Cyclistic Success
Selmin
2025-12-14
Cyclistic Strategy To Get More Annual Members
OBJECTIVES
Cyclistic, a fictional bike-sharing company operating in Chicago, needs to attract more customers and maximize its annual membership numbers in order to increase its profits. We want to understand how casual riders and annual members use Cyclistic bikes differently. We need to design a new marketing strategy to convert casual riders into annual members.
INSTALLING PACKAGES
# Downloading datasets
if (!require(pacman)) install.packages("pacman")
pacman::p_load( tidyverse,
here,
inspectdf,
plotly,
janitor,
visdat,
highcharter
)
Trips_2020 <- read.csv("~/Coursera/data analytics/Part 8/proje/proje_GSheets/Trips_2020.csv")
Trips1_2019 <- read.csv("~/Coursera/data analytics/Part 8/proje/proje_GSheets/Trips1_2019.csv")
# Calculating duration times for 2020
Trips_2020 <- mutate(Trips_2020,
duration_time = as.numeric(difftime(
ended_at , started_at, units = "sec")))
Trips_2020_dt <- filter(Trips_2020, duration_time > 0)DATA ANALYSIS
# Data Analysis for 2020
## To get an idea about the data set for 2020
summary(Trips_2020)## ride_id rideable_type started_month started_at
## Length:426887 Length:426887 Length:426887 Length:426887
## Class :character Class :character Class :character Class :character
## Mode :character Mode :character Mode :character Mode :character
##
##
##
##
## started_day ended_at trip_duration start_station_name
## Length:426887 Length:426887 Min. : -552 Length:426887
## Class :character Class :character 1st Qu.: 329 Class :character
## Mode :character Mode :character Median : 550 Mode :character
## Mean : 1327
## 3rd Qu.: 949
## Max. :9387024
##
## start_station_id end_station_name end_station_id start_lat
## Min. : 2.0 Length:426887 Min. : 2.0 Min. :41.74
## 1st Qu.: 77.0 Class :character 1st Qu.: 77.0 1st Qu.:41.88
## Median :176.0 Mode :character Median :175.0 Median :41.89
## Mean :209.8 Mean :209.3 Mean :41.90
## 3rd Qu.:298.0 3rd Qu.:297.0 3rd Qu.:41.92
## Max. :675.0 Max. :675.0 Max. :42.06
## NA's :1
## start_lng end_lat end_lng member_casual
## Min. :-87.77 Min. :41.74 Min. :-87.77 Length:426887
## 1st Qu.:-87.66 1st Qu.:41.88 1st Qu.:-87.66 Class :character
## Median :-87.64 Median :41.89 Median :-87.64 Mode :character
## Mean :-87.64 Mean :41.90 Mean :-87.64
## 3rd Qu.:-87.63 3rd Qu.:41.92 3rd Qu.:-87.63
## Max. :-87.55 Max. :42.06 Max. :-87.55
## NA's :1 NA's :1
## duration_time
## Min. : -552
## 1st Qu.: 329
## Median : 550
## Mean : 1326
## 3rd Qu.: 949
## Max. :9383424
## filter(Trips_2020, duration_time == -552) # to find wrong data## ride_id rideable_type started_month started_at started_day
## 1 6FABADDD595AF922 docked_bike Mar 2020-03-27 15:15:26 Friday
## ended_at trip_duration start_station_name start_station_id
## 1 2020-03-27 15:06:14 -552 HQ QR 675
## end_station_name end_station_id start_lat start_lng end_lat end_lng
## 1 HQ QR 675 41.8899 -87.6803 41.8899 -87.6803
## member_casual duration_time
## 1 casual -552Trips_2020 %>% slice(1:2000) %>% vis_dat()
# Data Analysis for 2019
## To get an idea about the data set for 2019
summary(Trips1_2019)## trip_id start_month start_time start_day
