We ain't seen nothin' yet
Editor’s note: Humphrey Taylor is chairman of The Harris Poll and a member of the executive committee at Harris Interactive, New York.
History teaches us that new technologies are used, initially, to do old things better, or faster, or cheaper but that eventually they are used to do completely new things which nobody had ever thought of doing before - advertising on TV, driving to the beach on Saturday, watching Neil Armstrong’s first step on the moon, traveling thousands of miles for a week’s vacation or just calling home on Mother’s Day.
When new technologies are available, human ingenuity, greed and creativity find a thousand new ways to use them. So it is with the Internet.
Much of the buzz, and discussion, about on-line research has missed the most exciting thing about how the research industry will use the Internet. Of course, it will replace (and is already replacing) some traditional quantitative and qualitative research methodologies. But, it is much, much more than just a faster, more cost-effective way of doing what we already do reasonably well on the telephone, in the mall or in people’s homes. Its greatest potential, by far, is not as a replacement technology, but as an enabling technology that will allow us to do things we could hardly dream of doing before. After less than two years of conducting research on-line, and having conducted on-line research for 186 clients, we find that only about 20 percent is replacing traditional research methodologies. Fully 80 percent is work which would never have been conducted without the Internet. That is what is truly exciting!
There are several reasons why the Internet is such a powerful technology, enabling us to do so much exciting new research:
- We can survey humungous samples. Some on-line surveys have samples of over 100,000 people.
- We can screen for and survey substantial samples in tiny segments of the population. Some on-line surveys have been of segments of smaller than 1 percent of all adults, with samples of hundreds, and even thousands.
- We can survey very hard-to-find people. We have surveyed the very rich, people with unusual medical conditions, people who have traveled to out-of-the-way places or who have bought the most expensive cars - even divorced senior business executives.
- We can show them copy, products and packaging. We can test concepts, print ads and TV commercials. We can show and rotate different products and packages to test their appeal.
- On-line respondents speak eloquently to us in their own words. The quantity and quality of verbatim responses far exceeds what we obtain by telephone, or face-to-face - even when we record and transcribe their words.
- We can re-contact respondents or key sub-samples quickly and easily. It is much easier and faster to re-interview respondents multiple times on-line.
- Respondents critique our surveys. After many surveys, we invite respondents’ comments and criticisms. Some of them are excellent critics, and have helped us to improve our questionnaires.
Reality is more exciting
All that sounds pretty good in theory, but the practical reality is even more exciting. These are just a few examples:
- A three-step experimental design involving 22,000 women. In the first stage, women were segmented into 16 personality types. In the second stage 24 different advertising options were tested, involving different media and different products, which were displayed on-line. The third stage involved the collection and segmentation by additional psychographic variables and Web security. In addition, data were collected about brand images and brand usage in 12 different product categories.
- Measuring corporate reputations. We are surveying very large samples of the public about corporate reputations with substantial sub-samples of people who are familiar with the companies involved, consume their products, invest in and are employed by their industries. We are doing this both for specific industries and across all industries. We are also building dedicated panels and key stakeholder groups including money managers, IT decision-makers and health care professionals for these surveys.
Because of the speed of the Internet, we can measure, almost instantaneously, the impact of a crisis in a company’s reputation, and track it thereafter.
- Strategic and segmentation advertising research with multiple market segments of investors. We recently surveyed 12,000 individual investors in 10 days to develop segmentation analysis so that our client’s marketing and advertising could be targeted at multiple, diverse market segments.
- Concept testing with multiple market segments. We recently tested four different product concepts with 1,500 people who had recently bought a relatively unusual kind of product. Because we were on-line we could show pictures of four different products, each with a brief written description. The sample was large enough to look at multiple market segments. Data collection took five days.
- Banner advertising effectiveness research. A typical study involved screening our 5.3 million on-line panel to find people who had visited specific sites, and adding samples of registered visitors to our clients’ Web sites to measure recall and advertising impact.
- New qualitative methodologies. While we are conducting traditional real-time focus groups on-line, we are finding that new, interactive qualitative methods may work better. Modified bulletin board and chat room techniques - well-moderated - can be used to develop powerful, new qualitative, research tools.
- Recruiting for clinical trials. We are using our on-line panel to recruit patients to participate in clinical trials for the pharmaceutical industry and academic medical centers. This will enable drug companies to get their drugs to market much faster, so they can enjoy a longer period of patent-protected life.
- Evaluating the impact of sports sponsorship. We are also finding and tracking large samples of people who closely follow specific sporting events and buy the sponsor’s products (and the competitive products) to measure changes in buying behavior related to attending and/or viewing the sponsored events.
- Evaluating 275 e-commerce Web sites. By regularly surveying samples of over 100,000 people, we are measuring and analyzing on-line activities, user attitudes and both on-line and off-line buying behaviors. This service enables on-line retailers, e-commerce strategists, and financial analysts to determine which sites are drawing the most customers and why, and to profile visitors and buyers to uncover which sites have the best quality traffic. It helps sites and advertisers evaluate where they should invest their marketing and advertising dollars, and to determine why consumers leave a site dissatisfied and how customer experience can be improved.
- Survey of cancer patients who have participated in clinical trial. For a study to help understand why so few cancer patients participate in clinical trials, we are currently surveying cancer patients who have, and who have not, participated in clinical trials to find ways to increase recruitment into future trials.
- Survey of parents of children with a particular medical condition. Later this year, we expect to publish a survey of parents of children who were diagnosed with relatively rare medical condition to measure how appropriate and effective they believe the treatment was.
- Providing data for tomorrow’s vital presentation. Several advertising agencies have called us up late in the afternoon asking for overnight research among consumers or purchasers of specific products, to use in tomorrow’s pitch to a new client. We have been able to screen many thousands of households to find and survey the customers they need and deliver the data by 9:00 a.m. the next morning.
- Consumer products and services - customer loyalty and retention. For the first time it is economic to communicate regularly on-line with very large numbers of consumers to increase the likelihood that they will continue to buy and consume our clients’ products. (But because this is a marketing service - to change consumer behavior - this service must be kept separate from research work.)
Exploded capacity
A year and a half ago, we had not thought of, and would not have predicted, most of these applications of on-line research. In another 18 months, the research industry will have exploded our capacity to do a thousand things nobody has yet thought of. We surely ain’t seen nothin’ yet!