ELECTRIC VEHICLE
MARKET SEGMENTATION
PROJECT REPORT
P.PRAGATHI SREE
7-6-2023
PROBLEM STATEMENT:
The project is to conduct an in-depth analysis of the Electric Vehicle
Market in India using market segmentation to identify particular
categories within the market and comeup with a feasible strategy to enter
the market that targets the segments most likely to use electric vehicles.
ABSTRACT:
Electric vehicles are the critical technology for decarbonizing road
transport, which accounts for 16% of world emissions. Electric car sales
have expanded exponentially in recent years, owing to enhanced range,
wider model availability, and improved performance. Passenger electric
cars are gaining popularity; we anticipate that 13% of new cars sold in
2022 will be electric; if the growth shown in the last two years continues,
CO2 emissions from cars can be reduced to meet the Net Zero Emissions
by 2050 Scenario. Electric vehicles, however, are not yet a worldwide
phenomenon. Due to increased purchasing costs and a lack of charging
infrastructure availability, sales in underdeveloped and emerging
countries have been slow.
As seen by the nearly 0.32 million vehicles sold in 2021—an increase of
168% YoY—the Indian EV market is also developing quickly. The Paris
Agreement, which aims to reduce carbon emissions, improve the quality
of the air in urban areas, and decrease oil imports, is the foundation for
India's ongoing adoption of electric vehicles.
The market segmentation analysis of the electric vehicle market in India
is the main objective of this report. The market will be divided into
segments in the research depending on, among other things, consumer
demographics, vehicle types, and pricing points. This research will offer a
clearer view of the Indian electric vehicle market by identifying particular
market segments and assist in creating a successful market entry
strategy. Additionally, the study will discuss the market's key competitors,
present market conditions, and future market estimates.
EV MARKET IN INDIA:
The Indian automobile sector, which ranks fifth globally, is anticipated to
grow to advance to third-largest status by 2030. According to the India
Energy Storage Alliance (IESA), the Indian EV market would grow at a
CAGR of 36% as population rise. Since India imports about 80% of its
crude oil needs, dependency on conventional energy resources is not a
viable choice as vehicle demand develops. By 2030, NITI Aayog wants to
see EV sales penetration for all commercial vehicles reach 70%, for
private vehicles reach 30%, for buses reach 40%, and for two- and three-
wheelers reach 80%.
According to the Ministry of Heavy Industries, 0.52 million EVs have been
registered in India over the last three years. EVs experienced strong
growth in 2021, which was aided by the government's adoption of
beneficial laws and initiatives. With 66,704 units sold across all
categories in 2021, Uttar Pradesh led all Indian states in EV sales,
followed by Karnataka with 33,302 units and Tamil Nadu with 30,036
units. While Karnataka and Maharashtra lead the two-wheeler and four-
wheeler segments, respectively, Uttar Pradesh dominated the three-
wheeler market.
REDUCTION OF OIL CONSUMPTION BY VEHICLES
USING EV:
In 2021, the global EV fleet consumed around 50 TWh of electricity,
accounting for less than 0.5% of total global final electricity consumption.
In 2021, the use of EVs displaced approximately 0.3 Mb/d of oil. To meet
the Net Zero Scenario, EVs would need to displace more than 7 Mb/d of
oil by 2030.
Energy density is critical for ensuring that BEVs have enough range. The
energy density of EV batteries has increased over the last year, and some
of the finest performing battery cells can now reach energy densities of
over 300 Wh/kg, up from roughly 100-150 Wh/kg a decade ago -
meaning that electric cars can now travel twice as far with the same
mass. This advancement has been made possible by continuing
advancements in battery chemistry and cell design. Tesla's forthcoming
4680 cells and LG Energy Solution's Ultium cells are two prime instances
of this.
