Sie sind auf Seite 1von 15

Mobile VAS Study

Paper
Submitted by :

Rahul Nagpal

Table of Contents
Research Background......................................................................................................3
Research Objectives and Scope.......................................................................................3
VAS Overview.................................................................................................................3
i) User Segmentation in the Mobile VAS.................................................................4
ii) Mobile VAS Value Chain.................................................................................5
Impact of Long Tail on Mobile VAS..............................................................................6
Impact of Crowd-Sourcing on Mobile VAS....................................................................8
Cross-Selling Models for Mobile VAS.........................................................................10
i) Statistical Approach............................................................................................10
Up-selling models for Mobile VAS...............................................................................12
i) Proposed up-selling model for CRBT’s..............................................................13
ii) Up-selling Model.............................................................................................14
Conclusion.....................................................................................................................15
Conclusion

Research Background
The Indian mobile telephony market has grown at a rapid pace. Declining call
tariffs in conjunction with favorable regulatory policies have lead to a
tremendous increase in the subscriber base, crossing the 100 million mark in
2006. While the growing subscriber base has positively impacted industry
revenues (which have risen consistently over the past few years), operator
margins also have shrunk, pulling down “Average Revenue per User” (ARPU).
As ARPU declines and voice gets commoditized, the challenge is to retain
customers, develop alternative revenue streams, and create a basis for
differentiation in high-churn markets.

Telecom operators are now looking at “Mobile Value Added Services” (MVAS)
for growth and a large chunk of revenue is expected to flow from VAS in the
near future.

Research Objectives and Scope


1) Mobile VAS Trends
2) Impact of Long-Tail on Mobile VAS
3) How to leverage crowd-sourcing in Mobile VAS
4) Cross-selling and up-selling models in Mobile VAS
5) Statistical Model and Analysis

VAS Overview
From the early days of “Person-to-Person Short Message Service” (P2P SMS),
the industry has witnessed an emergence of a growing portfolio of services
including graphics/wallpapers downloads, ringtones and caller ring back
tones (CRBT), SMS contests, and games It is estimated that the Indian VAS
market, including messaging, mobile internet, social networking and mobile
payments, stood at Rs 16,650 crore in 2009. It is expected to cross Rs 21,940
crore in 2010.compound annual growth rate (CAGR) of more than 50%.

M-VAS can be broadly classified into the following 4 categories

COMMUNICATION ENTERTAINMENT

SMS Ringtones
MMS CRBT
Mobile Email Wallpapers/Screensavers
Chat Music
IVRS Games
USSD Video
Jokes

INFORMATION M-COMMERCE

Personalized Info Message m-banking


(Horoscope, Weather, Stock, News, payments
Movie Reviews, Cricket updates) movie ticket reservation
Browsing m-shopping
Location Based Services
(GPS, Restaurants etc.)
Missed Call Alert

In terms of current segmentation, mobile entertainment themed VAS,


including mobile music, games, images, TV and video, accounts for 29 per
cent. Mobile messaging and mobile Internet accounts for 49 per cent and 21
per cent respectively and 1 per cent revenue is generated from other value
added services.

Entertainment VAS is expected to drive the growth of the market going


forward with Video/TV and games registering the highest growth.

i) User Segmentation in the Mobile VAS

Apart from the general demographic segmentation of the customer like age,
sex, location, education, profession, income etc, we can also segment the
customer based on the MVAS service usage like Type of VAS service, tenure
of usage, download medium (IVRS, SMS, USSD, Internet), number of times
the user has opted out of the service in the past 1 year, number of times the
user has made use of a promotional offer.

Faced with diminishing marginal returns on voice revenues, a lot of telecom


companies have abandoned their one size fits all demographic segmentation.

