A Project Report On Using

Personalised E-Marketing
Strategy For Smart E-marketing
With Reference To
Litz-Technology

Project Work Done By:

S.Vishnu Kumar
Under The Guideness Of:
Dr.Suganthalakshmi

INTRODUCTION

In 21st century if we are not on a social networking site and

dont have an e-mail, we are not on the internet.
Its as true for advertisers as it is for consumers.
Browsing history (user generated content) and social
networking is the ultimate manifestation of user generated
content, and as such, holds more potential for growth than
any other form of content on the web today.
Todays marketing model is collaborative, collective,
customized and shared.
Its a world in which the consumer is the creator, distributor
and consumer of content.

INTRODUCTION-conti

Personalized marketing is a way of delivering

tailored messages to an individuals through
personalized web sites (Dynamic Sites).
Personalized feeds and collaborative filtering have
become extremely popular in these websites.
These sites can change based on the users buying
and surfing habits i.e. automatically modifying the
textual and multimedia features based on past
usage of the site, demographics, and a multitude of
other attributes which could be collected from the
users online by tracking their activities.

PERSONALIZATION

The fundamental concept of personalization as a part of the E-Marketing mix lies in the
need of recognizing, identifying a certain customer in order to establish relations (i.e.
Establishing relations is a fundamental objective of Marketing).
It is crucial to be able to identify our customers on individual level and gather all possible
information about them, with the purpose of knowing our market and be able to develop
customized, personalized products and services.
For example, an internet cookie strategically placed on the website visitor's computer
can let us know vital information concerning the access speed available: in consequence,
if we know the visitor is using a slow connection (eg. dial-up) we will offer a low volume
variation of our website, with reduced graphic content and no multimedia or flash
applications.
This will ease our customer's experience on our website and he will be prevented from
leaving the website on the reason that it takes too long to load its pages. Personalization
can be applied to any component of the Marketing mix; therefore, it is a moderating
function.

NEED FOR STUDY

We want better, more relevant results from our Internet searches, but we dont want to be tracked to
get them. So concludes the survey report conducted by LITZ Technologies.
Its yet another bucket of cold water splashed on the future of online advertising and the publishers it
supports.
According to the digitally administered questionnaire survey of about 106 adults from Coimbatore
region which mostly consisted of college goers and office goers early this year, 65% of people who
were asked how theyd feel if a search engine kept track of their searches and then used the
information to personalized future results said its a bad thing because it limits information they get.
Some 73% said its bad because its an invasion of their privacy.

And 68% said theyre not OK with targeted online ads because they dont like their behaviour tracked
and analysed, a finding consistent with earlier studies by others.

Theres a paradox here, of course. People say they want relevant results, and in fact they say search
engine results are getting more relevant. But better relevance is happening largely because of
personalized results based on data collected about peoples activities online. People have legitimate
concerns about their privacy violations

OBJECTIVES OF THE STUDY

The main objective of this study is to find answers for the following
questions

Do really personalisation marketing works.

What is the value of Personalised E-Marketing ROI compared to others

Do users welcome and are open to getting tracked.

How likely are users to click the pay per advertisement.

The use of AD-ROSA system for personalised E-Marketing.

FINDINGS

As a result of constant and overwhelming exposure to the modern consumer, traditional

advertising has endured intense scrutiny due to its intrusive nature.
With the advent of digital media on the mobile phone and the internet and the use of modern
technology in television, marketers are attempting to minimize traditional advertising as a form of
promotion.
Many existing approaches to advertisement recommendation are based on demographic targeting
or on information gained directly from the user. This one is not much welcomed by the user, as
they feel their privacy is breached.
Personalized rendering of web pages gives the users greater control to view only what they prefer.
While personalized feeds and collaborative filtering have become extremely popular on media
websites, content personalization has received much less attention.
Automatically modifying the textual and multimedia features of an article offers significant
opportunities.

