Company surveys of their consumers: How reliable?

Corporations have always used consumer surveys for marketing purposes. Now, some are using so-called public opinion surveys to justify their own positions.

Results like these from such surveys were sent out to the press this year, for example:

* The majority of Americans think money market mutual funds are ''draining credit money away from local communities and consumer needs.''

* Most US households would like to have a satellite ''dish,'' to receive multichannel television programming.

* And 94 percent of all Americans want more product information about the products they buy.

The first finding, concerning money market funds, was commissioned by the American Bankers Association, the lobbying and trade association of commercial banks, which is not exactly a disinterested party when it comes to money market funds.

The second was paid for by the United States Satellite Broadcasting Company, a unit of Hubbard Broadcasting Inc., which would like to beam TV programs via satellite.

And the third survey was done for the General Electric Company, which subsequently began a multimillion-dollar advertising campaign to spread product information.

In each instance it appears the organizations received the findings they wished to receive, leading some outsiders to question the validity of the surveys. Only the results of the General Electric survey made the national news media. The New York Times advertising column ran a story about GE's new ad campaign and mentioned the survey.

At the Wall Street Journal, Charles Stabler, assistant managing editor, says that ''obviously this kind of thing is self-serving and, like any self-serving press release, is thrown in the wastebasket. I wouldn't say we never use one, but certainly if the release is pure puffery, it is treated as it deserves.''

Naturally, the institutions that sponsored the surveys defend them. Laura Jordan Bellis, a spokeswoman at the American Bankers Association, says the ABA sponsored its survey to inform the membership what the public was thinking, not just to send out a press release. Had most people disagreed with the ABA's findings, would she have sent out a press release? Probably not, she concedes.

Most people would probably not have disagreed with the ABA survey, however, because of the way the questioning was worded. Cambridge Reports Inc., a Cambridge, Mass.-based polling firm, asked 1,100 people in six regions of the country to agree or disagree with the following statement: ''Uninsured money market mutual funds offered by investment bankers are draining money consumers would normally put into savings accounts which are the basis to provide credit to meet the needs of the local community.''

The same is true with the General Electric survey, likewise done by Cambridge Reports. GE's first question was worded, ''How interested are you in having sufficient and useful information about the products or services you buy and use?'' Not surprisingly, 94 percent were either extremely interested or fairly interested.

James Murphy, a partner at Cambridge Reports, defends the questions. ''Often a question is loaded intentionally,'' he says, ''and then you begin to unbundle it - make it less loaded - to see how many people switch.'' And the fact that money market mutual funds are growing rapidly also does not bother him. ''Often there is a difference between how the public sees an issue and reality,'' he comments. Mr. Murphy predicts that use of such surveys will grow in the future, since ''the private sector is concerned with what the public thinks.''

Robert Franzan, a vice-president at the United States Satellite Broadcasting Company, says its survey, too, was not made so stories would get written. He says, ''We did the survey and filed it with the FCC (Federal Communications Commission) and told some people about it.'' He holds that the importance of the survey is that it showed people were willing to buy a satellite dish even though they didn't have much information about it.

Melissa Wrenn, an associate of Frank N. Magid Associates Inc., the Marion, Iowa, company that did the survey, denies it was constructed to meet the vested interest of its client but says it merely provided information from a well-constructed random sampling of the population.

But Humphrey Taylor, president of Louis Harris & Associates, says pollsters can get anyone to agree with practically anything according to the way they word a question. ''I could write four different versions of a question,'' he states, ''and each would look good to the layman, and each could have different nuances.''

He notes that a recent issue of Public Opinion magazine, published by the American Enterprise Institute, asked the question ''Do people lie to the pollsters?'' The magazine found that different surveys asking the same questions got different answers. For example, Mr. Taylor says his firm has done surveys for banks concerning money market mutual funds and found the public did not want to see the government restrict where they could invest their money. The implications of the ABA survey are just the opposite.

When looking at a survey, Mr. Taylor says, it is important to see who paid for it. ''A client often pays for what they are looking for,'' he comments, adding, ''If it was paid for by the media or a foundation, the chances are it is more impartial.'' The polling organization matters, too, he says, noting that ''if a national firm is used they can't afford to be known as a 'hired gun.' Most of the bad polling is done by the hungry firms.''

Not everyone does a survey to make a point. Recently, Warburg Paribas Becker-A. G. Becker, an investment banking house, commissioned a survey of 50 chief executive officers, which its president, Ira Wender, said was done because Becker felt CEOs were misunderstood.

According to Burns Roper, who did the survey for Becker, the firm commissioned the survey to help get some attention in the business community.

Suzanne Pederson, a vice-president of public relations at Becker, replied, ''Of course we hope to engender some goodwill among this group, which is important to us.'' She adds, ''But we believe it goes beyond that. There is a lot of misperception of these people.'' To try to change this misperception, she says, it cost Becker between $50,000 and $100,000. But no major publication wrote a story on the survey.