But it's not just about energy density. Reduced demand for essential
metals is another aim for EV innovation. The market share of lithium iron
phosphate (LFP) cathodes, which do not require nickel or cobalt, has
more than doubled in the last year. This was due in part to improved cell-
to-pack technologies that enable increased pack density by reducing pack
dead weight, but it was mostly due to manufacturers shifting to LFP to
reduce commodity price vulnerability. Another significant breakthrough
has been the announcement of significant supply chain development for
sodium-ion batteries by the world's largest battery manufacturer. Critical
metals could potentially never again be used because to this technique.
The IEA now rates sodium-ion technology at TRL 6, up from TRL 3 or 4.
Automakers are looking towards electrification strategies more and more
as a way to adhere to rules and stay competitive in a market that is
changing. A totally electric future was proclaimed as a goal by a number
of significant automakers in 2021.
MARKET SEGMENTATION FOR EV MARKET:
The main objective of this report is the market segmentation analysis of
the electric vehicle market in India. The market will be divided into
segments in the research depending on, among other things, consumer
demographics, vehicle types, and pricing points. This research will offer a
clearer view of the Indian electric vehicle market by identifying particular
market segments and assist in creating a successful market entry
strategy. Additionally, the study will discuss the market's key competitors,
present market conditions, and future market estimates.
The following is how the report is organised: Firstly we will present an
outline of the current situation of the Indian electric vehicle market. This
will include information such as market size, growth rate, and important
players. Following that, we will do a market segmentation analysis of the
Indian electric vehilces market using 2 datasets, defining particular
market groupings and we willl discuss some questions. Finally, based on
the findings of the market segmentation analysis, we will make advice for
how to design a successful market entry plan.
FOR DATASET-1 -> ElectricCarData_Norm.csv dataset.
IMPLEMENTATION IMPORTED LIBRARIES:
Packages/Libraries used:
• Pandas
• Numpy
• Matplotlib
• Seaborn
• Sklearn
• Plotly
DATA PRE-PROCESSING:
Data pre-processing is a stage that assists us in identifying variable
measurement levels, and the relationship between variables produces
hidden insights from raw data and informs us on the appropriate
algorithms to employ for market segmentation.
To gain insights from the raw data, we utilised data visualisation to
construct numerous plots:-
➢ We employed histograms and kernel density plots for numerical
variables.
➢ Bar plots, count plots, and pie charts for categorical variables.
➢ We used pairs plot, scatter plot, and heatmap to understand the
relationship between two numerical variables.
SEGMENT EXTRACTION (ML TECHNIQUE USED):
K-Means Clustering:-
Clustering is the task of dividing the population or data points into a
number of groups such that data points in the same groups are more
similar to other data points in the same group and dissimilar to the data
points in other groups. It is basically a collection of objects on the basis of
similarity and dissimilarity between them.
K-Means Algorithm is the simplest unsupervised learning algorithm that
solves clustering problem. K-means algorithm partitions n observations
into k clusters where each observation belongs to the cluster with the
nearest mean serving as a prototype of the cluster.
We will be using this tehcnique as here in this dataset we can observe the
clustering form of the datapoints. So we are using this technique.
IMPORTED LIBRARIES DATA PROCESSING:
DATA PROCESSING:
We can see that there are no null values in the dataset. So our data is
clean and we can go for further process.
USING CATEGORICAL VALUES:
So now we have our data in categorical form so now we can use pairplot
and heatmap for every column comparision.
PAIR PLOTS BASED ON THE RAPID CHARGING:
HEAT MAP FOR SHOWING THE CORRELATION IN
DATA:
Demographic analysis entails categorising the market based on factors
such as age, gender, income, education, and occupation. The research can
provide insights into the tastes and purchasing power of distinct groups
of customers by identifying specific sectors of the market based on these
demographic parameters, which can help to inform a successful market
entry strategy.
The analysis may, for example, reveal which age groups are more likely to
acquire electric vehicles, or which income groups are more likely to
afford higher-priced electric automobiles. Furthermore, it may
demonstrate differences in purchasing behaviour across different
geographies, urban and rural locations, and other demographic variables.