SCOPE AND FOCUS OF THE

STUDY

The explosive growth of the E-Marketing over the last two decades has
captured the attention of many researchers and academics from various
scientific fields.

Recognizing the diversity of the nature of an E-Marketing, this project work

tries to examine the topic of a personalized E-Marketing.

Although the main goal of the study is about analysing the impact of
personalised E-Marketing so far has created in consumer perspective and
what are the alternate solutions for the same. The work tries to analyse
the methods and the strategies that can lead to a successful E-marketing.

DIFFERENT PERSONALIZED EMARKETING MEDIUMS

Personalised E-mail Marketing

Personalised Dynamically Changing Websites

Personalised Search Engine Optimization

Personalised Search Engine Marketing

Personalised Online Advertising

Personalised Pay per Click Advertising

Personalised Social Media Advertising

Personalised Mobile Marketing

PROBLEM WITH CURRENT

PERSONALISED E-MARKETING
METHOD

Using part of the information that have be gathered by internet marketers

to achieve personalization is closely related to that of behavioural
targeting. This technique examines the consumers behaviour in terms of
websites visited and searches made in order to optimize the
advertisement that they receive.

An important factor to take into consideration is that this type of

marketing receives a great level of interest worldwide and data shows that
more and more money is being spent for its use.

This then provides a base for personalized marketing which makes use of
this technique and attempts to improve its efficiency by gathering a larger
data set.

The problem lies repeatedly in the fact that behavioural targeting is

already taking heat in terms of its privacy issues, which is a major issue
for personalization.

Server-side aggregation is being increasingly challenged by consumer and

PRIVACY-POLICY

Privacy is an element of the mix very much connected to the previous one
-personalization.
When we gather and store information about our customers and potential customers
(therefore, when we perform the personalization part of the E-Marketing mix) a
crucial issue arises: that of the way this information will be used and by whom.
A major task to do when implementing an E-Marketing strategy is that of creating
and developing a policy upon access procedures to the collected information.
This is a duty and a must for any conscious marketer to consider all aspects of
privacy, as long as data are collected and stored, data about individual persons.
Privacy is even more important when establishing the E-Marketing mix since there
are many regulations and legal aspects to be considered regarding collection and
usage of such information.

SECURITY

The "security" function emerged as an essential function of e-Marketing once

transactions began to be performed through internet channels.
What we need to keep in mind as marketers are the following two issues on
security:
security during transactions performed on our website, where we have to take
all possible precautions that third parties will not be able to access any part of a
developing transaction;
Security of data collected and stored, about our customers and visitors.

A honest marketer will have to consider these possible causes of further

trouble and has to co-operate with the company's IT department in order to
be able to formulate convincing (and true, honest!) messages towards the
customers that their personal details are protected from unauthorized eyes.

PROPOSED METHOD

Personalization of online advertising is a great challenge while the market

is moving and adapting to the realities of the Internet.
Many existing approaches to advertisement recommendation are based
on demographic targeting or on information gained directly from the user
the AD ROSA system for automatic web banner personalization, which
integrates web usage and content mining techniques to reduce user input
and to respect the users privacy.
In this paper we introduce the AD-ROSA system for automatic web sites
personalization, which integrates web usage and content mining
techniques to reduce user input and to respect the users privacy.
To enable online personalized advertising -the integration of all relevant
information is performed in one vector space.

TECHNICAL SUGGESTIONS FOR

THE PRIVACY AND
SECURITY ISSUES

The method proposed below solves the problem of

automatic personalization of advertisements with
respect to user privacy.
That is none of the users personal details are stored in a
database.
It is based on knowledge extraction from the web pages
content and historical user sessions, as well as the
current behaviour of the online user, and the use of data
mining techniques.
The implementation of data mining to web content and
web usage is usually called web content and web usage
mining, respectively.

THE AD-ROSA SYSTEM

Continued

The proposed method uses web mining technique to combine into one personalized
framework with several useful factors of advertising like the most suitable content, click
probability etc...
Historical user sessions are stored in the database and clustered to obtain typical,
aggregated user sessions.