## Min. :21742443 Length:365068 Length:365068 Length:365068
## 1st Qu.:21848766 Class :character Class :character Class :character
## Median :21961830 Mode :character Mode :character Mode :character
## Mean :21960872
## 3rd Qu.:22071824
## Max. :22178528
##
## end_time bikeid tripduration t_duration
## Length:365068 Min. : 1 Length:365068 Min. : 61
## Class :character 1st Qu.:1777 Class :character 1st Qu.: 326
## Mode :character Median :3489 Mode :character Median : 524
## Mean :3429 Mean : 1016
## 3rd Qu.:5157 3rd Qu.: 866
## Max. :6471 Max. :10628400
##
## from_station_id from_station_name from_address Latitude
## Min. : 2.0 Length:365068 Length:365068 Min. :41.74
## 1st Qu.: 76.0 Class :character Class :character 1st Qu.:41.88
## Median :170.0 Mode :character Mode :character Median :41.89
## Mean :198.1 Mean :41.90
## 3rd Qu.:287.0 3rd Qu.:41.91
## Max. :665.0 Max. :42.32
## NA's :39
## Longitude to_station_id to_station_name to_address
## Min. :-87.94 Min. : 2.0 Length:365068 Length:365068
## 1st Qu.:-87.65 1st Qu.: 76.0 Class :character Class :character
## Median :-87.64 Median :168.5 Mode :character Mode :character
## Mean :-87.64 Mean :198.6
## 3rd Qu.:-87.63 3rd Qu.:287.0
## Max. :-87.55 Max. :665.0
## NA's :39
## Latitude.1 Longitude.1 usertype gender
## Min. :41.74 Min. :-87.94 Length:365068 Length:365068
## 1st Qu.:41.88 1st Qu.:-87.65 Class :character Class :character
## Median :41.89 Median :-87.64 Mode :character Mode :character
## Mean :41.90 Mean :-87.64
## 3rd Qu.:41.91 3rd Qu.:-87.63
## Max. :42.32 Max. :-87.55
## NA's :30 NA's :30
## birthyear
## Min. :1900
## 1st Qu.:1975
## Median :1985
## Mean :1982
## 3rd Qu.:1990
## Max. :2003
## NA's :18023Trips1_2019 %>% slice(1:2000) %>% vis_dat()
# Categorical Analysis for 2020
min(Trips_2020$duration_time, na.rm = TRUE)## [1] -552max(Trips_2020$duration_time, na.rm = TRUE)## [1] 9383424show_plot(inspect_cat(Trips_2020))
# Numerical Analysis for 2019
show_plot(inspect_num(Trips1_2019))
min(Trips1_2019$t_duration, na.rm = TRUE)## [1] 61max(Trips1_2019$t_duration, na.rm = TRUE)## [1] 10628400# Numerical Analysis for 2020
show_plot(inspect_num(Trips_2020))
# Categorical Analysis for 2019
show_plot(inspect_cat(Trips1_2019))
Bilboards can be put in the Locations used by Casual users
For 2020 Most used locations
# Most used locations for 2020
ggplot(data = Trips_2020, mapping = aes(x = start_lat,
y = start_lng,
fill = member_casual)) +
geom_col() +
labs(title = "Most Used Location",
subtitle = "for the year 2020",
x = "Latitude",
y = "Longtitude")
Duration times per member for 2020
# Duration times per member for 2020
ggplot(data = Trips_2020_dt, mapping = aes(x = duration_time,
y = end_station_id,
color = member_casual)) +
geom_line(size = 0.55,
alpha = 0.65) +
scale_color_brewer(palette = "Dark2") +
theme_light() +
labs(title = "Showing duration time by end station",
subtitle = "For the year 2020",
x = "Duration Time",
y = "End Station Id",
color = "Member Casual")
While members are concentrated in the same stations, casual users are spread across different stations
Most used locations for 2019
# Most used locations for 2019
ggplot(data = Trips1_2019, mapping = aes(x = Latitude,
y = Longitude,
color = usertype)) +
geom_point(shape = 18, alpha = 0.75) +
geom_smooth(method = "glm",
se = FALSE,
color = "orange") +
labs(title = "Most used locations",
subtitle = "for the year 2019")
While members are concentrated in the same stations, casual users are spread across different stations
Busiest locations for 2019
# Busiest locations for 2019
ggplot(data = Trips1_2019, mapping = aes(x = Latitude.1,
y = Longitude.1,
color = gender)) +
geom_line() +
labs(title = "Busiest locations by gender",
subtitle = "for the year 2019",
x = "Latitude",
y = "Longitiude")
Female users appears to be less frequent
Duration times per member for 2019
