Answer:
a ) See step by step explanation
b) 50%   and  50%
Step-by-step explanation:
The definition of type of user x is:
if x < 400 minutes  per month  low user
if x > 400 minutes per month  high user
The value for these states are of course yes  or no  ( the user is low or high user)
Then:
if the user is low today  xt  (0)    we have the probability of 80% ( 0,8) of have the user in the same condition and 20% (0,2) of having him or her as high user
xt = 1       
If the user is high today xt (1) is high today we have the probability of 70% ( 0,7) of having  the user in the same condition and 30% (0,3) of having him or her as high user
a) Then we construct the stochastic matrix 
  Type of user
          (xt)               xt         xt+1  
       Low (0) or L   0,8        0,2
      High(1)  or H   0,3        0,7
The resulting equation system is:
xt  =  0,8* xt  +  0,2* xt+1
xt+1  = 0,3*xt + 0,7* xt+1
1  =  xt + xt+1
Solving for variables         xt+1  =  1  - xt
xt  = 0,8*xt + ( 1 - xt ) * 0,2
xt - 0,8*xt = 0,2 - 0,2xt
xt - 0,6*xt = 0,2
0,4*xt = 0,2
xt = 0,5       and     xt+1  = 0,5
b)  If in January 0,5  ( 50%) of customesr were  low user
In february  teh % of low user customer wll be
xt  =  0,8* xt  +  0,2* xt+1
0,5 = 0,8*0,5 + 0,2*xt+1
0,5 = 0,4 + 0,2 *xt+1
0,1/0,2 = xt+1        xt+1 = 0,5       or   50%
The same answer for March