Skip to main content

Budget allocation with Markov model

Import libraries​

from ChannelAttributionPro import *

Set the password​

password="mypassword"

Load data​

Data = pd.read_csv("https://app.channelattribution.net/data/Data.csv")

print(Data)

Set the total budget you want to allocate​

total_budget_new=100000

Set the percentage of the budget you want to reallocate from your previous allocation​

The parameter can be set in the range (0,1]. Setting it to 0.1 means that you will reallocate the 10% of your previous allocation. Drastic changes in your previous allocation are not suggested because the allocation algorithm is based on ROI and thus it is a local optimization algorithm. We suggest to reallocate a small percentage each time​

perc_reall=0.1

For each channel indicate its marketing spend​

This parameter is optional. If you do not know the marketing spends you can set it to None​

tab_costs=pd.DataFrame({'channel':['alpha','iota','eta','beta','theta','lambda','epsilon','zeta','kappa','gamma','mi','delta'],
                        'value':['41111.43','10387.21','23816.66','6743.46','1650.4','523.52','709.94','288.77','269.46','153.57','0.61','0.43']})

Perform attribution with Markov model​

var_value and var_null are optional and can be set to None if you do not have this information​

res_markov=markov_model(Data, var_path="path", var_conv="total_conversions", var_value="total_conversion_value", var_null="total_null", row_sep=";", cha_sep=">", password=password)

Perform budget allocation​

res=markov_budget_allocation(res_markov,total_budget_new,tab_costs,perc_reall,min_perc_budget,server="app.channelattribution.net",password=password)
print(res)