ONLINE FOCUS GROUPS: THE BACKLASH - In the late 1990s, the idea ofconducting online focus groups excited the PR community. Then realityset in. Kimberly Krautter weighs the pros and cons

Corporate America's love affair with research has always been

fickle. It likes to court it, but prefers a cheap date.



So when the internet emerged, offering the chance to discover people's

views at the click of a button, PR people were excited about the

possibilities, even for qualitative research. Not having to physically

get a group of people together in a room potentially offers substantial

savings on both costs and time.



There is also the factor that people's views will not be affected by

group dynamics and their responses will be a truer representation of

what they feel. The relative anonymity also makes it easier to poll

people's views on controversial or personal issues.



Kari Bjorhus, director of interactive public relations at Coca-Cola,

says, "We generally find people are more candid and truthful because of

the relative anonymity the net provides us."



Coca-Cola has begun experimenting with broader market studies on the

web. It is now using online panels in the same manner as mall intercepts

to test television advertising, new brand concepts, and consumer

attitudes toward their brands.



Limitations



However, given Coke's global presence and the uneven distribution of web

technologies overseas, Bjorhus says the major drawback is that "we have

limited ability to go outside the US."



Indeed, while they might have their blue-chip fans, many in the industry

are now increasingly cautious of conducting online focus groups.



Even early adopters like Duffey Communications are now hesitating to

recommend web-based focus groups to their clients. In the early '90s,

Duffey developed a full-service in-house focus group capability.



Although he says he welcomed the advent of internet focus groups, CEO

Lee Duffey is surprisingly vociferous in his opposition to them today:

"It's an interesting barometric test, but there is zero statistical

significance."



Duffey's disenchantment resonates throughout the PR industry. Jacqui

Walford, operations director and SVP-international with Echo Research,

says agencies were excited by the newness of the internet and the

supposed instantaneous turnaround it could provide. "I think people

expected more than it could deliver," says Walford. "It's a tool that

needed to shake down to determine its own level."



Early enthusiasm by dot-commers has a lot to do with today's

disillusionment.



In the dot-com heyday, dozens of ersatz research companies deluged the

internet offering price clubs and other incentives to users in exchange

for joining research panels. However, since most of these companies did

not understand the fundamentals of research science, the results they

posted were deemed specious, thus leading to their rapid demise.



Panel mismanagement was one culprit. Poorly constructed, long

questionnaires and too-frequent solicitations led to participant burnout

and attrition.



There was also a concern with the veracity of panel responses and

whether the prize payoffs led to panelists "just hitting yes all the way

down the page," says Duffey.



But it comes down to the single hot-button issue of web focus groups:

sampling. Coke's Bjorhus says that the company is concerned about

presumed biases of online panels, and does work to ensure the sampling

is representative of and weighted to the US population.



No diversity online



Even if you don't need to worry about getting a globally representative

panel, it is unlikely that you will be able to attract an online sample

that truly reflects the American population.



"The internet is all white people," exhorts Duffey. "The demographics

have moved down to 50-50 male to female, but if you reach people who are

likely to give information to an unnamed source, you're going to get

largely male respondents."



Duffey's anecdotal response is backed by US Department of Commerce

statistics.



As of August 2000, the ratio of males to females is now relatively even,

but internet usage still skews heavily to caucasians and the

college-educated.



Of the total US population over age 25, 50.7% of white males use the

internet, whereas only 27.9% of black males are online. Similarly, 49.9%

of all white women over age 25 are on the net versus 30.5% of black

women. Among Hispanics, the nation's largest ethnic group, only 22.7% of

males and 24.7% of females are active internet users.



The data extends to show that most users are upwardly mobile and

affluent.



According to Katharine Paine, CEO of Delahaye Medialink, the research

industry expects more people to want net-based research, but says at

present the concept is a hard sell. "The joke in the research industry

is, 'You want faster, cheaper, better - pick two," says Paine.



Brian Collins, head of survey research at CARMA, points out that the

labor-intensive nature of the task of coming up with a representative

sample often cancels out the savings made by not having to get people

physically together. "Although your cost per completed survey can be

more favorable, the underlying principle comes back to being able to

reduce, as much as possible, the bias of the respondent."



