Market Segmentation Using Cluster Analysis


Customer service professionals will tell us that each individual is different and therefore needs to be treated accordingly. While this is definitely true, retailers and marketing personnel cannot service this individuality when they are marketing mass products. Competition has made it necessary to ensure that your product is not a jack of all trades. Differentiated products for the specific needs of different kinds of people are a necessity today.

Market segmentation is a process that has helps in splitting and grouping customers into various clusters. The customers belonging to a single group are considered to be homogenous and comparable. A market segment is homogenous within and heterogeneous without. If the marketing segmentation has been carried out properly, a market segment is expected to react similarly to a market stimulus.

According to Philip Kotler, an effective marketing segmentation exercise should result in segments that are measurable, relevant, accessible, distinguishable and feasible. These are aspects that ensure that the marketing segmentation has been done on relevant parameters and that further action on these groups is practical.

Marketing segmentation helps marketers understand the otherwise extremely heterogeneous market so that the strategic and tactical marketing decisions can be taken in accordance with customer needs and desires.

Understand the market

At the first level marketing segmentation helps in a better understanding of the market in relation to the industry in question. When each individual is grouped, it is easy to understand the common characteristics of the group rather than assessing each individual customer (something that is quite impossible too) or treating everyone as a single group with similar needs and preferences.

Create product variations effectively

Once you have a certain number of groups that define your market, product variations can be launched depending on the specific needs of the group. So while those who treat a bath as a time to pamper themselves would buy soap with a moisturizer added in, those who feel that a bath is for cleansing may well prefer to have one that fights germs. An insight such as this can help you develop relevant products for specific people.

Perfect your positioning

Not only does marketing segmentation help in creating new products but it also aids in the creating of the appropriate positioning. The words used in the body copy and tag line of the advertisement can be modified based on the expected reaction of the targeted demographic too.

Create customer empathy

With targeted and more relevant products and positioning, the marketer can be sure to create an empathy with the customer; a bond that is difficult to break, thereby encouraging higher customer retention.

Identify need gaps

Conducted for exploratory research, marketing segmentation can allow you to identify need gaps in the market so that the research and development division can be given specific parameters to work with. Sometimes the gap identified can also be used for repositioning of an existing product.

Focus your resources

When the resources are limited, identification of the relevant market segments that are showing growth can be used to ensure that the budgets are allocated in an appropriate manner.

Targeted media selection

An understanding of the customers in each segment can help in better media selection, media scheduling

Using Cluster Analysis for Marketing Segmentation

Cluster analysis is a multivariate statistical tool that is used for marketing segmentation. It helps in identifying groups of people from the sample that have similar responses with respect to any of the parameters that are used for clustering.

Cluster analysis can use any data that has been collected to segment customers into various groups. Some of the broad parameters that can be used are:

  • Behavioral parameters like purchase history, amount of product bought, product usage details, brand loyalty, shopping behavior and more.

  • Attitudinal parameters that can be used in cluster analysis include (among others) reaction to various belief statements, likes and dislikes measured on a scale and general tendency towards innovation.

  • Psychographic variables like lifestyle, values, personality and more can also be obtained using pre tested scales for the same. Customized ones can also be developed depending on the category.

  • Demographic data is a key part of the analysis because irrespective of whether it is used in the cluster analysis and segmentation analysis or not, this data linked with each respondent helps in identifying the kind of target audience that you are looking to cater to.
With Mineful all that you need to do is the upload your database with the survey details and the clustering will be done with a few clicks. You can expect to get the specific clusters, cluster sizes, cluster naming and a detailed understanding of each of the clusters in terms of the relevant parameters. Not only will you get an understanding of each cluster but also the sketch of a typical customer in each group.
Read On

Survey Examples: New Concept Evaluation Survey


Before launching a new concept or product, marketing professionals first test it with a selected audience. A new product survey provides a way for marketers to understand when to start, modify existing ideas, how and who to market. Evaluating new products with online questionnaires provides a cost-effective way to obtain valuable information from the the people who matter most, the customer.

