Segmentation Pitfalls

A segmentation is one of the most essential market research exercises.

As segmentation can impact strategy and organization, as well as how people build and market products, the success of a segmentation project depends not only on research design and execution, but also on buy-in, stakeholder engagement, and a researcher’s thorough understanding of the client organization.

There are a few very important things to consider when planning and conducting a segmentation. We’ll provide you with the ‘DECK’ system that consists of 4 key areas you should pay special attention to throughout the process:

1) Design

Make sure the study objectives are clear. What is your primary goal for the study? Who will work with the segments and what will the results be used for? And consequently, will the questions offer the results you need?
On a related note, be sure not to have too many objectives. It might seem like a good idea to gain as much information as possible in a single swoop, but it can also lead to blurred lines within the results. If too many objectives overlap and lead to confusion, results might not be of any help but rather complicate matters. Instead, try to come up with clear hypotheses of how the segments might look in order to be relevant to your everyday business. These will help building a study design that includes essential questions and excludes irrelevant ones.

The easiest way to ensure alignment and good focus during this critical phase is to do a workshop on objectives, hypotheses and dimensions early on in the questionnaire design phase - with all important stakeholders atone table.

2) Execution

While fielding, try to avoid sampling bias as much as possible. A wide variety of people is needed to identify distinct segments – as a rule of thumb you should question at least 200-300 respondents per segment.  
When it comes to measuring responses, you’ll be looking for differences and variance in the data to find certain patterns describing these groups. By selecting the right question types, you can minimize response style to measure real differences instead.
Another pitfall which will end up being a major challenge for any segmentation is a simple lack of analytical skills. So, if you’re using K-means instead of advanced techniques like Latent Class, the segmentation might not yield the results you hoped for.

During the execution phase, remember to choose the right question types from the get-go and prepare yourself for very thorough data cleaning.

3) Communication

Communication is crucial in order to avoid internal resistance or a lack of buy-in across stakeholders.
Here, it is critical to communicate a realistic time frame. Only if there are shared expectations about the time the survey design and data analysis will take can the segmentation proceed smoothly.
A typical communication mistake is not explaining where segments were derived from. Make sure to establish clear base dimensions in an early workshop and only introduce segments that can be logically connected to these dimensions so as to not confuse others trying to understand the segments.

Before settling for a segment solution, gather all stakeholders involved in the project for a results workshop and make the decision together with them. It’s the easiest way to avoid this pitfall.

4) Knowledge

This key area is closely related to the 3rd point, as poor communication might turn out to be a hindrance when it comes to collaboration. When the team isn’t on the same page about the project and its execution, problems are bound to occur.
Ingrained habits or beliefs might turn out to be another stumbling block here, as one should never make assumptions based on simple beliefs and generalizations, but only on the data.

If you find out that the derived segments can’t be mapped to the data mart after everything is said and done, it’s already too late to restart the whole process. You can dodge this bullet by discussing early on how important data mart mapping will be for the project.

 

These are just some of simple things to keep in mind when planning or conducting a segmentation. If you want to get a deeper understanding of the power of segmentation, are looking for a partner to conduct such a project with, or have a simple question, we’d be happy to assist!

Back

You may also like

Blog Post

Company Outing 2019

For two days of team building and group activities, we left Berlin, just to admire another beautiful big city: Hamburg!

Read More
Blog Post

Webinar on the Future of Name and Concept Testing

Get a glimpse of the future of name and concept testing during our upcoming webinars on marktforschung.de (June 13th) and NewMR (June 20th).

Read More
Blog Post

New directions for open-end analysis

Read our newest article on using Text Analytics for open-end analysis on Research World.

Read More