Break out your notebooks and no. 2 pencils, because, in marketing class, this exercise is what we refer to as segmentation. For our second installment in the SurveyGizmo Case Study, we’re going to walk through the process of defining segments. To get caught up, check out installment 1 of the SurveyGizmo Case Study over here.

The motivation for this is that every time Scott (our CEO) does an webinar or demo, he inevitably fields questions about target customers. It sounds like an easy task (surveys are pretty universal), but it’s actually more complicated than it seems. Some of our clients are large corporations that use SurveyGizmo to measure customer satisfaction, conduct market research, gather new leads by integrating SurveyGizmo with SaleForce.com (check out tutorials here and here), and just about every other “traditional” survey you can imagine. We also have customers who are conducting their first survey ever and are looking to SurveyGizmo for technological and methodological guidance help make it easy. In a completely different segment, we have a strong group of academic researchers who use SurveyGizmo for conducting survey and experimental research studies, setting up student tests and quizzes, and even for teaching marketing research. And sometimes, people just need a quick way to ask a group to vote on options for an upcoming company outing.

SurveyGizmo can do all of these things (and a lot more), and that’s exactly why it’s so hard to pin down a target audience. Because SurveyGizmo is so versatile, Scott usually explains that SurveyGizmo is for “anyone who wants to design a great survey!” While that’s a true statement, it’s not exactly satisfying. So as part of our customer survey project, one of the big goals is to figure out who uses SurveyGizmo and how their needs differ as a function of their unique circumstances.

Our end goal in doing this is to better understand each segment’s unique (and common) needs, so that we can tailor our marketing/advertising efforts to reflect the needs of each segment. This will allow us to position SurveyGizmo appropriately for different marketing and advertising efforts based upon the needs of the target segments (more on this in future posts).

Our first step in segmentation is to think about how survey users might differ. We came up with four variables, each with two levels.

Variable (levels)
Sophistication Needs/Features (basic, advanced)
Number of Surveys (few, many)
Number of Responses per survey (low, high)
Survey Length (short, long)

This gives us a conceptual space in which to start placing our users. Take a look at how we filled out the chart below.

surveygizmocasestudyimage

Even though we identified four variables which could have given sixteen specific groups (2^4), once we started plugging customer types into the spaces, it turned out that the variables were flexible guides rather than hard and fast barriers allowing us to combine cells. This gives us plenty to work with without being overwhelmed by detail (being overly focused can be as bad as being too general).

As a reflection of our analysis and noodling over this problem, the schematic reveals that users choosing SurveyGizmo for advanced features probably aren’t doing a few short surveys. Additionally, it is evident that the number of responses likely correlates with the length of survey and that customers with advanced needs might be choosing SurveyGizmo for different reasons. At the same time, it is also clear that basic users occupy all areas of the space (note that you could imagine/arrange the advanced and basic users tables as three dimensional spaces with axes corresponding to number of surveys, number of responses, and length of surveys).

In the end, we came up with three relatively broad segments: Advanced-Short, Advanced-Long, Basic. We can now start thinking about the different things we want to know from each segment that we should build into our survey.

The ultimate goal of a survey is to understand how groups differ based upon defined characteristics. By going through a segmentation exercise, you can anticipate the types of questions you should ask in your survey to help you better understand the different needs of each separate group. This makes for better data and a better use of respondents’ time since you’re asking the right questions to the right customers.

Extra Credit
For those who want to go more in depth on segmentation, Chapter 1 in Kellogg on Marketing is a great resource. Links below for the Google Book (some pages missing) and Amazon where you can order a full version to read on your next plane trip.

Google Book — Kellogg on Marketing (some pages missing)
Amazon — Kellogg on Marketing

Tags: , ,