Harris Interactive: Reading the Internet’s Mind

PDF DownloadGordon Black, chief executive officer of Harris Interactive, issued a challenge to his technology team in the spring of 1997: Develop a market research system that could utilize the Internet to quickly and inexpensively poll thousands of people on a wide range of topics. The research not only needed to be accurate by the same standards used in traditional polling methods, but the team also had to find a way to translate the results to represent the world at large—not just the highly slanted views of the computer literate.

Black got what he asked for, and in some respects a lot more than he bargained for. Dropped e-mail and nagging problems that crashed the survey engine angered clients, resulting in the loss of several substantial accounts. The 61-year-old executive spent days crisscrossing the country apologizing to longtime customers for errors or delays. At another crucial stage of development, an anti-spamming group blacklisted the firm, threatening to put its entire method of polling in jeopardy. Then came the technology crash and general economic downturn, causing clients to cancel research projects at a time when the company was pouring money into research and development.

Yet Black considers himself fortunate. The Internet research side of Harris Interactive is now the company’s fastest growing—and most profitable—line of business. By year’s end the Rochester, N.Y., company expects Net research to account for about half its revenue. Internet revenue is now running at about $60 million a year, almost three times the company’s entire revenue before it launched the initiative in 1997. Despite the difficulties, Black says his technology team pulled off a remarkable feat that many in the industry said couldn’t be done. Today, 50% of all polls are conducted online.

“It was a painful process; there’s no denying that,” he says, recalling the last five years on a recent afternoon at the company’s headquarters in a Rochester office park complex. “But I gave our technology team an extremely difficult task. I told them to build me a platform with no upper boundary [in terms of number of panelists it could support]. I’ve also been through enough rounds of technology development to know that’s what happens.”