Gregory Elsey 2025-01-07
This R Markdown document is the case study for the Google Data Analytics course done through Coursera. This is the capstone project for said course. This case study was primarily done through R with Microsoft Excel being used briefly. If you want to work with the rmd file for yourself, see the Prepare section for more information on the data used for this case study.
You are a junior data analyst working on the marketing analyst team at Cyclistic, a bike-share company in Chicago. The director of marketing believes the company’s future success depends on maximizing the number of annual memberships. Therefore, your team wants to understand how casual riders and annual members use Cyclistic bikes differently. From these insights, your team will design a new marketing strategy to convert casual riders into annual members. But first, Cyclistic executives must approve your recommendations, so they must be backed up with compelling data insights and professional data visualizations.
● Cyclistic: A bike-share program that features more than 5,800 bicycles and 600 docking stations. Cyclistic sets itself apart by also oering reclining bikes, hand tricycles, and cargo bikes, making bike-share more inclusive to people with disabilities and riders who can’t use a standard two-wheeled bike. The majority of riders opt for traditional bikes; about 8% of riders use the assistive options. Cyclistic users are more likely to ride for leisure, but about 30% use the bikes to commute to work each day.
● Lily Moreno: The director of marketing and your manager. Moreno is responsible for the development of campaigns and initiatives to promote the bike-share program. These may include email, social media, and other channels.
● Cyclistic marketing analytics team: A team of data analysts who are responsible for collecting, analyzing, and reporting data that helps guide Cyclistic marketing strategy. You joined this team six months ago and have been busy learning about Cyclistic’s mission and business goals—as well as how you, as a junior data analyst, can help Cyclistic achieve them.
● Cyclistic executive team: The notoriously detail-oriented executive team will decide whether to approve the recommended marketing program.
In 2016, Cyclistic launched a successful bike-share oering. Since then, the program has grown to a eet of 5,824 bicycles that are geotracked and locked into a network of 692 stations across Chicago. The bikes can be unlocked from one station and returned to any other station in the system anytime.
Until now, Cyclistic’s marketing strategy relied on building general awareness and appealing to broad consumer segments. One approach that helped make these things possible was the exibility of its pricing plans: single-ride passes, full-day passes, and annual memberships. Customers who purchase single-ride or full-day passes are referred to as casual riders. Customers who purchase annual memberships are Cyclistic members.
Cyclistic’s nance analysts have concluded that annual members are much more protable than casual riders. Although the pricing exibility helps Cyclistic aract more customers, Moreno believes that maximizing the number of annual members will be key to future growth. Rather than creating a marketing campaign that targets all-new customers, Moreno believes there is a solid opportunity to convert casual riders into members. She notes that casual riders are already aware of the Cyclistic program and have chosen Cyclistic for their mobility needs.
Moreno has set a clear goal: Design marketing strategies aimed at converting casual riders into annual members. In order to do that, however, the team needs to beer understand how annual members and casual riders dier, why casual riders would buy a membership, and how digital media could aect their marketing tactics. Moreno and her team are interested in analyzing the Cyclistic historical bike trip data to identify trends.
This case study will use the Google data Analysis Process to complete this case.
-
Ask: Define the problem and confirm stakeholder expectations
-
Prepare: Collect and store data for analysis
-
Process: Clean and transform data to ensure integrity
-
Analyse: Use data analysis tools to draw conclusions
-
Share: Interpret and communicate results to others to make data-driven decisions
-
Act: Put your insights to work in order to solve the original problem
Task: Identify annual members and causal riders and analyze the different riding habits of the two listed groups. Then, come up with possible different ways on how to convert the casual rider to being a member.
Lily Moreno: The director of marketing and your manager.
Cyclistic marketing analytics team: A team of data analysts who collect, analyze and report data that helps guide Cyclistic’s marketing strategy.
Cyclistic executive team: A detail-oriented executive team that will decide whether to approve the marketing program you will recommend.
This case study uses Cyclistics public historical trip data to help analyze and identify trends. The data is avaliable on Motivate International Inc. under this license. If you would like to download this case study yourself download the csvs labeled 202312-divvy-tripdata to 202411-divvy-tripdata from the Motivate International link to be able to work with the data for yourself.
The case study uses the csv files ranging from December 2023 to November 2024 that were contain the zip files of the Motivate International data base. Microsoft Excel was used to make sure the data was all there. In each of those files were 13 different variables
The data set contains 13 different variables:
- ride_id: The id of the rideable being used
- rideable_type: The type of rideable used
- started_at: What day and time the ride began
- ended_at: What day and time the ride ended
- start_station_name: What station the ride began at
- start_station_id: The id of the starting station
- end_station_name: What station the ride ended at
- end_station_id: The id of the ending station
- start_lat: The beginning Latitude of the ride
- start_lng: The beginning Longitude of the ride
- end_lat: The ending Latitude of the ride
- end_lng: The ending Longitude of the ride
- member_casual: Whether the the rider is an annual member or a casual rider
This section will contain both processing and analyzing the data, and is where the r coding will take place.