Businesses can take advantage of knowledge gain from online marketing surveys such as the new product survey to:
  • Understand which benefits are most important to customers
  • Which features are essential in delivering the product's promise
  • Identify customer's needs
  • Measure advertising and placement factors
  • Understand how incentives, warranties, and bundling change price and value perceptions

If the concept being tested is for a new product, most of the responses are expected to be based on a perception of the product after reading. There are various theories on whether or not to reveal the company and the brand when exposing a concept or not.

New Concept Evaluation Survey

  • Overall likeability
  • Intention to purchase product after reading concept details
  • Intention to recommend
  • Credibility
  • Uniqueness
  • Perceived quality based on concept
  • Key message recall
  • Perceived benefits of the product or service
In all the studies mentioned above, capturing relevant demographic data is critical. Keeping a track of all the information areas that need to be covered in a survey can get extremely tough. To make sure that you do not falter and lose precious time and money in re-fielding a survey, check out this New Concept Evaluation Survey example from Mineful.
Read On

Survey Examples: Brand Awareness Survey


While there are a large number of studies that can be covered in the market research area and each survey will almost always have some unique survey specific questions, there are some areas of information that are common to brand awareness surveys.

The following matrix of information areas allows you to calculate the awareness, penetration, trial ratio and repeat ratio. These parameters can help in assessing the success of a new product launch.

Brand Awareness Survey
  • Brand awareness in terms of top of mind, spontaneous and aided recall
  • Claimed ad awareness for the brands
  • Usage details – current usage, past 3 months, lapsed usage, repeat purchase
  • Purchase details – when bought, quantity purchased, store bought from and the like
  • Performance of product on specific attributes among users
  • Diagnostic evaluation among users
  • Reasons for non usage among those who are aware but have not tried the product
See our Brand Awareness Survey example.
Read On

Survey Examples: Product Evaluation Survey


Typically current brand ratings in a product evaluation survey are captured before the product is placed for usage. This is true especially if it is a new product or a modified version that is slated to be introduced in the market. The diagnostic ratings can help in tweaking some of the elements of the offering. Key survey questions to have in a product evaluation survey that will bring insight and actionable information to better the product, offering, messaging, and features include:

Product Evaluation Survey - Questions/Topics to Have
  • Current brand/product usage details
  • Overall evaluation of current brand
  • Overall performance rating
  • Overall satisfaction
  • Likelihood to recommend
  • Likelihood to use again
  • Purchase intention without a price
  • Reasons for the same
  • Performance rating on specific product attributes
  • Diagnostic rating for specific attributes (too sweet, just right, too salty)
  • Comparison with current brand
  • Purchase intention at a price
  • Open ended questions to understand specific likes and dislikes
See our Product Evaluation Survey example template from our survey library.
Read On

Survey Examples: Customer Satisfaction Survey


In a customer satisfaction survey, typically the overall questions are used to arrive at a customer satisfaction index so that a single number can be used to compare products, stores, brands, or companies. If previous experience is lacking, one may need to arrive at the specific parameters of evaluation after conducting a small qualitative research with experts and customers in the industry. For example, broad areas covered in an employee satisfaction survey include compensation, work environment, policies, relationship with superior, estimate of future growth and the like.

A customer satisfaction survey may include aspects like responsiveness, attendant behavior and demeanor, performance, billing and more. A box plot of claimed importance and derived importance (calculated by regressing the performance of specific attributes with overall satisfaction) along with the performance on each attribute can actually tell you the whole story.

