“How are you?”
It’s one of the first questions you ask when you meet someone, and the answer sets the tone for your entire chat. Will you share a joke or commiserate? Will you make plans for a fun outing or help your friend through their latest personal crisis?
As in conversation, the key to success in the advertising world is empathy. Cluep—a Toronto startup founded in 2012 by three idealistic millennials—is well aware of this, using algorithms gleaned from artificial intelligence to assess consumer emotions based on their public posts on Twitter, Pinterest, YouTube and Tumblr. Cluep’s ad platform then seamlessly serves a diet of emotionally appropriate messages to users’ social media feeds. Genius, right? Hundreds of brands, including marquee names like McDonald’s, Air Canada, Mercedes and Victoria’s Secret seem to think so.
Karan Walia, Cluep’s 29-year-old CEO, has shepherded the company from an underfunded pipe dream to a powerful media platform that has since managed upward of 2,000 major campaigns. That success has elevated the company to a 40-employee operation, sales offices on four continents and a gangbusters five-year growth rate of 11,433%—good for 5th place on the 2019 Growth 500 ranking.
Walia genuinely believes that serving consumers better-targeted messages enhances their lives by connecting them to relevant solutions to their problems: “We have lots of case studies on how businesses have leveraged our platform to get in front of people who weren’t familiar with their product lines,” he says. But aside from the obvious monetary upsides, using AI to enhance people’s quality of life has been—and continues to be—the company’s mission critical, a priority that sometimes gets lost in the rapid-fire world of automation.
Walia’s warm approach to tech started when he was young: growing up in a small village in northern India, he always felt a close sense of community. At 14, Walia convinced his parents to let him move to Ontario to find better opportunities, and lived there with an uncle’s family. Four years later, he was dazzled by the first iPhone, pondering how the new tools of computing could be used to reinforce connection in the broader world. His dream: to use natural language processing and speech recognition to create a digital personal assistant that would connect people to information, education and, importantly, each other.
But Walia soon realized his obsession required more resources. Apple alone had hundreds of people working on similar projects (NB: Siri would debut in 2011). He tried to raise venture capital, but, as a one-man operation, he had trouble landing meetings, let alone money. In November 2011, he recruited his brother Sobi—who had followed him to Canada in 2008—and teen programming whiz Anton Mamonov (now Cluep’s chief technical officer).
While Walia was studying information technology at Toronto’s York University, he and his teenaged partners built a text-analysis engine to suss out the emotions behind millions of words and phrases. It wasn’t easy: moving beyond keywords to understanding the grammatical structure of entire sentences required deep neural network architectures based on the latest AI research. Training the algorithm on endless data sets required more computing power than the team could afford, so they commandeered a roomful of York computers meant for wider student use.
But their goal of an all-empowering personal assistant increasingly looked like a moonshot, particularly for a company funded exclusively by Walia’s student loans. They realized they needed an alternate product, something they could actually sell quickly. With coaching from a few advisors, they decided to leverage their language-processing work to build a more efficient ad platform.
Cluep’s breakthrough arose from the team’s confidence that they could improve on the keyword advertising that vaulted Google (and other search engines) to success two decades ago. What would happen, they wondered, if you matched ads, not just to the words used in social media posts, but also the emotions behind them? Put simply, if someone wrote, “My iPhone is sick,” could you discern whether that user was planning to turn in their phone for repairs or buy a high-end pair of earbuds?
On the strength of this software, Cluep raised $500,000 from its advisor group They also secured the enthusiasm of Stacey McInnes, a marketing consultant with 10 years of agency experience, who become the firm’s unpaid mentor. She arranged the new platform’s first demo, with a Toronto ad agency, in 2014. With her help, Walia closed deals with the first five clients he pitched—and McInnes became Cluep’s first employee.
“I thought this was the holy grail of marketing,” she says. “We used to be able to do research projects to find out what people felt about a brand, but with this product, marketers could engage people one-on-one.” Besides being chief strategy officer, McInnes says that, at the time, she was the only staffer old enough to rent a car.
Today, Cluep clients can target prospects using both keywords and the feelings behind them. So when McDonald’s buys the keyword “lunch,” it can assume that its prospects are actually talking about their next meal, not complaining about yesterday’s. Advertisers can also target consumers who seem to be in the right frame of mind to entertain their product or service, based on the eight core feelings identifiable by Cluep’s algorithm: scared, sad, horrible, angry, happy, loved, excited and hopeful.
Never underestimate the power of empathy: according to Walia, Cluep’s campaigns generate average click-through rates of 1.7%—a 300% improvement on the industry average.
Since its first deal—helping 20th Century Fox Canada sell more tickets for its 2014 tearjerker, The Fault in Our Stars, Cluep has created two new services: Cluep Pics helps brands target prospects based on the images they post, while Cluep Places tracks where consumers are actually located, where they go and where they shop. Another new service, yet to be announced, is in the testing phrase, and a fifth will be released next year.
In the meantime, Cluep is continuing to double down on its people-first mission, hiring based on such values as fun, kindness and respect for others. With its granular focus on relationships, some observers wondered why, last fall, Cluep agreed to a takeover by Impact Group, a food-distribution company in Boise, Idaho. Walia explains that Cluep welcomed the $50-million deal as a chance to tap Impact’s resources to develop more new products. Impact, now owned by New York investment firm CI Capital Partners, hopes to use Cluep’s AI expertise to develop—you guessed it—relationships with its packaged-goods clients.
In no hurry to sell, Walia says the founders held out for autonomy in running their business and maintaining a head office in Canada (the market responsible for 40% of Cluep’s revenue). Still, alert readers may be wondering: whatever happened to Walia’s ambition to build an AI-powered personal assistant? He says that dream is still alive. For now, he’s offering no details. Siri might be listening.