First we must load the libraries that we will be used for this case study.
# Load all needed libraries
library(tidyverse)## ── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
## ✔ dplyr 1.1.4 ✔ readr 2.1.5
## ✔ forcats 1.0.0 ✔ stringr 1.5.1
## ✔ ggplot2 3.5.1 ✔ tibble 3.2.1
## ✔ lubridate 1.9.4 ✔ tidyr 1.3.1
## ✔ purrr 1.0.2
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors
library(janitor)##
## Attaching package: 'janitor'
##
## The following objects are masked from 'package:stats':
##
## chisq.test, fisher.test
library(lubridate)Then the environment needs to be emptied in order for everything to work properly (Note: If you are planning on downloading this markdown then you will need to adjust this step to whatever directory you will be using to run this code).
# Cleaning enviroment
rm(list=ls())Now lets load all of the csvs and combine them into one big file for an easier analysis!
# Csv calling and compiling into a single file
bike_rides <- list.files(path="./compiled csvs", full.names = TRUE) %>%
lapply(read.csv) %>%
bind_rows# Head call to see combined data
head(bike_rides)## ride_id rideable_type started_at ended_at
## 1 C9BD54F578F57246 electric_bike 2023-12-02 18:44:01 2023-12-02 18:47:51
## 2 CDBD92F067FA620E electric_bike 2023-12-02 18:48:19 2023-12-02 18:54:48
## 3 ABC0858E52CBFC84 electric_bike 2023-12-24 01:56:32 2023-12-24 02:04:09
## 4 F44B6F0E8F76DC90 electric_bike 2023-12-24 10:58:12 2023-12-24 11:03:04
## 5 3C876413281A90DF electric_bike 2023-12-24 12:43:16 2023-12-24 12:44:57
## 6 28C0D6EFB81E1769 electric_bike 2023-12-24 13:59:57 2023-12-24 14:10:57
## start_station_name start_station_id end_station_name end_station_id start_lat
## 1 41.92
## 2 41.92
## 3 41.89
## 4 41.95
## 5 41.92
## 6 41.91
## start_lng end_lat end_lng member_casual
## 1 -87.66 41.92 -87.66 member
## 2 -87.66 41.89 -87.64 member
## 3 -87.62 41.90 -87.64 member
## 4 -87.65 41.94 -87.65 member
## 5 -87.64 41.93 -87.64 member
## 6 -87.63 41.88 -87.65 member
There’s almost 6 million objects within this data set! We need to weed out any data that would hinder the analysis later on.
# Check to see if there are any extra pieces of data
unique(bike_rides$member_casual)## [1] "member" "casual"
unique(bike_rides$rideable_type)## [1] "electric_bike" "classic_bike" "electric_scooter"
There may be some unnecessary and null data within the data set, so lets get rid of that.
# Getting rid of the extra data and null data
bike_rides <- drop_na(bike_rides)There might also be some test data so lets make sure we aren’t missing anything.
# Check for test cases
unique(bike_rides$start_station_name[grep("TEST", bike_rides$start_station_name)])## character(0)
unique(bike_rides$end_station_name[grep("TEST", bike_rides$end_station_name)])## character(0)
unique(bike_rides$start_station_name[grep("test", bike_rides$start_station_name)])## character(0)
unique(bike_rides$end_station_name[grep("test", bike_rides$end_station_name)])## character(0)
There doesn’t seem to be any test data, so we can go forward with parsing the dates to make them easier to work with.
# Changing the format of started_at & ended_at to a date format
bike_rides$started_at <- lubridate::ymd_hms(bike_rides$started_at)
bike_rides$ended_at <- lubridate::ymd_hms(bike_rides$ended_at)# Creating an hour column for both start and end hour
bike_rides$start_hour <- lubridate::hour(bike_rides$started_at)
bike_rides$end_hour <- lubridate::hour(bike_rides$ended_at)# Parsing date from started_at and ended_at
bike_rides$start_date <- as.Date(bike_rides$started_at)
bike_rides$end_date <- as.Date(bike_rides$ended_at)# Parsing month the bike was rented from started at
bike_rides$month <- format(as.Date(bike_rides$start_date), "%B")# Parsing what day of the week the bike was rented
bike_rides$day_of_week <- format(as.Date(bike_rides$start_date), "%A")Now lets order the Months and Days of the week to make them much easier to work with
# Convert months order from December to November
bike_rides$month <- ordered(bike_rides$month, levels=c("December","January", "February",
"March", "April", "May", "June", "July", "August", "September", "October",
"November"))
# Convert days of week order to Sunday to Saturday
bike_rides$day_of_week <- ordered(bike_rides$day_of_week, levels=c("Sunday",
"Monday", "Tuesday", "Wednesday", "Thursday", "Friday", "Saturday"))You may have noticed that the order is from December to November and not January to December. The reason it’s in this order is because the data starts in December 2023, so we need to make sure that when we graph the data we start from the earliest time frame, which happens to be in December.
