July 10, 2012
In a small village south of Moscow, three programmers are writing software that is learning why we like certain websites and not others, and what makes us buy a product online. It’s incredibly valuable data that gets at the eternal obsession of marketers: how to generate return customers who do word-of-mouth advertising for your business.
Nearly 9,000 miles away, at the University of Melbourne, Dr. Brent Coker is working as a lecturer of marketing. A data fanatic, he traveled all the way to Russia to recruit those 3 programmers. He needed mathematical geniuses to handle all the algorithms behind his new startup, Webreep.
“Russians are brilliant but very, very difficult to work with,” says Coker. “If I tell them to do something, they’ll go ahead and do it their way because they think their way is better.”
Webreep began in 2004 as part of Coker’s PhD in information systems. He knew that consumer satisfaction was the jackpot for online businesses: it would make buyers more loyal, spend more and more often, and tell their friends. But he wanted to test what caused satisfaction: was it a website’s nice design, or easy navigation, or something else? So he drew up a list of hypothetical causes from past research on consumer behavior and turned them into equations that could measure their effects.
“The Internet is just the environment. The psychology of consumers – the way that we think – doesn’t change,” explains Coker.
Then came the hard part – developing the simplest tool possible to collect and analyze behavioral data. With the help of his programmers, he turned Webreep into a simple line of code that any business could insert into its website. It would generate a feedback form asking visitors about their experience on the site, the kind of survey we see all the time.
But the questions in the survey aren’t direct, like Will you be loyal to this business? Questions like that don’t always work, Coker explains, because we might not know the answer or we might lie – the way we tell a waitress that our bland, overpriced meal was “good.” Instead, Webreep surveys use psychometrics techniques to get at what we really think. “The way that these questionnaires are structured is that they’re very difficult to game,” says Coker.
With data flowing in from tons of websites, Coker and his team do statistical analysis and run correlations to see what factors, like design or navigation, are most important for consumer satisfaction. Meanwhile, the website owner gets a simple report: an overall satisfaction score (1-5), suggestions for areas to improve, and warnings about what is causing dissatisfaction.
“I like the data,” Coker says. “My passion is figuring out why people behave the way that they do on the Internet.” For example, based on Webreep data, he’s discovered that people value “ease of use” the most. They get the most dissatisfaction, respectively, from not finding information, slow download speed, and slow response to email.
When Coker’s plane landed in Russia 2 years ago, his friend picked him up at the airport (and then introduced him to Russia’s pastime: vodka drinking). Coker wanted to recruit researchers at Moscow University to help with Webreep, which was just getting started. He handed over his passport to enter the university and met with a few academics. “They’re freakishly good at mathematics,” he recalls – they were working with phenomenally advanced technologies, like wireless electricity. But soon it became clear that Webreep wouldn’t be their next project: the government denied their request to “share knowledge” with Coker, a foreigner.
The trip wasn’t all in vain, however. Coker used to hang out in IRC chat rooms – basically early forums on the Internet – with hardcore programmers and brilliant hackers (including the shady variety). He met some of them in Russia, and they eventually put him in touch with those 3 Moscow programmers, Dimitry, Kol, and Alexei.
To keep this trio engaged, Coker uses a little employee psychology: he’s continually observing what makes them tick. “It was really difficult to try and create these relationships,” says Coker. “They get bored really easy.”
But he soon discovered the carrot that would entice these wayward geeks: the prospect of learning how to channel their genius into deceptively simple tools. Coker remembered visiting Russia and spending 30 minutes scratching his head in front of an ATM: the “most complicated machine” he had ever seen. It even had a string hanging off the side with a manual.
“Russians are brilliant,” he realized. “But they don’t know how to turn their brilliance into something that’s usable.”
So Coker is constantly coming up with little challenges for his team: most recently, they created a feature where a user giving feedback could take a screenshot and even make notes on the screen – like “This button doesn’t work” or “This text is too small.” It was an astronomical task, but it took them just one week to build a prototype.
For the release of Webreep version 2, Coker’s programmers requested permission to rewrite about 80% of the software in a new framework – just as they had requested other rewrites in the past. Many managers would have vetoed the ideas and settled for the current version, but Coker routinely gives in to their nerdy fancies.
“They don’t like working on old technologies. They want the new shiny thing,” he says. “They’re like kids with a new bicycle – Can I buy some new handgrips? Can I buy a new seat? Or can I put these new stickers on my bike? I let them do that, and it keeps them happy.”
Two years after traveling to Russia, Coker is in his version of heaven: piles upon piles of data and brilliant minds to help him parse it. He may look like an entrepreneur – and his ideas certainly have real market potential – but he’s an academic at heart: he publishes in academic journals, presents at conferences, and releases studies: like what causes satisfaction on websites (based on Webreep) or what makes a 30-second video go viral. In a way, entrepreneurship is just a means to an end.
“I’m not a Silicon Valley guy who’s trying to raise capital all the time,” says Coker. “I’ve realized my dream … having massive amounts of data so I can do my academic-type stuff.”
Did you like this article?
Get more delivered to your inbox just like it!