|
||||||||||
|
Richard Alonzo Fyfe
The types of surveys available are extremely wide-ranging. One type of survey consists of the registration cards you get any time you buy a large-ticket item (like some new computer software). All of those questions: age, salary, where did you purchase this product, why did you purchase this product, etc. - that is a marketing survey.
Some surveys (or, more precisely, data-collection techniques) are subtle. A company will buy a mailing list from, say, a magazine and send out a mass mailing. That mailing will use a unique product code that signals whether a respondent is answering this mailing or some other (built from some other magazine's mailing list). As the company processes the orders, it records the purchases - comparing the numbers and types of units sold to the demographic profile of the magazine's subscribers. This is just an example. But you are communicating a lot of information about you, depending on the product code you use - which depends on how you got the product code - which depends on facts about you.
There is also the "drop your business card in the fish bowl for a drawing" type of customer survey. The company notes the name, rank, and occupation of its primary customers - so it knows where to target advertising (or where it needs to change its image).
Sales are a marketing survey of sorts. "If we lower our price X%, what will be the change in volume? Does it justify lowering the price permanently?"
The types of surveys I'm most familiar with involve, as I said, drawing up a list of huge list of questions concerning customer preference (with questions added aimed at discovering the reliability of answers to other questions and the degrees that free-ridership or free-drivership have influenced the answers. These are used for really big issues (including, in some cases, whether or not to call it quits, close the company, and try to get the best deal on its assets - or whether or not to buy a company and start expanding in a whole new direction).
Reasonable accuracy requires getting hold of around 1000 people, carefully picked so as to be willing to spend a half hour or so answering questions. They are heavily labor-intensive, and should be done only by a trained expert (which means paying an expert's salary, unless somebody wants to volunteer). These surveys cost hundreds of thousands of dollars. It's the type of survey I would recommend having before deciding whether the first trip should be lunar-orbit, manned landing, or habitat-landing.
And there are surveys that involve bringing certain types of customers in for a very detailed discussion of some product. These survey participants are generally paid - $15 to $30 per hour - for their time, and walk away (often) with about $50 in their pocket. These surveys often involve 100-500 people.
This is just a really quick overview of some of the types of market survey techniques used. Some are, obviously, far more expensive than others.
|
|
|