Editor’s note: Humphrey Taylor is chairman of The Harris Poll and a member of the executive committee at Harris Interactive, New York.

Andy Grove, the founder of Intel, says that "In five years’ time, all companies will be Internet companies or they won’t be companies at all." The Economist recently wrote that "The Internet (will) change everything -- the way we work, the way we learn and play, even, maybe, the way we sleep or have sex. What is more, it is doing so at far greater speed than the other great disruptive technologies of the 20th century, such as electricity, the telephone and the car."

I will leave others to describe how it will change the way we have sex. But even a modest assessment of how it will change the research industry would lead to the conclusion that in a few years, most successful international market and opinion research firms will be conducting much of their research on-line.

It is a cliché but it is true: the Internet is by far the fastest growing technology in the history of the world. At the beginning of 1995 only 7 percent of all adults in the U.S. were on-line, whether from the office, from home, school, a library or somewhere else. Less than five years later that number has risen to approximately 50 percent (49 percent by our last measure), and looks set to reach about 60 percent by the end of the year 2000. Globally the same phenomenon is occurring. While most developed countries are behind the U.S. in terms of Internet usage penetration, they are moving rapidly to close the gap. Today, approximately 58 percent of the world on-line population resides in the U.S., 26 percent in Europe, and 12 percent in Asia. By the year 2002 it is estimated that the U.S. share will decrease to 46 percent as more Europeans go on-line and claim approximately 36 percent of the market, while Asia will stake 14 percent.

A revolutionary change in marketing research

Marketing and opinion research has always changed to take advantage of new advances and technology. However, the use of the Internet to conduct marketing and opinion research is a much more revolutionary development than the other, more modest, changes I have witnessed in my 36 years in the industry. In the ’70s we began to use the telephone for data collection. In the ’80s we started using CATI and CAPI systems. These changes did move around the world, but were restricted in a very real sense to the infrastructure development within countries. Countries with well developed telephone systems such as Europe, Australia, Japan moved their data collection to CATI, but countries like China continued to depend on face-to-face interviews. And if you wanted information from around the world, data were still collected on a country-by-country basis. You needed research support in a lot of countries to make that research happen.

While these were major advances they did not fundamentally change our thinking about how to collect and analyze data. They did not fundamentally change the way we designed a questionnaire or how we collected data.

Such is not the case with the Internet. The Internet as an audio-visual medium changes the way questionnaires are designed. And, not only can you collect data from many countries from one site (really anywhere in the world), countries such as China have leapfrogged over the telephone and CATI revolution, directly to the Internet. This has created tremendous possibility as well as more than a few problems with which market research professionals must deal.

As market research firms saw an opportunity to harness the Internet and offer a cost-effective alternative to data collection in the U.S., they found themselves leading the charge, while clients waited on the side for results. Once its viability was apparent, clients came to appreciate its cost-effective approach and now drive the demand for on-line research on a global scale. Where large-scale international projects were out of reach for many clients, on-line research makes it viable in a number of growing pockets, most notably in Northern Europe, Canada, Australia, New Zealand, Hong Kong and Singapore, where traditional data collection costs are relatively high. It is interesting to note that while Europe’s share of the on-line market is nearly half that of the United States, there are a number of European countries which have overall Internet penetration rates higher than the U.S., including Sweden, Denmark, and Finland, which offer the greatest opportunities for on-line research.

Conducting global Internet research

Before conducting or commissioning Internet surveys internationally consider:

  • Do you know who is on the Internet in the countries you are surveying?
  • Do you know the code of ethics and privacy standards as well as any laws that might apply to Internet research in these countries? Does the software you are using support all the various fonts necessary for the languages in which you are interviewing?
  • Do you have the data to understand what the population you are surveying represents? Raw on-line data substantially under-represents some groups (and those under-represented groups vary by country). The importance of these groups to your final analysis may be more or less important depending upon the objectives of the study.
  • Have you considered variables other than demographics in determining weighting variables? In many cases behavioral data such as time on-line and number of years using the Internet, which varies substantially by country, may be more important weighting factors than age, income or education.
  • Have you considered cultural issues that might have been handled by the field service in the country? The ability to contact people around the world from any country does not mean that people around the world answer surveys the same way. Cultural concerns and questionnaire norms must still be taken into account and a local expertise is still invaluable in research design.
  • Have you considered the "method effect" in your questionnaire design and have you thought about how that might vary by country? Using the same questionnaire around the world may not give you the same results, whereas some variation in design may give you the comparable data you are looking for. We know for example that different cultures respond to scales differently -- so do on-line and off-line populations.

Replacement technology?

On-line research has already been described both enthusiastically, and critically, in many different ways. "A replacement technology" is one such phrase and it will surely replace much of the qualitative and quantitative research work currently done face-to-face or by telephone. However, it will not fully replace other methodologies. Printing did not fully replace handwriting. Radio did not replace newspapers. Television did not fully replace the movies or the radio. We will continue to do in-person and telephone research, even if this amounts to a rapidly shrinking part of our work.

What is much more exciting to me is that the Internet will enable us to do many things we could not do (or afford to do) before, greatly enhancing the value of our services.

Very specifically, the Internet enables us to:

  • survey huge samples of people;
  • survey tiny sub-samples of the population;
  • do everything we can do on CAPI/CATI;
  • show lists, still and moving images;
  • obtain much richer verbatim replies.

And it enables us to do this:

  • incredibly fast;
  • at affordable costs;
  • all over the world.

The future

Because of the speed of the Internet revolution, new and exciting research applications of on-line research are appearing every week. We repeatedly stumble on new ideas, new and better ways of doing things, as well as things nobody thought of doing at all. Moore’s law (Moore, of course, was the other founder of Intel) is that computing speed and power doubles every 18 months. Internet traffic is reported to be doubling every six months. Our knowledge and understanding of how to use the Internet to conduct both qualitative and quantitative research is probably growing (from a zero base!) even faster. If there is one certainty, it is that we ain’t seen nothing yet.