COMPARISION PLOTS:
TOP SPEED ACHIEVED BY A BRAND
NUMBER OF SEATS IN A CAR
INDIAN AND EURO PRICE
FREQUENCY OF BRANDS
NUMBER OF SEATS IN A CAR:
ELBOW METHOD NUMBER OF SEATS IN A CAR:
PREDICTION OF PRICES MOST CARS USED:
A machine learning algorithm built on supervised learning is linear
regression. The task is a regression one. Based on independent variables,
regression models aim to predict a value. It is mostly used to determine
how variables and forecasting relate to one another. Here, we employ a
linear regression model to forecast the costs of various Electric vehicles
in various businesses. The dependent Prices are represented by y,
whereas the independent variables are contained in X. We split the data
into a 4:6 ratio to train our model, which means that 40% of the data is
utilised.
LinearRegression().The data collection is fitted into the model using the
fit(Xtrain,ytrain) command. The figure gives information about the
intercept, coefficient, and cumulative distribution function (CDF) values.
We test the remaining 60% of the data on the model when the model-
training process is complete. The resulting results are validated using a
scatter plot for the dependent variable, acquired as a straight line as
shown in the figure, between predicted values and the original test data
set, and the density function is likewise normally distributed.
FOR DATASET-2 -> EV Stats-1.csv dataset.
For this Dataset we used Linear Regression Machine Learning Algorithm.
This gave the good value of accuarcy. Almost 99.99% of the dependent
variables has been explained in independent variables. So we used this
algorithm here.
LINEAR REGRESSION:
Linear Regression is one of the most basic Machine Learning algorithms
and is used to predict real values. It involves using one or
more independent variables to predict a dependent variable. Although it
is one of the simplest algorithms we will encounter it is extremely
powerful and robust in nature, making it an essential tool for aspiring
data professionals. As our dta is linearly varying so we can just consider
this.
IMPORTED LIBRARIES DATA PROCESSING:
DATA PROCESSING:
As we have null values we filtered and cleaned the dataset by dropping
the null values.
PAIR PLOTS BASED ON THE RAPID CHARGING:
HEAT MAP FOR SHOWING THE CORRELATION IN
DATA:
COMPARISION PLOTS :
Total vehicles in different states
Comparision plot for all columns
SCATTER PLOT:
As we can see that linear variation of the datapoints so we used Linear
Regresssion.
HISTOGRAM:
MODEL USED IN ELECTRIC VEHICLE
SEGMENTATAION:
In the electric vehicle market segmentation project, the K-means
clustering ML model played a crucial role. K-means clustering is an
unsupervised learning algorithm that helped us group and categorize the
electric vehicles based on their similarities in key features such as range,
top speed, efficiency and price. By applying the K-means algorithm, we
were able to identify distinct clusters or segments within the electric
vehicle market, each representing a specific group of vehicles with similar
characteristics. This segmentation enabled us to understand the market
better, target specific customer groups, and tailor our marketing
strategies accordingly. Additionally, K-means clustering helped us
uncover patterns and trends within the electric vehicle market, leading to
valuable insights and informed decision-making for our project. For
analysing different Electric Vehicles in different states, Linear Regression
model was useful.
HOW WOULD I IMPROVE THE PROJECT IF EXTRA
TIME PROVIDED:
I would concentrate on expanding the dataset collection and looking into
other machine learning models if given more time and money to better
the Market Segmentation Project. I would approach it as follows:
➢ Dataset Gathering;
• Consumer Demographics: Gathering data on consumer
demographics, such as age, gender, income level, and
location, would enable more thorough customer profiling
and targeted advertising campaigns.
• Competitor Analysis: Information on rival electric car
models, their specifications, costs, and market success will
help identify market positioning and potential blind spots.