The clusters centroid corresponds to one usage pattern of the publishers web site.
Each user session is linked up to the set of advertisements visited (clicked) by the user
during this session (visited ad vector).

Having a cluster of sessions, the AD ROSA system can also extract information about
related, visited advertisements, by counting the mean vector (ad visiting pattern vector)
from all visited vectors related to sessions from the cluster.
Thus, one web usage pattern (centroid) corresponds to exactly one ad visiting pattern
(centroid).

Continued

The site content of the publishers web pages is automatically processed in the
similar way. Content thematic groups - conceptual spaces - are received using the
clustering of term vectors extracted from the HTML content of web pages.
In order to recommend a suitable advertisement for the user we have to know its
general subject matter. This is achieved by text (HTML) content analysis of the
advertisement target web site.
The AD ROSA system automatically downloads advertisers web pages and processes
only the terms, which occur in the publishers web pages.
As a result we obtain advertising conceptual spaces corresponding to the appropriate
publishers conceptual spaces.
A user requesting a web page is assigned to both the closest usage pattern and the
closest conceptual space, based on the users previous behaviour during the session
(the active user page session vector).

Continued

Assuming that he or she will behave like others(usage pattern) and is

interested in web pages similar to those recently visited (conceptual
space), the AD ROSA system can recommend to the user the most suitable
advertisements.

Advertisements related to the closest conceptual space link to web sites

with content appropriate for the current user, so they should be displayed
to the user. The closest usage pattern and in consequence the closest ad
visiting pattern enables the selection of advertisements that are most
likely to be clicked by the user.

User behaviour (active user session vector) as well as information about

already displayed or visited banners (active user ad session and active
user ad visited vector) are stored by the system separately for each user.

Continued

They all prevent the too frequent emission of the same advertisement for
one user and provide control over the number of emissions contracted in
the advertising campaign.

A new, personalized advertisement ranking is determined for the user

after each users request. This process exploits information mentioned
above and the targeting parameters established by the advertiser
(advertising policy), like limited emissions per user during a single session
(active user ad emission acceptance vector).

As a further advantage for the publisher, some additional priority features

of each advertisement can be set up manually (advertising policy). This
provides an opportunity to increase the rankings value for more profitable
advertisements.

Finally, the personalized ranking list is filtered using additional advertising

policy features: limitation to certain web browsers, time of day of the
emission, etc.

CONCLUSION

To conclude, we have seen that e-Marketing implies new dimensions to be

considered aside of those inherited from the traditional marketing. First of
all, personalized marketing seems to have quite a good foundation upon
which to build its success.

A large part of the consumers believe that this form of promotion

constitutes advancement and provides better information than the current
platform. However, marketers need to be careful not to antagonize the
consumer by bombarding them with advertisement, so that they will be
able to maintain this positive view and gain permission to collect a users
personal information.

Furthermore, there is certain information that appeals to some consumers

more than others, and marketers should focus their advertisement toward
their inclusion.

Marketers need to find ways in which to create the best advertisement

with this information, or provide better privacy and security measures to
convince the consumer to disclose this data.

From a consumers point of view, the internet seems the appropriate and
most wanted medium from which to receive personalized advertisement,
even though there is a considerable amount of individuals who would
prefer other mediums.

Upon reflection on all aspects, personalization appears most inevitable for

the future of advertising. Regardless of the challenges it faces, the need
for appropriate targeting and increased relevance will ensure its success.

LIMITATIONS OF THE STUDY

To start with, the length of the survey might have constituted a drawback as it
leads to lack of concentration or willingness to carefully answer each
question.

On the other hand, its already extensive size did not permit the inclusion of
additional questions.

The number of respondents is limited to 106 which is bias based on the

sample size.

The research is limited to 3 months so a deep analysis about the concept

cannot be made.

Moreover, the fact that some were not aware of personalization provides the
ability to consider their views as compared to those who have experienced
this form of promotion.