Airtel came up a strategy where it reclassified users into different segments


based on attitudes, VAS adoption traits and average revenue per person
(ARPU). Airtel focused its attention on two segments in particular. The first
group was known as "funsters" -- consumers aged between 18-35 years old,
who share a common trait -- a high adoption of VAS. Airtel believes that with
some targeted marketing, spending from this group can increase quite
significantly. The telco has exclusive tie-ups with application providers like
Google and has identified music-on-demand as a key value proposition.

The other segment of focus is the "achievers", who are the top five percent of
mobile users. While this segment isn't a high adopter of VAS, it contributes to
revenues nearly 10 times that of Airtel's ARPU, currently at US$10.1

For selling data intensive services new marketing strategies have to be


adopted. There has to be shift from ‘Traditional segmentation’ based on
Demography, Income, Location to ‘Evolved segmentation’ based on Degree-
of-Involvement of consumers.

ii) Mobile VAS Value Chain

The value chain is in the evolving stage and the roles are not clearly defined.
There is lot of role overlapping. Technology enablers play the role of content
aggregator, operators venturing into the domain of technology enabler, etc.
The market will get consolidated in future with clear role definitions, however
it is likely to take around 3 years.

The current revenue distribution is in favor of mobile operators who capture


anywhere between 60-80%. Technology enablers get anywhere between 10-
20% and content aggregators get 10-15% of the revenue. The content owner
gets 5-10%.

Operators cite infrastructure development, cost of acquiring subscribers, and


marketing costs as the reasons for the higher revenue share on their side.
Thus, operators have the highest bargaining power. Since there is only
handful of operators with market share in double digits, it is not business
sense to ignore any of these. Once the number of operators increases, the
revenue share of operators is expected to dilute.

Impact of Long Tail on Mobile VAS.


Anecdotal evidence suggests that Internet markets has helped shift the
balance of power from a few best-selling products to niche products that are

1 Imedia Connections : Segmenting India’s mobile market


previously obscure. For example, Frank Urbanowski, Director of MIT Press,
observes that the increased accessibility to backlist titles through the
Internet has resulted in a 12% increase in sales of these titles (Professional
Publishing Report 1999). This increase has happened despite flat growth in
overall book sales. Similar observations have been made in electronic
markets for music, DVDs, and electronics. Rhapsody, an online music
provider, streams more songs each month beyond its top 10,000 than it does
its top 10,000. While “new release” movies account for a dominant share of
revenue in a video rental shop, DVDStation, a company that allows
consumers to search and reserve movies online and pick them up in a DVD
kiosk, finds that more than 50% of their rental revenue come from titles that
are not new releases (DVDStation 2005).

Consumers are no longer limited to “key” products that are heavily


promoted, advertised and, as a result, highly visible. The increased sales of
niche products can lead to changes in the firm’s profit function and changes
in the firm’s strategies in producing and marketing its products.

Proposition1. Equilibrium profits decrease when consumer search costs for


niche products are higher.

Proposition2. Product sales become more concentrated when consumer


search costs for niche products are higher

Proposition3. Consumers, whose search costs are lower, have a higher


tendency to purchase niche products.

For instance, a firm can respond to lower consumer search costs (a change in
the demand side) by increasing its product selection (a change in the supply
side)

As far as the demand side is concerned, in case of mobile VAS (CRBT) the
cost of searching is high. A customer has to either call through IVRS and
browse through the list of songs one by one or interact through P2A SMS in
order to download the songs via SMS. Also, the call rates for these IVRS calls
are very high (Rs 6-8 per minute). Hence, it is not technologically feasible for
the user to search the niche content.
The Internet channel provides online search and recommendation tools that
help customers locate niche products, while the catalog channel does not
have these tools. However, with more and more people using internet as a
mean to download VAS like music, CRBT, ringtones etc. the cost of searching
is likely to go down in the future. The long tail could then come into play.

As far as the supply side is concerned, the cost of holding the long tail
inventory comes into the picture. In the case of VAS with services like CRBT,
ringtones and songs the cost of holding additional inventory is minimal since
it is digital media. Hence, it is more logical to store the niche products which
would cater to the “long tail”.

