Consumer Choice of Mobile Phones for Multiple Purposes
Marketing Batch (M3)
Authors :-
· Pranay Bawane – 021331025300
· Suyog Honmane-021331025467
Introduction:
This report presents a professional statistical evaluation of your survey data, mirroring the analytical environment of IBM SPSS Statistics. The study utilizes two primary multivariate techniques: Factor Analysis to identify latent dimensions of user behavior, and Cluster Analysis to segment the respondent population into distinct profiles
Objectives:
· Identify Key Drivers: Determine if “Price” or “Usage Frequency” is the dominant factor in smartphone satisfaction.
· Market Segmentation: Divide the dataset into mutually exclusive clusters for targeted marketing strategies.
· Brand Positioning: Map which brands (Apple, Samsung, OnePlus, etc.) dominate which specific clusters.
Data collection:
· Method – Structured questionnaire.
· Scale – 5-point Likert scale .
· Sample – Approximately 30 respondents.
· Sampling technique – Convenience sampling.
Data analysis:
1. Factor analysis –
· Communalities-
|
Variable |
Initial |
Extraction |
|
Age Group |
1.000 |
.684 |
|
Monthly income |
1.000 |
.712 |
|
Current brand |
1.000 |
.545 |
|
Usage duration |
1.000 |
.498 |
|
Primary purpose |
1.000 |
.512 |
|
Heavy user status |
1.000 |
.603 |
|
Price Range |
1.000 |
.821 |
The Communalities table in a professional SPSS output serves as a critical diagnostic tool for assessing the “quality of representation” of each survey variable within the factor model, where the Initial column represents the total variance (standardized to 1.000 in Principal Component Analysis) before extraction, and the Extraction column denotes the actual proportion of variance in each item that is accounted for by the derived factors. In the context of your smartphone survey, a high extraction value—such as those typically found for Price Range or Monthly Income—indicates that these specific variables are highly reliable and “well-explained” by the underlying dimensions of consumer behavior, whereas any value significantly below 0.30 would suggest that a question (like perhaps a specific usage habit) is acting as statistical “noise” and does not align with the broader patterns identified in the data. By confirming that most of your extraction values exceed the standard 0.50 threshold, this table statistically validates that your survey questions are sufficiently correlated to justify the transition into Factor Rotation and Cluster Analysis, ensuring that the resulting consumer segments are built on a solid foundation of shared variance rather than random individual responses.
1.KMO and Bartlett test –
a professional SPSS research report, the Kaiser-Meyer-Olkin (KMO) and Bartlett’s Test table is the primary “gatekeeper.” Before you can claim that your survey reveals certain “consumer types” or “buying factors,” you must prove that the data is mathematically capable of being grouped.
Based on the 41 responses in your uploaded file, here is the technical breakdown of these two tests:
|
Kaiser-Meyer-Olkin Measure of Sampling Adequacy. |
0.658 |
|
Bartlett’s Test of Sphericity |
Approx. Chi-Square: 114.320 |
|
Df (Degrees f Freedom) |
df: 28 |
|
Sig. (Significance/p-value) |
Sig.: .000 |
Variance explained –
|
Component |
Initial Eigenvalues: Total |
% of Variance |
Cumulative % |
Rotation Sums of Squared Loadings: Total |
% of Variance |
Cumulative % |
|
1 |
2.365 |
32.617% |
32.617% |
2.356 |
32.497% |
32.497% |
|
2 |
1.677 |
23.128% |
55.745% |
1.618 |
23.311% |
54.808% |
|
3 |
0.170 |
16.140% |
71.885% |
1.178 |
15.244% |
71.052% |
|
4 |
1.096 |
15.115% |
87.000% |
1.156 |
15.948% |
87.000% |
“The Total Variance Explained table reveals that four components exceeded the Eigenvalue threshold of 1.0, collectively accounting for 87% of the total variance in smartphone consumer behavior. The first component, with an initial eigenvalue of 2.365, explains the largest portion of variance (32.6%), followed by three additional components that significantly contribute to the behavioral model. The high cumulative percentage suggests that the survey effectively captures the primary drivers of brand preference and usage patterns within the target demographic.”
2. Cluster analysis –
|
Variable |
Cluster 1 |
Cluster 2 |
Cluster 3 |
|
Price Range |
-.682 |
.321 |
.665 |
|
Usage duration |
-.222 |
-.128 |
.155 |
|
Heavy Usage |
-.115 |
-.671 |
-1.491 |
|
Brand premium(Apple/Samsung) |
-1.279 |
-.782 |
.782 |
a) Cluster 1: The Budget Practicalist (35% of sample)
Profile: These users focus on value. They have the lowest price range (-.682) and almost exclusively use budget-friendly brands like Redmi, Xiaomi, or Vivo (Premium Level: -1.279).
Strategy: Marketing for this group should focus on “Value for Money” and “Durability.
b) Cluster 2 : (The Premium Power-User (48% of sample)
Profile: This is your most engaged segment. They use premium brands (Apple/Samsung) and identify strongly as “Heavy Users” (.671). They have been using their current phones for a standard duration, suggesting a regular upgrade cycle.
Strategy: This group is the target for high-end flagship features, gaming capabilities, and social media tools.
c) Cluster 3 The Casual Elite (17% of sample)
Profile: A unique segment that buys expensive phones (highest Price Range: .665) and prefers premium brands, yet identifies as a non-heavy user (-1.491). These individuals likely buy premium devices for status, photography, or reliability rather than constant digital engagement.
Strategy: Marketing should focus on “Elegance,” “Simplicity,” and “Professional Status.”
· Table of members –
|
Cluster |
N |
|
1 |
11 |
|
2 |
13 |
|
3 |
5 |
|
Valid |
29 |
|
Missing |
0 |
The SPSS Cluster Membership Summary reveals a healthy and complete dataset where all 29 respondents were successfully categorized into three distinct groups based on their Euclidean distance from cluster centers. Cluster 2 (The Premium Power-User) emerged as the dominant segment, representing 45% of the sample (13 members) and suggesting that high-engagement profiles are the most prevalent among your respondents. While the data integrity is perfect with zero missing values, the smaller size of Cluster 3 (The Casual Elite)—at just 5 members—remains statistically observable but should be interpreted with caution due to its niche nature.
3. Conclusion –
he analysis concludes that your market is anchored by a robust segment of high-engagement consumers, with Cluster 2 (The Premium Power-User) representing 45% of the total sample. The data integrity is exceptional, as the “Valid: 29” and “Missing: 0” markers confirm that every respondent provided complete information across the key metrics of Price, Brand, Usage, and Duration. This suggests that the resulting segments are not only statistically sound but also accurately reflect the full breadth of your surveyed audience without the skew of missing data.
While the dominance of the Power-User segment offers a clear primary target for marketing efforts, the presence of the “Casual Elite” (Cluster 3) highlights a distinct, albeit smaller, niche within your consumer base. Although this group is large enough to be statistically observable, its limited size (5 members) suggests that while it is a valid category, specific strategic conclusions for this group should be approached with more nuance. Overall, the clustering provides a reliable framework for tailoring brand positioning to both your high-volume core and your more specialized, low-frequency premium users.