- New JobMondelezCI Engineer
A lot of time can be saved in the planning stages of a survey research project by simply answering three short questions. By the time people from organizations have contacted me about a Six Sigma voice of the customer (VOC) study or research project, they have usually spent at least a fair amount of time discussing the research. Yet, when we talk about the project, it is unfocused and often the research originally contemplated does not even meet the company’s real needs. However, a VOC or survey research project can quickly get on track by answering three short questions.
Research requires focus or specificity; otherwise, it lacks direction and may result in costly miscommunications. Typically, VOC research answers a question such as how satisfied are customers with our services or how many people plan to purchase a new computer in the next year. Often, however, the question is too broad resulting in data that is not what the users of the research expected or needed. To ensure everyone involved in the project is on the same page, the research needs to be focused.
These three questions can help focus the research:
Answering these questions is a dialogue or process. It may turn out that in answering question 2, the answer to question 1 must be revised. Or you may know the answer to question 3 before question 1. A focused research project or VOC study will emerge when these three questions are constantly being asked and revised.
Below are two real-world examples of how these three questions are essential to defining a VOC study or survey research project. The first situation is straightforward and the other scenario provides a somewhat more complex project.
What do you want to know?
Acme International recently acquired a small company that produces a related but different product and a larger company that had been a competitor. One acquisition was welcome although it meant some internal restructuring; the other required government scrutiny and some negative press about Acme resulted. Additionally, productivity is flat and a few highly skilled and valued employees have left the company.
Management is concerned about this situation and decides it needs to gauge employee morale. For now, the obvious answer to “What do you want to know?” is “What is the condition of employee morale?” But the question is too general – too open to individual interpretation – and requires some probing such as “Why is company morale so important right now that spending thousands of dollars and a lot of time to conduct research is warranted?”
In answering this question, a management committee formed to deal with this issue learns that Acme’s president has noticed some signs among employees that the company’s progressive corporate culture, which he engendered and considers responsible for the company’s success, is not working well. The company has turned into a rumor mill, for example, that has under minded relationships and caused a great deal of insecurity. The president wants to bring employees from the acquisitions into the Acme fold to create a unified company. Acme also does not want to lose any more of its skilled employees.
Who do you want to know it from?
Seems obvious in this case: the employees. However, does Acme want to hear from all employees from the janitorial staff to the vice presidents? Acme has many sales people out in the field and they are an independent group not prone to answering a survey. Should they be queried? Also, Acme is global. Does it want to hear from employees in all offices whether in India, London, or New York?
The committee decides it is not nearly as concerned about the morale of unskilled workers as it is about well-trained employees and top sales people, so there is little reason to ask the unskilled workers their opinions if the information will not be used. It might be nice to know but it is not essential to this research project. Therefore, particular grades or levels of employees can be identified as the population that will be surveyed.
What are you going to do with the information?
The short answer to this question is, “improve morale.” But the survey should first reveal whether morale is indeed low, and if it is, the causes behind it as well as determine the best way to communicate with employees. Without knowing what is going on with its employees, Acme’s efforts to bolster morale may be wasted. For example, the research may find that there are indeed many rumors, that employees feel uncomfortable asking their managers too many questions, and that virtually all employees constantly use email and the Internet but few read the company newsletter. To counter rumors, a communication strategy can be developed such as an Intranet site with an online question and answer session where employees email their questions anonymously to an ombudsman or upper management, or a DMAIC (Define, Measure, Analyze, Improve, Control) project can be initiated to address issues. Company managers can now act effectively to counter low morale. Without this understanding, Acme would be shooting in the dark.
The research project is now defined well enough to proceed to the next step in the survey research process. The answers to these three questions narrow the scope of the project so that researchers understand what types of questions to ask, and management knows what kind of data they will obtain from the research. The survey research process will still take time as the survey instrument is finalized, but the three questions swiftly move the research process along particularly when it gets bogged down in committees.
Acme’s situation was relatively simple. A more complicated scenario follows.
Example 2: Increasing Sales
Acme Telecommunications Inc. produces cell phones and wants to increase sales. Again, a committee has been formed to begin strategizing. At this stage their question is, “How can we increase sales?” This is a broad topic but acceptable for now.
Who do you want to know it from?
Although this question forces Acme to focus on various ways to increase sales, committee members have very different ideas about how to do it. They identify three groups: people who use a competitor’s product, those who have never used a cell phone, and current customers. Acme needs different information from each of group, which brings the project back to question 1.
For instance, researchers might want to know what people like least about their competitors’ products so that they can capitalize on their shortcomings. On the other hand, researchers will want to know what keeps people from using cell phones, and they may want to find out how their cell phone could be more attractive so that current customers will want a new one. The research is narrowing down but it is still too general and requires more probing.
Acme knows that 40 percent of adults in the U.S. have never purchased a cell phone, which represents a huge, untapped market that could significantly increase profits. The company, however, has no idea if it is realistic to expect different demographic groups to even consider purchasing a cell phone. For example, will a 70-year-old retiree ever use a cell phone? Some initial research needs to determine the age, ethnic and income groups that are most likely to try cell phones.
Next, Acme knows the market share of each brand of cell phone. How many people can they reasonably expect to switch to their product? They need more information on customer loyalty and specifics about what customers dislike about competitors’ phones to answer this question.
Acme, of course, knows how many phones it now sells and wants its customers to buy new ones even though they may not really need one. To begin a strategy for that segment, the company may want to find out what, if anything, would prompt the purchase of a new one.
What are you going to do with the information?
The organization wants to use the research findings to initiate a new marketing campaign. By knowing that the company is in the early stages of creating this campaign, the researcher can recommend a research design and determine what questions to ask each group.
In this case, the researcher might first recommend a statistically valid survey. The survey initially asks about cell phone use and separates the respondents into two groups: current users and those not currently using a cell phone. Current users can be further divided into those who use Acme’s product and those who do not.
The cost of a survey is tied to how long the questionnaire is, how easily the people the survey wants to question can be reached, and how many people are needed to provide the necessary data.
To generalize from a survey to a population, a rule of thumb among survey researchers is that a minimum of 100 answered questions is needed in each subgroup. Therefore, if Acme wants to know whether men or women who have never used a cell phone differ in their willingness to use one, at least 100 men and 100 women who have never used a cell phone must be questioned. Or, if the company wants to know if willingness to use a cell phone varies among age groups, each age group must contain at least 100, and so forth. If fewer than 100 are used in each group, the margin of error is so high that the results can present a false picture.
If a person is currently using a competitor’s cell phone, the questions might focus on the brand of cell phone, why they purchased that particular cell phone, and what they like and do not like about it. If a respondent has never used a cell phone, the questions would center around why that is and the circumstances under which they might be willing to use one. And, of course, if the respondent currently uses Acme’s brand of cell phone, the question might focus on their satisfaction with the product and what new features they want from a cell phone. The data will then be analyzed according to these three groups.
By going through the exercise of answering the three questions, the research is focused, the methodology is identified, and Acme knows what it can expect from the data.
As is often the case, the information from the survey may lead to more research because it has only partly answered the overall question about how to increase sales. This research provides the foundation to learn how to approach people who have never used a cell phone. For example, if the research shows that women ages 50 to 65 that live in rural areas are open to using cell phones, the next research study can look at how to talk to them. This topic may require a different methodology such as focus groups made up of the demographic groups the company wants to target. The procedure – the three questions – should again be asked to define the project.
Research is most useful when it is clear what information is needed. If you do not know specifically what you need to learn, you risk not learning anything useful. These three questions help focus the thinking and discussions.