➢ Further ML Models:
• Gaussian Mixture Models (GMM) is a probabilistic clustering
approach that can recognise overlapping clusters and
capture complex cluster structures. It would provide more
refined segmentation in cases where clusters are not clearly
separated.
• Decision Trees and Random Forests: These tools can be used
to pinpoint important characteristics and their role in
market segmentation. These models can provide
interpretability and insights into what influences customers
the most.
• Neural Networks: To find non-linear relationships and
identify complicated patterns in the data, neural networks
like multi-layer perceptrons or convolutional neural
networks can be used. They might be helpful in uncovering
hidden trends and drawing deeper conclusions from the
dataset.
➢ Advanced Feature Engineering:
• Sentiment Analysis: Using techniques for sentiment analysis
on customer reviews or social media data might assist
determine how the general public feels about various electric
vehicle models. Market segmentation might be improved
even more using this information.
➢ Outside Information Sources:
• Environmental Factors: Considering environmental
information, such as data on carbon emissions or air quality
indices, might help shed light on the connection between
eco-awareness and the uptake of electric vehicles.
The Market Segmentation Project would be enhanced with more in-depth
insights and increased accuracy in identifying and targeting various client
categories by extending the dataset collection, utilising more ML models,
and adding sophisticated feature engineering approaches.
TOP FOUR VARIABLES FOR CREATING MOST
OPTIMAL MARKET SEGMENTS:
Here are four variables/properties that are typically relevant and should
be taken into account for constructing ideal market segments:
➢ Range: For prospective purchasers, the range, or the distance an
electric car can go on a single charge, is essential information. The
driving requirements and preferences of various client segments
may differ with regard to the driving range of their electric
vehicles. Longer ranges may be more important to some
consumers for long-distance travel, but others may just need their
vehicles for short commutes and favour those with a smaller range.
➢ Price: In any market, buyers give price a lot of thought. Based on
elements including battery capacity, features, brand, and market
positioning, electric vehicles frequently come in a variety of price
ranges. Different client categories' purchase decisions may be
influenced by price sensitivity and financial restrictions. While
other segments may prioritise affordability, segments with greater
purchasing power may be willing to spend more on premium
electric vehicles with cutting-edge technologies.
➢ Availability of Charging Infrastructure: Access to and
availability of charging infrastructure are crucial elements in the
adoption of electric vehicles. Based on the availability or lack of
dependable charging infrastructure in particular geographic areas
or target markets, segments can be created. Shorter commutes may
be preferable in some segments due to inadequate charging
infrastructure, while longer commutes may be possible in others
due to improved access to charge stations that allow longer
excursions and reduce range anxiety.
➢ Vehicle Size and Type: There are many different sizes and types
of electric vehicles, including sedans, SUVs, hatchbacks, and even
electric motorbikes. Based on elements including lifestyle, family
size, and usage requirements, customer segments' choices for
vehicle size and type can differ. Customers' preferences for
compact, mid-size, or bigger vehicles, as well as their particular
requirements for passenger capacity, cargo space, or off-road
capabilities, could be used to form segments.
The best features for market segmentation should be chosen
depending on the unique objectives, target market, and data available
in your electric car market domain.
FINAL CONCLUSION:
Therefore, based on the data from 2 datasets, it is clear that the best
target market should fall into one of the following groups:
Behavioural:
➢ Most of the automobiles in our analysis had Two seats.
Demographic:
➢ Top Speed & Range: Due to the size of the market, cost is influenced
by the top speeds and maximum range of the cars.
➢ Efficiency: The segments tend to be the most efficient.
➢ Total in state: The total number of Electric vehicles in Maharastra
are more compared to other states.
So our over all conclusion of our target market should include vehicles
with the highest levels of efficiency and top speed often have two seats.
Most of them are in Maharastra state following with Gujarat, Uttar
Pradesh and Haryana.
GITHUB LINK:
https://github.com/Pragathisree13/EVs_Market_Segmentation_Pragathi
THANK YOU