Customer Satisfaction Survey
  • Overall satisfaction based on a typical 5 point scale
  • Likelihood to recommend on a typical 5 point scale
  • Likelihood of repeat purchase on a typical 5 point scale
  • Assessment of competitive advantage
  • Claimed importance of various parameters on a scale
  • Assessment of performance for client brand/company and competition for the same set of attributes used in the importance question above
You can access any of the following customer satisfaction survey from our survey library, check them out:

- Customer Satisfaction Survey
- Customer Service Survey
- Customer Satisfaction Survey - for Service Firm
Read On

Survey Examples: Key Questions to Have


Over the next two weeks, we will have a series of posts that go in depth about Key Questions to Have in specific survey examples. In this post, we'll introduce the topic and cover key steps in preparing to do research with a questionnaire.

The analysis that you do on the data collected is as good as the specific questions that you have asked and covered in a survey. It is therefore pertinent that the survey examples be crafted in a manner that covers all details that are required to analyze the data once all the responses are in. Survey examples help, but each research has its unique properties.

There are various newcomers in the area of market research that realize after the survey has been completed and closed that a certain piece of information has not been covered in the questionnaire. With no data on the parameter, certain kinds of analysis may not be possible. Simplistically speaking, if you have not asked the importance of a specific product feature understanding how importance varies by customer segment may not be possible.

It is therefore pertinent that the objective of the study be internalized completely. Most seasoned market research professionals decide on the specific analysis that they will perform to confirm or negate the marketing hypothesis before the launch of the survey. This allows for modifications to me made of required at a stage where no harm has been done. A what-if scenario, therefore, needs to be built in.

Key Steps in Preparing a Survey Questionnaire
  • Go through the specific market research objectives
  • Understand the broad areas of information that need to be asked like usage behavior, purchase details etc.
  • Detail these further and jot down specific information points under each of the areas
  • Make a rough analysis plan detailing the kind of analysis that you will want to carry out.
    This includes simple univariate analysis of the frequencies and averages of the variable distributions along with the cross tabs and multivariate analysis that you may want to look at.
  • Make sure that the format of the data that is being gathered is conducive to the kind of analysis that you envisage. This involves an understanding of the kind of data (ordinal, nominal, interval or ratio) that is required to run some specific analysis.
  • Revisit the information areas and ascertain that you have covered everything.
  • Move on to devise a questionnaire flowchart to ensure that the right respondent is answering the right question. This can help you in formulating the filters and the skip instructions better.
  • Word each question in the questionnaire.
  • Test the questionnaire on a dummy respondent and script it for an online study.
  • Make sure that you test the link before launching the survey.

Survey Examples and Information Areas for Each

While there are a large number of studies that can be covered in the market research area and each survey will almost always have some unique survey specific questions, there are some areas of information that are common to specific survey examples.

Four of the most common market research studies that are conducted in the industry include a customer satisfaction survey, product evaluation studies, brand awareness studies and new concept evaluation studies. Here again, the questions may differ based on the kind of product test or concept test that one carrying out. For example, a sequential monadic product test will have a different flow to the questions asked as compared to a monadic design.

Stay tuned as we will deep dive into each of these four survey examples and provide you with specific and key questions to have in each of these survey examples.
Read On

Apple iPad: Industry changer or another overpriced Apple fanboy product?


Mac Users Like It, Of Course
According to our non-scientific survey of 48 respondents, those who consider themselves to be a "Mac" are more likely to purchase the iPad than those respondents who consider themselves a "PC". Mac users have the highest interest level in an iPad, followed by those who consider themselves neither a Mac or a PC. Mac users are significantly more interested than PC users at nearly twice the interest level.

39% of respondents indicated they were both a Mac and interested in buying an iPad this year. Only 14.7% of respondents were both a PC and plan on buying an iPad during this year. A total of 27.1% of total respondents indicated they would wait at least one year before deciding whether or not to purchase an iPad. 10.4% said they would never buy an iPad (all indicated they were a PC).

iPad 64GB Wi-Fi +3G is the way to go
The most popular configuration was the 64GB Wi-Fi + 3G iPad, followed by the two 16GB versions (both with and without 3G). Buyers are most interested in having 3G internet as indicated by the popularity of those models. People also indicating that "being able to connect to the internet anywhere" was the most important feature, narrowly edging ease of use and having all your entertainment options in one place.