# Duration times per member for 2019
ggplot(data = Trips1_2019, mapping = aes(x = t_duration,
y = to_station_id,
color = usertype)) +
geom_line(size = 0.55,
alpha = 0.65) +
facet_wrap(vars(gender)) +
scale_color_brewer(palette = "Dark2") +
theme_light() +
labs(title = "Showing duration time by end station",
subtitle = "for the year 2019",
x = "Duration Time",
y = "End Station Id",
color = "Usertype")
Member and Casual users summary and mean for 2020
# Memeber and Casual users summary and mean for 2020
Trip2020_sum1<- Trips_2020 %>%
filter(duration_time > 0) %>%
group_by(member_casual) %>%
summarize(sum(duration_time),
mean(duration_time) , count = n()) %>%
arrange(desc(count))
colnames(Trip2020_sum1)[2] <- "sum_duration_time"
colnames(Trip2020_sum1)[3] <- "mean_duration_time"
barplot(table(Trip2020_sum1$member_casual), names.arg = Trip2020_sum1$'sum_duration_time', main= "Duration times by member and casual For 2020", xlab = "member and casual", col = terrain.colors(2))
In 2020, the number of members was greater than the number of casual users
Subscriber and customer users summary and mean for 2019
# Subscriber and customer users summary and mean for 2019
Trip2019_sum1 <- Trips1_2019 %>%
group_by(usertype, gender) %>%
summarize(sum(t_duration),
mean(t_duration), count = n()) %>%
arrange(desc(gender))
colnames(Trip2019_sum1)[3] <- "sum_duration_time"
colnames(Trip2019_sum1)[4] <- "mean_duration_time"
user = c("Customer", "Subscriber")
colors = c("pink", "blue")
barplot(Trip2019_sum1$sum_duration_time, names.arg = Trip2019_sum1$'gender',
main = "Duration times by Gender For 2019 ", xlab = "Gender", ylab = "Duration Times", col= colors, beside = TRUE)
legend("topleft", user, cex = 0.6, fill = colors)
In 2019, the usage time for Male was longer than the usage time for Female users
Subscriber and customer users duration time by day for 2019
# Subscriber and customer users duration time by day for 2019
Trip2019_day <- Trips1_2019 %>%
group_by(start_day, usertype) %>%
summarize(sum(t_duration),
mean(t_duration), count = n()) %>%
arrange(desc(start_day))
colnames(Trip2019_day)[3] <- "sum_duration_time"
colnames(Trip2019_day)[4] <- "mean_duration_time"
Trip2019_day <- Trip2019_day %>%
mutate(nstart_day = recode(start_day,
"Wednesday" = "Wed",
"Tuesday" = "Tue",
"Thursday" = "Thu",
"Sunday" = "Sun",
"Saturday" = "Sat",
"Monday" = "Mon",
"Friday" = "Fri"))
colors = c("green", "orange")
user = c("Customer", "Subscriber")
barplot(Trip2019_day$sum_duration_time,
main = "Trip Times by Days for 2019",
names.arg = Trip2019_day$'nstart_day',
space=c(0.5,0.1,0.5,0.1,0.5,0.1,0.5,0.1),
xlab = "Days", ylab = "Trip Times",
col = colors,
beside = TRUE)
legend("topleft", user, cex = 0.5, fill = colors)
In 2019 Customers use bicycles more on Tuesdays and Saturdays, while subscribers use them more during the weekdays
Member and Casual users duration time by day for 2020
# Member and Casual users duration time by day for 2020
Trip2020_day <- Trips_2020_dt %>%
group_by(started_day, member_casual) %>%
summarize(sum(trip_duration),
mean(trip_duration), count = n()) %>%
arrange(desc(started_day))
colnames(Trip2020_day)[3] <- "sum_duration_time"
colnames(Trip2020_day)[4] <- "mean_duration_time"
Trip2020_day <- Trip2020_day %>%
mutate(nstarted_day = recode(started_day,
"Wednesday" = "Wed",
"Tuesday" = "Tue",
"Thursday" = "Thu",
"Sunday" = "Sun",
"Saturday" = "Sat",
"Monday" = "Mon",
"Friday" = "Fri"))
colors = c("red", "blue")
user = c("casual", "member")
barplot(Trip2020_day$sum_duration_time, names.arg = Trip2020_day$'nstarted_day',