If we look at the Apple App Store or Android market place, the continued
adoption of these devices is due to the ubiquity of the Apps that are built for
these devices. Very few of the applications are popular and frequently
downloaded. However, the end user gets attracted to a platform which has
more apps. There is a concept of long tail here.

Conclusion

With the increasing use of Smart Phones/WAP Phones and Mobile Internet,
the cost of searching has decreased dramatically over the years. Hence, it
makes sense to cater to the long tail in mobile VAS. The only thing to keep in
mind for the supplier is the cost of storing the additional inventory (VAS
service like CRBT).

Impact of Crowd-Sourcing on Mobile VAS.


With the increase of web applications and more and more users being
attracted to the web platform, several crowd-sourcing models have emerged.
An important example of web-based crowd-sourcing is social bookmarking.
In social bookmarking systems, users assign tags to resources shared with
other users, which has given rise to a type of information organisation that
emerges from this crowdsourcing process. Other useful crowd-sourcing type
is web based idea competitions like threadless T-Shirts.

The unique benefits of mobile technology in ubiquity, media convergence,


and global mass reach will continue to drive new, innovative crowd-based
service concepts over mobile

Few of the applications in the mobile VAS world would be:-

1. User experience optimization. A lot of companies like Google,


Amazon, Netflix etc. use wisdom of the crowds to enhance user
experience. For e.g., Netflix has a unique recommendation model
where it uses a complex matrix algorithm to suggest movies based on
your viewing preferences. It asks users to rate the movies that they
have watched and creates micro-segmentation of users with similar
profiles. It uses the concept of wisdom of the crowds (rating the
movies) to create powerful predictive algorithms. It not only helps
enhance the user experience but also helps Netflix tap into the long
tail of its movie catalogue by providing the users with lesser known
movies of his/her favorite genre.
If we try to translate this into mobile VAS, we have CRBT’s, Caller
Ringtones, Movies, Wallpapers, and Songs etc. which the users could
rate. Based on their ratings, one could create profiles of users based
on their past preferences and hence, enable more customized,
accurate and targeted predictions.
2. Geo-social Networking. It is a type of social networking in which
geographic services and capabilities such as geocoding and
geotagging are used to enable additional social dynamics. User-
submitted location data or geolocation techniques can allow social
networks to connect and coordinate users with local people or events
that match their interests. Geolocation on web-based social network
services can be IP-based. For mobile social networks, texted location
information or mobile phone tracking can enable location-based
services to enrich social networking. Google latitude (location-aware
mobile app) tries to leverage this aspect of social networking.
3. Product testing. Crowd sourcing is a good platform to get user
feedback for new applications. This is especially useful when you try to
launch new VAS products in the market and want to gauge user
response. The use of Companies such as uTest and Mob4Hire employ
crowds to test newly developed mobile applications and provide
feedback on functionality, load and performance, and usability.
4. User Generated Content. Although this feature of crowd-sourcing
has not gained ubiquity on mobile platform, a few companies like Qik
and Kyte enable users to generate and share a large variety of content
(mainly videos) with their network. The role is much similar to the role
of Youtube on PC Web.

Cross-Selling Models for Mobile VAS


Cross-selling is a strategy of selling new products to a customer who has
made other purchases earlier. Except for the obvious profit from extra
products sold, it also increases the dependence of the customer on the
vendor and therefore reduces churn. This is especially important in the area
of telecommunications, characterized by high volatility and low customer
loyalty.

i) Statistical Approach

As a part of the data mining models used for cross-selling of products, the
following models were studied
1. Regression (simple and step-wise)
2. Correlation
3. Market Basket Analysis (Product Association)
4. Neural Network
5. Bayesian Network

Although all the above methods have their own set of advantages and
drawbacks, the Market Basket Analysis has been the most preferred
method amongst the options because of its ease of implementation and
predictive power.

