Methodologies

Medium for Survey

ComSim enables you to capture the sentiment of your customers via three channels – Phone, Web, and Hybrid surveys, each with its own set of advantages:

1. Telephone surveys: offer the benefit of human contact and personalized follow-up questions. These are especially effective for customer responses, because respondents can reply with minimal effort. Since these surveys are initiated by the interviewer, the response rate can be managed. Voice recordings are possible. The telephone medium works well with short surveys.

2. Web surveys: are conducted with large populations simultaneously and assure uniformity of the survey. They allow for the inclusion of logic, graphics, links and references and can accommodate long surveys and a variety of rating scales. They require the respondent to proactively decide to take the survey, and there is effort involved in entering verbatim comments; thus the response rate and volume of comments vary.

3. Hybrid surveys: are used to take advantage of the benefits of both methodologies.

Web followed by telephone – When an email invitation does not result in a completed survey within a pre-determined time period, ComSim can attempt to interview the contact by telephone.

Telephone followed by Web – When a number of attempts to reach a contact by phone yield no success, an email invitation to do the survey online can be sent.

Rating Scales

There is always a debate as to whether to use a satisfaction scale with a midpoint or not – for example, a 5-point scale, which has a midpoint of 3 or a 4-point scale that requires the answer to come down on one side or the other. No matter which philosophy is chosen, ComSim follows these guidelines:

1. The scale should be easily understood, for example: Very satisfied to Very dissatisfied.
2. The intervals between points on the scale should be equal or perceived as logical progression: for example, Excellent, Good, Fair, Poor.
3. It is preferable to use words for points on the scale, rather than solely using numbers, to avoid confusion. An anchored scale could have 10 as Critically important and 0 as Not at all important.
4. Too many points on a scale may provide more granularity than is warranted. This may be the case when certain points on the scale are seldom selected as answers.
5. Some questions do not require a rating scale but are Yes/No or multiple choice.
6. The Likert scale of Strongly Agree to Strongly Disagree is generally reserved for opinion surveys rather than surveys following an event about which the customer has specific information.
7. The Net Promoter Score®1, as defined by Fred Reichheld, uses a likelihood to recommend scale from 10 to 0, where 10 represents Extremely likely and 0 represents Not at all likely.

Question Types

1. Rating – These questions can ask the degree of satisfaction or the quality of service or product on a pre-defined scale.
2. Open-ended – ComSim uses open-ended questions to ask for suggestions for improvement, product ideas, and competitor comparisons. Customers are more comfortable with open-ended questions later in the survey. Topics can be coded in order to tabulate them.
3. Multiple choice – Multiple choice questions work well in web surveys, where the respondent can see all the choices. They can be used in telephone surveys if the choices are not too lengthy.
4. Binary – Yes/No answers are an example. These are frequently used to determine the survey path; for example, “Did the field service technician solve the issue on the first visit?” If the answer is no, the next question can ask about the total length of time before the issue was resolved.
5. Ranking – These questions are used sparingly to measure such concepts as preference or importance.
6. Follow-up – These questions are posed in order to obtain greater detail and clarity. Elaborations by respondents often provide the narratives that support their rating choices. These are a hallmark of ComSim interviews.