main = "Trip Times by Days for 2020",
space=c(0.5,0.1,0.5,0.1,0.5,0.1,0.5,0.1),
xlab = "Days", ylab = "Trip Times", col = colors)
legend("topleft", user, cex = 0.5, fill = colors)
In 2020 Casuals use bicycles more on Sundays, while Members use them more during the weekdays
Subscriber and customer users duration time by month for 2019
# Subscriber and customer users duration time by month for 2019
Trip2019_month <- Trips1_2019 %>%
group_by(start_month, usertype) %>%
summarize(sum(t_duration),
mean(t_duration), count = n()) %>%
arrange(desc(start_month))
colnames(Trip2019_month)[3] <- "sum_duration_time"
colnames(Trip2019_month)[4] <- "mean_duration_time"
colors = c("green", "orange")
user = c("Customer", "Subscriber")
barplot(Trip2019_month$sum_duration_time, main = "Trip Times by Months for 2019",
names.arg = Trip2019_month$'start_month',
xlab = "Months", ylab = "Trip Times",
col = colors, beside = TRUE)
legend("topleft", user, cex = 0.5, fill = colors)
In 2019 Customers and Subscribers use bicycles more in spring times than winter times
Member and Casual users duration time by day for 2020
# Member and Casual users duration time by day for 2020
Trip2020_month <- Trips_2020_dt %>%
group_by(started_month, member_casual) %>%
summarize(sum(trip_duration),
mean(trip_duration), count = n()) %>%
arrange(desc(started_month))
colnames(Trip2020_month)[3] <- "sum_duration_time"
colnames(Trip2020_month)[4] <- "mean_duration_time"
colors = c("red", "blue")
user = c("casual", "member")
barplot(Trip2020_month$sum_duration_time, main = "Trip Times by Month for 2020",
names.arg = Trip2020_month$'started_month',
xlab = "Months", ylab = "Trip Times",
col = colors, beside = TRUE)
legend("topleft", user, cex = 0.5, fill = colors)
In 2020 Casuals and Members use bicycles more in spring times than winter times
Most of used start stations for casual for 2020
# Most of used start stations for casual for 2020
Trip2020_sum2 <- Trips_2020 %>%
filter(member_casual == "casual") %>%
group_by(start_station_id) %>%
summarize(count = n()) %>%
arrange(desc(count))
# highcharter::hchart(Trip2020_sum2$start_station_id)
top15_rows2 <- head(Trip2020_sum2, 15)
top15_rows2## # A tibble: 15 × 2
## start_station_id count
## <int> <int>
## 1 675 3766
## 2 76 1590
## 3 35 1530
## 4 3 998
## 5 90 779
## 6 85 631
## 7 177 567
## 8 43 495
## 9 6 490
## 10 341 465
## 11 199 428
## 12 268 402
## 13 255 398
## 14 623 391
## 15 2 335Most of used start stations for customer for 2019
# Most of used start stations for customer for 2019
Trip2019_sum2 <- Trips1_2019 %>%
filter(usertype == "Customer") %>%
group_by(from_station_id) %>%
summarize(count = n()) %>%
arrange(desc(count))
# highcharter::hchart(Trip2019_sum2$from_station_id)
top15_rows1 <- head(Trip2019_sum2, 15)
top15_rows1## # A tibble: 15 × 2
## from_station_id count
## <int> <int>
## 1 35 1219
## 2 76 1142
## 3 3 834
## 4 90 627
## 5 85 386
## 6 341 362
## 7 43 344
## 8 6 342
## 9 97 299
## 10 2 256
## 11 195 239
## 12 177 228
## 13 77 220
## 14 623 206
## 15 268 204SUMMARY
Although the total trip time appears higher for members, the average trip time is higher for casual users. To convert casual users into members, billboards could be placed at frequently used stations. Advertisements targeting women would also increase the number of female members. Since casual users use the service more on weekends and during the spring months, seasonal or weekend memberships could be offered. Digital media can be used to inform people about these options.