Decision analysis: profit
Good job translating the posterior click rates into cost distributions! In the meantime, a new company policy has been released. From now on, the goal of the marketing department is not to minimize the costs of campaigns, which was quite ineffective, but rather to maximize the profit. Can you adjust your findings accordingly, knowing that the expected revenue per click from a mobile ad is $3.4
, and the one from a desktop ad is $3
? To calculate the profit, you need to calculate the revenue from all clicks, then subtract the corresponding cost from it.
Everything you have calculated in the previous exercise is available in your workspace: the ads_cost
dictionary as well as the number of click distributions: clothes_num_clicks
and sneakers_num_clicks
.
This exercise is part of the course
Bayesian Data Analysis in Python
Exercise instructions
- Create a dictionary
ads_profit
with four keys:clothes_mobile
,sneakers_mobile
,clothes_desktop
, andsneakers_sneakers
, each holding the profit distribution from corresponding clicks. - Draw a forest plot of
ads_proft
using the credible interval of 99%.
Hands-on interactive exercise
Have a go at this exercise by completing this sample code.
# Calculate profit distributions for each product and platform
ads_profit = {
"clothes_mobile": ____,
"sneakers_mobile": ____,
"clothes_desktop": ____,
"sneakers_desktop": ____,
}
# Draw a forest plot of ads_profit
____
plt.show()