Market basket analysis (MBA) is a marketing technique that facilitates the


study of customer purchase trends together with the promotion and pricing
of certain products. It reveals how purchase patterns shift with time and
helps retailers understand product-timing dependencies, such as holiday or
seasonal purchase trends. The purpose of a market basket analysis is to find
dependencies and relationships between various retail products, on the basis
of which retailers arrange catalog or shelf items, design promotions and build
cross-marketing strategies.

This system has been successfully translated to online retail services as well.
Online retailing giant Amazon offers its recommendations using the market
basket analysis using the famous Association data mining algorithm.

The association task for data mining is the job of finding which attributes go
together. The task of association seeks to uncover rules for quantifying the
relationship between two or more attributes. Association rules are of the form
“If antecedent, then consequent,” together with a measure of the support
and confidence associated with the rule.

If we translate this association rule into mobile VAS market basket, we could
easily find whether there is an association between the purchases of various
VAS services. For example, a particular historical dataset reveals that out of a
1000 customers, 200 subscribed to the CRBT’s, and out of those 200
customers, 50 have also downloaded related wallpapers. Thus, the
association rule would have 2 components to it.
a) Confidence. The confidence of an association rule is a percentage
value that shows how frequently the rule head occurs among all the
groups containing the rule body. The confidence value indicates how
reliable this rule is. The higher the value, the more often this set of
items is associated together. In this case, the confidence {CRBT} =>
{Wallpaper} is 50/200 = 25%
A problem with confidence is that it is sensitive to the frequency of the
consequent (Y) in the database. Caused by the way confidence is
calculated, consequents with higher support will automatically produce
higher confidence values even if there exists no association between
the items

b) Support. The support is simply the number of transactions that


include all items in the antecedent and consequent parts of the rule.
(The support is sometimes expressed as a percentage of the total
number of records in the database). Here, the support for the item-set
{CRBT,Wallpaper} is 200/1000 = 20%.
The disadvantage of support is the rare item problem. Items that occur
very infrequently in the data set are pruned although they would still
produce interesting and potentially valuable rules.

In order to improve on the disadvantages of the confidence and support


principles of association, there are more components designed for
association rules:-

a) Leverage. Leverage measures the difference of X and Y appearing


together in the data set and what would be expected if X and Y where
statistically dependent. The rational in a sales setting is to find out how
many more units (items X and Y together) are sold than expected from
the independent sells. Leverage is a very good means to eliminate
noise due to low frequency events. For example, we could eliminate
the items which have minimum leverage threshold of 0.1%, or 10
occurrences in a dataset of 10,000 transactions.
leverage(X -> Y) = P(X and Y) - (P(X)P(Y))
b) Lift. Lift measures how many times more often X and Y occur together
than expected if they where statistically independent. Lift does not
suffer from the rare item problem like support does.
lift(X -> Y) = conf(X -> Y)/supp(Y)

The association rule can also be used to segment the different items into
logical group sets and hence, is effective to understand the ontological
relationship between various products and services in case of mobile VAS.
Once we group these together, we can use this relationship to launch single
level as well as multi-level campaigns and determine the lift from these
campaigns.

Another major advantage of association rule is that it is easy to understand


and implement. As explained above, in its very simplistic form, it only
consists of the 2 components i.e., confidence and support. At the same time,
various complexities can be implemented into the model based on the
requirements like leverage and lift to make it more robust and useful.

Up-selling models for Mobile VAS


Up-selling is a sales technique whereby a retailer induces the customer to
purchase more expensive items, upgrades, or other add-ons in an attempt to
make a more profitable sale.

