A survey takes place when a fixed set of questions is asked of a sample of customers. The sample is usually large and in many cases is carefully selected to represent the total population of customers. The comments that follow assume a telephone survey executed with a reasonably large and carefully selected sample in which the questions are largely descriptive.
Surveys can play a supporting role in environmental scanning. If you need a fairly exact factual description of the behaviors and simple perceptions of some customer group, and if such data cannot be gleaned from existing secondary research, then it may make sense to execute a survey. If, however, good secondary data already exist, it is rarely cost-effective to do your own survey, unless this takes the form of a small, fast, tailored survey directed at filling in a few gaps in the avail- able secondary data. If the needed secondary data do not exist, and if you simply must have precise descriptive data on such matters as the frequency of certain applications among particular customer groups, or the average dollar amount of equipment purchases, or the average rating of your speed of response relative to key competitors, then a survey may make sense.
You should ask yourself, however, whether you really need precise descriptive data at this early point in the decision cycle. Is it really that important to be able to state with precision that 54% of the time this medical instrument will be used on auto accident victims, 24% on mothers undergoing childbirth, 18% on victims of gunshot wounds, and 4% with others? At this early point, what is the value added by these precise percentages, as opposed to what you could gain from a program of customer visits? A couple of dozen visits would probably reveal that auto accidents, childbirth, and gunshot wounds were “major” applications, even though the exact percentages would be unknown. In addition, and in contrast to the limited data supplied by a survey, the visits would provide opportunities to describe in depth how each of these applications place different demands on the instrument and on hospital staff, how this instrument interfaces with other equipment in the hospital, and so forth. Such rich descriptive data are often more useful, early in the decision cycle, than the thinner but more precise data yielded by surveys.
It is even more important to understand that surveys are far less useful in the generation of options than customer visits or focus groups. The relative weakness of surveys at this point in the decision cycle has several sources: (1) the fact that the questions to be asked are fixed in advance; (2) the reality that the phone interviewers who will implement the survey probably lack the ability, the motivation, or the opportunity to deeply probe customer answers; and (3) the unfortunate truth that the impersonal nature of the survey contact-the certain knowledge that one's responses are but grist for the statistical mill-will inhibit and limit the customer's investment of the energy required for discovery, exploration, and depth. Surveys are a confirmatory tool whose proper purpose is to limit, narrow, and specify; hence, this tool is largely incapable of expanding, broadening, and re-configuring your understanding. Go easy on surveys early in the decision cycle.
Survey research comes into its own at the third stage of the decision cycle. All of the features that had been of dubious relevance or even liabilities at the earlier stages are here either neutralized or converted into strengths. In stage three the time for discovery and in-depth insight is past; now it is time to make hard choices and allocate limited resources. Perhaps you only have the resources to write new software for one or at most two of your instrument's applications, and you must determine which application predominates. Large investments may follow from decisions of this type, and it makes sense to invest a good sum of money in determining precisely which application is largest, or is growing the fastest, or has the weakest competitive presence.
Survey research is also of primary importance in the evaluation of outcomes. The classic example is the customer satisfaction surveys now conducted by many firms. These are usually telephone surveys, often conducted by a neutral outside firm, in which a standard series of questions is asked, focusing on product and vendor performance. The surveys are often repeated on a quarterly basis so that changes in satisfaction can be tracked over time. Another example is the tracking studies conducted after initiating an advertising campaign. These telephone surveys track awareness, brand attitude, and perceptions in those areas addressed by the advertising campaign. Here again, descriptive precision is an absolute requirement; otherwise, comparison over time becomes impossible.
Reference: “The Market Research Toolbox” Edward F. McQuarrie
Comments