It's all in the sampling



Many companies have now realized that it is better to reserve the use of

online focus groups to situations where a sample is easy to find.

Internal communications and existing customer research are two of the

widest applications.



Employee and customer panelists are most likely to be wired into the web

(or an intranet), and they have a vested interest in the subject, thus

assuring high response rates.



Cingular, the nation's second-largest wireless company, has used online

focus groups to aid the birth of its new identity. Starting immediately

after the October 2000 announcement of the joint venture between SBC and

BellSouth that formed Cingular, the company developed an "employee

promise" designed to help shape the corporate culture that it wanted to

test.



The company gathered employees around the nation online to ask what

barriers were present in their work groups to impede follow-through on

the promise.



"We were able to identify offices around the country that were doing

things very well, and we also learned that talking about the promise

from headquarters wasn't going to get people living it, and we had to do

things to make that happen," says Monica Mears, Cingular PR manager.



If you want to use the internet for a group other than customers or

employees, one option is to use traditional means to find your sample,

then employ the internet to actually run the group. This is the system

used by Knowledge Networks of Menlo Park, CA.



At a rumored cost of $1,000 per household, Knowledge Networks has

amassed 250,000 American households through random-digit dialing with no

quota sampling. The group includes both active and non-internet

users.



Once qualified as panel subjects, the company provides them with web and

TV access, and pays for the internet service. Panelists agree to

participate in a ten-minute survey once a week on a variety of topics,

with care taken to ensure that households are not queried on the same

market category within a three-month period. Furthermore, new

participants are trained and tested to make sure they are capable of

using the system.



"We gather a lot of data about these customers, their opinions, their

attitudes, media usage, and even what they purchase at the grocery

store," says Karen Lorentson, VP of marketing communications. "We can go

back and do a profiling study to do additional segmentation based on the

results."



To date, the only concern registered about Knowledge Networks' approach

is whether, by virtue of the frequency of testing, the participants

become "professional panelists" and how that may engender bias. There

was a concern that internet panelists have more instantaneous access via

search engines to the broad market information being tested than

traditional panelists, providing an easy "cheat" facility.



A study by Marketing Research magazine tested whether by joining a

panel, "research participants might feel obliged to prepare for future

surveys and take steps to increase their awareness and knowledge of

current events or advertising." It further tested whether Knowledge

Networks' focus groups were over the long term developing an increased

awareness of obscure brands by virtue of their participation. The study

concluded there was "no support for any panel conditioning." The

magazine also found no basis for the idea of "professional

panelists."



Knowledge Networks is attempting to leverage its product competitively

against other research firms. Roper Starch VP Rob Barrish says the

"opt-in" approach used by Knowledge Networks and his own institution

(which also operates online focus groups) requires certain disclosures

in order to understand the data compiled.



"You want to know how the list was compiled and how often the

participants are contacted (on a topic)." Barrish says he is concerned

that "there is a certain type of person who will join and agree to be

contacted a couple times a month," and he wonders whether testing

behavior is affected by this.



Behavior problems



Beyond the sampling dilemma, there are behavioral issues inherent in

conducting groups online.



Duffey wonders how researchers can be assured that the panelist, for

example, actually is a teenager just because the person logs on as

one.



Furthermore, Linda Hadley, director of strategic planning and research

at Porter Novelli, says she has concerns about the ability of online

focus groups to gauge the more subtle qualitative responses of

panelists. Until broadband becomes a more universal standard, she

observes, researchers are unable to observe body-language cues that

enable a moderator to probe deeper with a subject.



"Right now, there is no replacement for getting people in a room and

sitting behind a screen and watching them," says Hadley. She believes

focus groups on the internet are useful only to gather quick and basic

information. "The less deep you want to go into emotions, the better off

your findings are going to be."



Current attitudes toward online focus groups are largely affected by the

pendulum swing against all things dot-com. However, just as the web

didn't evaporate with the dot-com crash on Wall Street, web-based

testing will not disappear. The genie is clearly out of the bottle, and

although the idea of "cheaper, faster, better" is relative to the

target, advances in sampling ensure that the internet will most likely

become a resource with increasing validity to the science of marketing.



Have you registered with us yet?

Register now to enjoy more articles and free email bulletins

Register
Already registered?
Sign in

Would you like to post a comment?

Please Sign in or register.