It's Personal
The most popular reason for using an iPad? 91% said they would use an iPad for personal use (personal emails, calender, and web). Mac and PC users have similar views on the uses for an iPad except that Mac users are more likely to use an iPad as a music device than as an e-reader, while PC users would prefer to read on an iPad than listen to music. The 64GB 3G version was equally popular among all the choices due to its higher capabilities, while the bare bones 16GB with no 3G would be most popularly used for personal use followed by music and reading.

We Will Pay $829 to Surf the Web, in Cool Fashion
With all the excitement over the iPad and all it's capabilities, are consumers really ready to spend $829 to send emails and browse the web? According to our results, yes. Consumers are ready to invest money into putting life's simple pleasures and every day tasks into one device.

For the price of an iPad you can buy any one of the devices the iPad straddles the line between (laptop, smart phone, media player, e-reader, netbook), but not all of them. Even those who currently own most of these devices indicated they would still like to purchase an iPad. It should be noted, however, that 60% of respondents reported household income over $75K and 44% had over $100K, which may indicate more discretionary income to spend on items like the iPad.

PC Users Are Not Buying
So what is the biggest deterrent to purchasing an iPad? Apparently being a PC owner. While there is some enthusiasm, most PC owners are taking a wait and see approach to the iPad. Does this mean that Apple has introduced another niche product that it's loyal fan base will eat up, but most others won't? Or is this the next iPod or iPhone, a product that will take charge of the market and change the way we email/web browse/consume media? Patriots or Pinheads? You decide.
Read On

Political Research Gives Candidates an Edge


Elections place a heavy burden on mail carriers. Their bags are stuffed with political advertising telling voters why they should support this or that candidate. Most of this mail is ignored or discarded with barely a glance, mainly because it does not address issues that really matter to people.

It doesn’t have to be this way. With a little effort and a relatively small investment, candidates can conduct their own political research using simple survey tools to discover what people care about and what they are looking for in their elected officials.

Discovering What’s Important to People
Effective political surveys have a lot in common with surveys used for market research or any other effort intended to learn about personal interests and preferences. They should begin with an upbeat invitation, such as “Thank you for taking the time to make a difference in our community,” or “Thank you for letting us know what’s on your mind.”

Like other types of surveys, political surveys should be brief and easy to complete. The early questions should be pleasant and fairly simple. Questions about difficult or controversial topics should be put off until the end of the survey, when people feel more comfortable about expressing their opinions. If these questions appear too early, people may abandon the survey because they find it too challenging.

On multiple choice questions, it’s good to give people an option of “Don’t Know” or something similar. On other types of surveys, the Don’t Know response usually doesn’t provide useful information, but on a political survey it can be highly instructive. If a large number of people don’t know what they think about an issue, a candidate might decide that the issue is not very important to them. Of course there are other ways to determine what people care about most. The simplest is just to ask them to rank a number of issues in terms of their importance. Responses to such questions probably won’t lead candidates to change their positions, but they will help candidates decide which issues to emphasize.

The Online Advantage
In the past, conducting political opinion surveys was too expensive and labor-intensive for most candidates. The Internet has changed that. Online surveys yield results very quickly, often within a day or two after they are posted. Once a survey has been posted online, it costs very little to collect and interpret the results. Online surveys can use graphics, colors, and even video to make it easy and entertaining for people to complete them. And, in the Facebook era, many people will feel more comfortable sharing their opinions online than they would responding to a phone survey or even filling out a paper questionnaire.

Online survey software, such as Mineful’s products, can make it easy for candidates to get the most out of surveys. For example, candidates can use this software to sort responses by location or by political affiliation. They can also see how opinions on one issue might be correlated within opinions on another. Affordable, easy-to-use software tools can make it easier for candidates to deliver the right message to the right people.
Read On

Back to TOP