Markov Chain

Assumptions underlying Markov

Analysis
There are limited or finite number of possible states (they
are mutually exclusive and collectively exhaustive)
Probability of changing states remains same over time
Probability of next event depends upon the outcome of
last event and thus called first order Markov chain
(Second order Markov chain is the one where choice
depends on last two months choice)
Change from one state to another take place only once
during each time period and time period is equal in
duration. For example, if a customer change brand on
monthly basis then monitoring is done on month by
month basis. If customer changes brand in shorter time
period then reduce time of monitoring.
Example
Suppose an Orange juice company (Brand A) is having
20% market share. They hire a Market Researcher to
predict the effect of aggressive ad campaign on market
share of the company.
He concluded that:
Brand A will stay using Brand A with 90% probability
Those not using Brand A will switch from Brand B to A with
70% probability.
Prepare the transition matrix.
Prepare Initial state matrix.
What % of market share A has after 1 week? After 2
weeks?
In long run what will be the market share of the company?
Example
A new rapid transit system has just started operating. In
the first month of operation, it was found that 25% of
commuters are using the system while 75% are still using
automobiles. The following transition matrix was
determined from records of other rapid transit systems:
Rapid Transit Automobile
Rapid Transit 0.8 0.2
Automobile 0.3 0.7

What is initial state matrix?

What % of commuters will be using new system after 1
month? After 2 months?
What % of commuters using each type of transportation
after it has been in service for a long time? (Stationary
matrix)
Example
Suppose transition matrix for a Markov
Chain is
0.6 0.4
0.2 0.8

Find the stationary matrix S.

 Example
Recently, a market research team has conducted a
survey of consumer buying habits with respect to three
brands of talcum powder in an area. It estimates that,
at present 20% of the customers buy brand A, 50% buy
Brand B and 30% buy brand C. in addition, the firm has
analysed its survey data and has determined the
following brand switching matrix:
Brand Brand next bought
just
bought
A B C
A 0.6 0.3 0.1
B 0.4 0.5 0.1
C 0.2 0.1 0.7

What will be expected distribution of customers two

time periods later?
Example
In a certain, market only two brands of lipsticks A and B,
are sold. Given that a lady last purchased A, there is 80%
chance that she would buy the same brand in the next
purchase, while if a lady purchased B, there is 90% chance
that her next purchase would be brand B. using the
information, develop the transition probability matrix.
Now, calculate:
The probability that if a customer is currently a Band A
purchaser, she will purchase brand B two periods from now;
The probability that if a customer is a brand B purchaser, she
will purchase brand A three periods from now,
The probability that three periods from now, a customer shall
buy brand B, given that the current market share of the two
brands is as follows: Brand A – 70%, Brand B – 30%
The steady state probabilities.
Example
Suppose that new razor blades were introduced in the market
by three companies at the same time. When they were
introduced, each company had an equal share of the market,
but during the first year the following changes took place:
Company A retained 90% of its customers, lost 3% to B and 7%
to brand C
Company B retained 70% of its customers, lost 10% to A and
20% to brand C
Company C retained 80% of its customers, lost 10% to A and
10% to brand B
Assuming that no changes in the buying habits of the consumer
occur.
What are the market shares of the three companies at the end of
first and second year
What are the long run market shares of the three companies.