Page Visits Funnel

Page Visits Funnel

Cool T-Shirts Inc. has asked you to analyze data on visits to their website. Your job is to build a funnel, which is a description of how many people continue to the next step of a multi-step process.

In this case, our funnel is going to describe the following process:

1.A user visits CoolTShirts.com

2.A user adds a t-shirt to their cart

3.A user clicks “checkout”

4.A user actually purchases a t-shirt

# load packages
library(readr)
library(dplyr)

# load data
visits <- read_csv("visits.csv")
cart <- read_csv("cart.csv")
checkout <- read_csv("checkout.csv")
purchase <- read_csv("purchase.csv")

1.Inspect the data frames using head():

(1)visits lists all of the users who have visited the website

(2)cart lists all of the users who have added a t-shirt to their cart

(3)checkout lists all of the users who have started the checkout

(4)purchase lists all of the users who have purchased a t-shirt

# inspect data frames
head(visits)

head(cart)

head(checkout)

head(purchase)

2.Combine visits and cart using a left join.

# define visits_cart here:
visits_cart <- visits %>%
  left_join(cart)

visits_cart

3.How long is the visits data frame?

# define total_visits here:
total_visits <- nrow(visits)
total_visits

[1] 200

4.How many of the timestamps are NA for the column cart_time?

What do these null rows mean?

*Hint:

You can filter a data frame df to only include rows where column_1 has a value of NA using the following code:

”

df %>%

filter(is.na(column_1))

”

# define visit_no_cart here:
visit_no_cart <- visits_cart %>%
  filter(is.na(cart_time)) %>%
  summarize(count = n())

visit_no_cart

5.What percent of users who visited Cool T-Shirts Inc. ended up not placing a t-shirt in their cart?

# calculate visit_no_cart_percent here:
visit_no_cart_percent <- visit_no_cart / total_visits

visit_no_cart_percent

6.Repeat the left join for cart and checkout and count NA values. What percentage of users put items in their cart, but did not proceed to checkout?

# define cart_checkout here:
cart_checkout <- cart %>%
  left_join(checkout)

cart_checkout

# define total_carts here:
total_carts <- nrow(cart)
total_carts

[1] 48

# define cart_no_checkout here:
cart_no_checkout <- cart_checkout %>%
  filter(is.na(checkout_time)) %>%
  summarize(count = n())

cart_no_checkout

# calculate cart_no_checkout_percent here:
cart_no_checkout_percent <- cart_no_checkout / total_carts

cart_no_checkout_percent

7.Join all four steps of the funnel, in order, using a series of left joins. Save the results to the variable all_data.

Examine the result using head().

# define all_data here:
all_data <- visits %>%
  left_join(cart) %>%
  left_join(checkout) %>%
  left_join(purchase)

head(all_data)

8.What percentage of users proceeded to checkout, but did not purchase a t-shirt?

# define total_checkout here:
total_checkout <- nrow(checkout)

total_checkout

[1] 28

# define checkout_no_purchase here:
checkout_no_purchase <- all_data %>%
  filter(!is.na(checkout_time)) %>%
  filter(is.na(purchase_time)) %>%
  summarize(count = n())

checkout_no_purchase

# calculate checkout_no_purchase_percent here:
checkout_no_purchase_percent <- checkout_no_purchase / total_checkout

checkout_no_purchase_percent

9.Which step of the funnel is weakest (i.e., has the highest percentage of users not completing it)?

How might Cool T-Shirts Inc. change their website to fix this problem?

weakest_funnel <- max(visit_no_cart_percent, cart_no_checkout_percent, checkout_no_purchase_percent)

visit_no_cart_percent

cart_no_checkout_percent

checkout_no_purchase_percent

weakest_funnel <- 'visit_no_cart_percent'
weakest_funnel

[1] "visit_no_cart_percent"

Average Time to Purchase

10.Using the giant joined data all_data that you created, let’s calculate the average time from initial visit to final purchase. Start by adding the following column to your DataFrame:

# update all_data with time_to_purchase column here:
all_data <- all_data %>%
  mutate(time_to_purchase = purchase_time - visit_time)

all_data

11.Examine the results using:

# inspect the updated all_data data frame here:
head(all_data)

12.Calculate the average time to purchase

# define time_to_purchase here:
time_to_purchase <- all_data %>%
  summarize(mean_time = mean(time_to_purchase, na.rm = TRUE))

time_to_purchase