just buy a bunch of reports to see what’s going on in the marketplace, that’s not as likely to get you a return on your market research dollars as a specific need,” says Robert Rubenstein, who spent three decades in the market research business at Canadian corporate heavyweights Molson Breweries and TD Canada Trust before recently founding his own startup, Horizn. “There has to be a concrete project or objective that leads you to at least the hope of profitability, as opposed to jumping on a bandwagon.”

Gutsche believes Trend Hunter’s offerings stand out in a crowded field. “Most of these companies will subscribe to all sorts of different trend services that have been feeding them trends for decades,” he notes. “The word ‘trend’ is almost ruined in a way, because they’re just overwhelmed by information.

“Everybody is telling them everything, so we’re kind of like a chaos filter.”

This article originally appeared in Canadian Business.