Perceptual mapping and Multidimensional scaling are the two terminology which are linked with each other. Brand positioning depends on the strategy and marketing procedures which enables a brand in the customers mind depending on the customer needs and brand attributes. To see brand positioning compared to competitors, various techniques are being used like perceptual mapping, preference mapping and joint space mapping. Perceptual map is one of the popular methods for perceptual mapping, could be done depending on similarity-based methods or attributes-based methods. On the other hand, preference maps could be generated using two methods called, ideal point model and vector model. If the two types of mapping are applied at a same time than it is called joint space maps which could be done by external analysis and simple joint space maps. Multidimensional scaling is type of unsupervised machine learning and one type of perceptual mapping technique to visualize multiple variables in the form of two dimension.
Table of Contents
What is Perceptual Mapping:
Perceptual mapping is used to know the positioning of brand in an industry. It is mostly used by the marketers depending on various factors or attributes like price, benefits etcetera. There are various techniques are available to create a mapping based on customers perception. It could be two dimensional where two factors are considered and could be multi-dimensional where multiple factors or more than two factors are used. Moreover, it is used to make marketing strategy as per present brand positioning and customer’s perception about the brand. It also helps to set objective, mission and vision of a product or product line by judging market situation. Pictorial representation of perceptual mapping has been shown below with bubble chart where bubble size represents the market share and each bubble represents each brand. Any two attributes could be considered in X and Y axis or in case of Multidimensional Scaling, multiple attributes could be brought down into two dimensions and plot in the two-dimensional graph.
Usefulness of Perceptual Mapping in Marketing/ Market Research:
There are many usefulness of perceptual mapping that’s why marketers are preferring the method.
1. Visualize the brand position compared to competitor brands
2. Two or more variables can be used as an influential factor for brand positioning
3. Understand the view or perception of the target customers on a specific product
4. Measure the impact of marketing campaign by comparing two perceptual mapping as before and after.
5. Segmentation of products depending on one or more variables. Also, segment the brands based on the competitive positioning.
Types of Perceptual Mapping:
Perceptual mapping is considered as a technique which represent a data into visual display by diagrammatic representation. By the perceptual mapping, positioning of brands, products can be seen within the market as compared to the competitors.
There are various types of tools which are being used to do the mapping,
Factor analysis– Factor analysis can be described as similar factors or variables which influences dependent variable or business outcome. So, in the factor analysis method, the relevant factors getting find out or evaluated.
Multidimensional scaling– In the multidimensional scaling method, multiple product attributes has been considered to generate a visual map where multiple brand positioning could be seen.
Discriminant analysis– Discriminant analysis is initiate to find out groups depending on the discriminant function. Moreover, the relation between one or more variable being established.
Spider chart– Spider chart is also used for perceptual mapping where multivariate data is represented through a chart called spider chart. The spider chart is looks like a radar.
Multidimensional Scaling:
Multidimensional scaling is to plot multivariate data by a diagrammatic representation to visualize the whole data set.
There are two types of multidimensional scaling, nonmetric multidimensional scaling and metric or classical multidimensional scaling.
Metric data and Non-metric data concept
At the time of market research or when try to measure or to get insights, we talk about types of data. Broadly there are two types of data, qualitative and quantitate. Qualitative data is represented by expression, word etc. On the other hand, quantitative data is represented by numbers or numerical data. Quantitative data can be classified by metric data and nonmetric data. Metric data is defined by scaled data which is further divided by discrete/interval and continuous. On the other hand, nonmetric data is further sub divided by ordinal, nominal and binary data.
Multidimensional Scaling in R:
R is the open-source programming language or statistical software which is used to do statistical analysis. Various packages and functions are available to compute and get output of statistical modelling. In this case, to do multidimensional scaling, R software have been used to create the model.
Steps to create MDS:
Distance matrix– Distance matrix can be categorized by metric distance and non-metric distance. To create multidimensional scaling, it is required to determine distance between points where distance matrix helps to determine the same. It is also used to do clustering depending on the shortest distance to the points which represents brands.
CMDSCALE/ ISOMDS function– By using “cdmscale” or “isoMDS” function, multiple dimensions got accumulated into two dimensions and the data can plotted.
Plot the data– The objective of MDS is to visualize the data in the form of two dimension
Classical/ Metric Multidimensional Scaling in R
By using cmdscale function, classical MDS will be performed.
To find Euclidian distance:
d<- dist (data)
MDS:
fit<- cmdscale(d,eig=TRUE, k=2) ## k denotes number of dimension
Visualize result:
fit
Plot the data:
x<- fit$points [,1]
y<-fit$points [,2]
plot(x,y, xlab= “Dim 1”, ylab=”Dim 2”, main=”Classical/ Metric MDS”, type=”n”)
Nonmetric Multidimensional Scaling in R
By using isoMDS function, nonmetric MDS will be performed.
Library to install:
library(MASS)
To find Euclidian distance:
d<- dist (data)
MDS:
fit<- isoMDS(d,k=2) ## k denotes number of dimension
Visualize result:
fit
Plot the data:
x<- fit$points [,1]
y<-fit$points [,2]
plot(x,y, xlab= “Dim 1”, ylab=”Dim 2”, main=”Nonmetric MDS”, type=”n”)
Create perceptual mapping in Excel:
In you want to create perceptual mapping in excel by using two variables, simply insert scatter plot or bubble chart. Follow the steps to create scatter plot or perceptual mapping in excel.
1. Select the data set with two or three variables.
2. Go to insert tab
3. Go to chart and select scatter plot of bubble chart from the list
Thanks.