Up-selling is generally harder than cross-selling since here, you are inducing
customer to buy a more expensive variation of the same product. Hence, you
would need to know the customer buying habit and price sensitivity. You
might also have to induce a fear of durability in case of the current line of
product that the consumer is buying and at the same time, have a strong
value proposition for the product that you are up-selling the consumer. An
industry where this works the best is software/computer industry where the
retailers induce customers to buy add-ons and upgrades frequently, either
through discounting (10 % off on external hard drive/mouse with a PC) or
version upgrades (how version 2.0 of a software is better than version 1.0).

In the below section, I will mention some of the successful techniques that
Amazon uses for up-selling of products and how they can be translated into
mobile VAS:-
1) Offer free shipping for items purchased more than $X value.
This is to induce the customer to purchase that extra bit more to avail
the benefit of free shipping. However, these amounts have to differ for
difference customers. Similarly, in mobile VAS, telecom providers could
offer customers with small discounts to urge them to purchase more
expensive VAS services. This will have to be customized for different
customers based on their purchasing habits and the average amount
they spend.
2) Offering Catchy discounts in the recommendation section.
Amazon uses intelligent phrases in the recommendation section to
entice customers to buy upgrades like
“You might also like”, “More items from this category”, “More items
you may need”. It might also entice you for buying a more expensive
version of its current product. Let’s say if you are looking at a 17”
computer monitor for $99. Amazon might come up with this
recommendation “Take a look at this 21” flat screen monitor for only
$139!”
In case of mobile VAS, telecom operators should push the up-sell of
products through various means like providing recommendations, IVRS
and SMS and make sure that they have a strong value proposition for
the product that they are up-selling.

i) Proposed up-selling model for CRBT’s

CRBT’s can be broadly classified into:-

RBT Type Explanation

Default RBT immediately receive a song

Caller/ Group based RBT Different callers are played different


ring back tones
Special Condition RBT The RBT is set depending on the day
(Valentine Day, Independence Day,
Holi, Rakhi, Diwali etc.)
Gift a Tone Ability to gift a tone to friends and
family

Copy a Tone Subscribers can adopt the RBT of


someone they call if they like the
tune
Fixed Album The subscriber’s RBTs comprise those
of a complete album and are played
randomly to callers
Juke Box Ability to select the playlist that
callers hear
Corporate RBT Employees set corporate jingles on
their RBTs
Roaming Tone RBT Set a roaming tone to be played even
when out of the network

ii) Up-selling Model.

Level 0
Default RBT
Immediately Receive a Song

Segment:-
Corporate Customers Middle Aged Group (25-35)
Students
Level 1

Copy/Record a Tune
orporate RBT Special Condition RBT Occasion Based RBT
Copy a tune from another caller or record a
t Corporate JinglesBased
on their
onRBTs
a specific day e.g., Valentine, Holi,Birthday, Anniversary
Independence

Level 2

Roaming RBT Fixed Album


Caller/GroupRBTBased RBT of a complete album which is played
comprises Gift a Tone
randomly
Set a RBT when out of NetworkDiff callers are played different RBT Gift a Tone to Friend and Fami
I have categorized the above different RBT types into 3 levels based on the
economic benefits to the telecom operator once the customers subscribe to
them. Level 0 is the default level and the starting point of the customers.
Based on the customer type, 3 arrows branch out to level 1 and finally, level
2. Level 2 is where the telecom operators want the customers to come since
that would add maximum economic value to them.

Conclusion
This paper touches on the evolution on MVAS market and how the concepts
like the long tail and crowd-sourcing can be leveraged in order to increase
the overall value of the mobile VAS market. The success of the long tail and
crowd-sourcing hinges on the adoption of WAP and Smart Phone markets
which enable easier access to web-based content and offers an entirely new
market for the MVAS service providers.

The paper also talks about the concept of cross-selling and up-selling in the
internet era and how these concepts are translated into the MVAS solutions.
The concept of market basket analysis using the association data mining rule
has been discussed in detail.

For further study, these cross-selling and up-selling models proposed in the
paper can be refined based on the actual data.

Das könnte Ihnen auch gefallen