Bankruptcy at GM: What Do Consumers Want?

Last November, GM CEO Rick Wagoner flew to Washington on a corporate jet and told Congress that bankruptcy would lead to the liquidation of the firm because research surveys indicated that consumers are unwilling to buy vehicles from a bankrupt automaker. And yet, just four months later, Wagoner agreed that a pre-planned bankruptcy might be manageable after all. Motor Trend magazine, the automobile bible for gearheads, labeled Wagoner’s recent comment a stark reversal of his earlier position.

Stark reversal indeed … and, at first glance, seemingly unjustified. After all, consumer spending in America hasn’t yet bounced back in any noticeable manner. So why would Wagoner suddenly change his mind about buyer sentiment? And should our government pay any attention to sentiment at all when making mammoth decisions about whether to allow GM to fail?

What Exactly Is Consumer Sentiment?

Consumer sentiment data are produced by snapshot polls that simply measure transient attitudes at fleeting moments in time. Politicians, for instance, are well versed in the “here today, gone tomorrow” nature of survey data. The first President George Bush’s Gallup poll approval rating soared to 89% in March 1991, and yet he was decisively voted out of office the following year. Likewise, his son’s rating peaked at 90% in September 2001, but then slumped to 25% later in his tenure.

Why did so many people change their minds about these presidential figures so quickly? Perhaps the world changed in dramatic fashion during these turbulent periods of time, triggering changes in the public’s opinion of their political leaders. And yet some polling experts warn us that the polls themselves are subject to exaggerations and misstatements.

In the business world, the Reuters / University of Michigan’s Index of Consumer Expectations is incorporated into the U.S. Index of Leading Economic Indicators as a monthly measurement of consumer attitudes and expectations. However, many analysts consider it to be a mere “reflection of uninformed opinion.” That hardly seems to provide a suitable foundation on which to make massive corporate bail-out decisions …

… or does it?

The Value of Snapshots

To give Rick Wagoner the benefit of the doubt, it is important to remember that snapshots do generate some value; after all, that’s why we build digital cameras into mobile telephones. We don’t expect to use our phone cameras to produce another Sistine Chapel, which reportedly took ten years to paint in the 1500s and twice as much time to restore in the 1990s. We simply want to create images that will be emailed to friends one day, and deleted from inboxes the following day.

We can certainly debate whether our federal government’s perspective should be short term or long term in nature regarding our economic revitalization strategy. But Rick Wagoner was pretty clear about his mental time frame last November: he bluntly stated that GM would go bankrupt in a matter of weeks if not bailed out by the government immediately. And President Obama maintains that his massive government deficits will help stimulate spending in the short term, even though the U.S. Congressional Budget Office warns that they may hurt the national economy in the long term.

So there is some support among the experts for Wagoner’s position that research polls of consumer attitudes may be helpful to decision makers, particularly when our considerations are short term in nature. And thanks to the internet … thank you, internet! … there are many free online web applications that business men and women can use to develop their own consumer polls.

Care To Develop One?

Would you care to develop your own consumer sentiment poll? If not for the entire economy, then perhaps for your own product or service?

It’s actually not very difficult to develop a poll.  First, we must create a list of questions and find a representative group of people to poll; most statisticians recommend using Likert Scale questionnaires for randomly selected samples. But not every pollster uses such scientifically pure approaches; in the Tonight’s Poll section of the home page of CNN’s Lou Dobbs Tonight, for instance, any web surfer is welcome to supply a yes-or-no answer to rather uni-dimensional questions like “Do you think President Obama should follow Lou’s advice and send troops to secure our border with Mexico?” In fact, any web surfer is welcome to vote as often as (s)he pleases!

The next step is to gather all of the responses and record them in a data base, perhaps by employing any of the various web applications that are freely available online. Survey Monkey, for instance, allows users to create and maintain a wide variety of sophisticated surveys. And WordPress allows users to post simple poll questions directly on their blogs.

A Sample Illustration

The final step is to analyze and interpret the results. Let’s assume, for instance, that we ask five people a set of three questions: (a) on a scale of 1 to 5, is the cost of our product the most important determinant of your buying decision?, (b) is the quality of our product the most important determinant?, and (c) considering the level of cost and quality that we are promising to deliver to you, will you buy our product?

Now let’s assume that these five people respond to questions (a), (b), and (c) as follows:

#1: (a) 3, (b) 3, (c) 3

#2: (a) 1, (b) 3, (c) 1

#3: (a) 4, (b) 1, (c) 4

#4: (a) 5, (b) 5, (c) 5

#5: (a) 1, (b) 2, (c) 2

Although this data only presents a snapshot of their opinions at a single moment in time, it might still be helpful to know what our prospective customers would prefer: a cheap no-frills product or an expensive luxury product.

There is a clear answer to this question, one that requires a free statistical web application to quantify it. By using a correlation analysis applet,  we can learn that the correlation between the responses to questions (a) and (c) is an impressive 97%. But the correlation between the responses to questions (b) and (c) is a meager 32%. So the answer is quite clear; far more of our consumers’ purchase intentions are correlated with cost considerations than with quality considerations.

Of course, these very consumers might change their minds tomorrow … as they often do when they consider their voting selections, their automobile purchasing decisions, and other significant choices. Nevertheless, if our time horizon is restricted to simply surviving the day, such information may be all that we want or need to live to see tomorrow.