After parsing lets take a quick look at the data frame to see all of the changes made.
# Check on data
head(bike_rides)## ride_id rideable_type started_at ended_at
## 1 C9BD54F578F57246 electric_bike 2023-12-02 18:44:01 2023-12-02 18:47:51
## 2 CDBD92F067FA620E electric_bike 2023-12-02 18:48:19 2023-12-02 18:54:48
## 3 ABC0858E52CBFC84 electric_bike 2023-12-24 01:56:32 2023-12-24 02:04:09
## 4 F44B6F0E8F76DC90 electric_bike 2023-12-24 10:58:12 2023-12-24 11:03:04
## 5 3C876413281A90DF electric_bike 2023-12-24 12:43:16 2023-12-24 12:44:57
## 6 28C0D6EFB81E1769 electric_bike 2023-12-24 13:59:57 2023-12-24 14:10:57
## start_station_name start_station_id end_station_name end_station_id start_lat
## 1 41.92
## 2 41.92
## 3 41.89
## 4 41.95
## 5 41.92
## 6 41.91
## start_lng end_lat end_lng member_casual start_hour end_hour start_date
## 1 -87.66 41.92 -87.66 member 18 18 2023-12-02
## 2 -87.66 41.89 -87.64 member 18 18 2023-12-02
## 3 -87.62 41.90 -87.64 member 1 2 2023-12-24
## 4 -87.65 41.94 -87.65 member 10 11 2023-12-24
## 5 -87.64 41.93 -87.64 member 12 12 2023-12-24
## 6 -87.63 41.88 -87.65 member 13 14 2023-12-24
## end_date month day_of_week
## 1 2023-12-02 December Saturday
## 2 2023-12-02 December Saturday
## 3 2023-12-24 December Sunday
## 4 2023-12-24 December Sunday
## 5 2023-12-24 December Sunday
## 6 2023-12-24 December Sunday
Now we should convert the ride length into a minute format for easier calculations when we analyze the data.
# Calculate trip time into minutes
bike_rides$ride_length <- difftime(bike_rides$ended_at, bike_rides$started_at)
bike_rides$ride_length <- bike_rides$ride_length/60
bike_rides$ride_length <- round(bike_rides$ride_length, 2)Now lets get a more in depth look into the changes we made.
# See all the new columns with str called
str(bike_rides)## 'data.frame': 5898929 obs. of 20 variables:
## $ ride_id : chr "C9BD54F578F57246" "CDBD92F067FA620E" "ABC0858E52CBFC84" "F44B6F0E8F76DC90" ...
## $ rideable_type : chr "electric_bike" "electric_bike" "electric_bike" "electric_bike" ...
## $ started_at : POSIXct, format: "2023-12-02 18:44:01" "2023-12-02 18:48:19" ...
## $ ended_at : POSIXct, format: "2023-12-02 18:47:51" "2023-12-02 18:54:48" ...
## $ start_station_name: chr "" "" "" "" ...
## $ start_station_id : chr "" "" "" "" ...
## $ end_station_name : chr "" "" "" "" ...
## $ end_station_id : chr "" "" "" "" ...
## $ start_lat : num 41.9 41.9 41.9 42 41.9 ...
## $ start_lng : num -87.7 -87.7 -87.6 -87.7 -87.6 ...
## $ end_lat : num 41.9 41.9 41.9 41.9 41.9 ...
## $ end_lng : num -87.7 -87.6 -87.6 -87.7 -87.6 ...
## $ member_casual : chr "member" "member" "member" "member" ...
## $ start_hour : int 18 18 1 10 12 13 9 8 18 6 ...
## $ end_hour : int 18 18 2 11 12 14 9 8 18 6 ...
## $ start_date : Date, format: "2023-12-02" "2023-12-02" ...
## $ end_date : Date, format: "2023-12-02" "2023-12-02" ...
## $ month : Ord.factor w/ 12 levels "December"<"January"<..: 1 1 1 1 1 1 1 1 1 1 ...
## $ day_of_week : Ord.factor w/ 7 levels "Sunday"<"Monday"<..: 7 7 1 1 1 1 1 1 2 1 ...
## $ ride_length : 'difftime' num 3.83 6.48 7.62 4.87 ...
## ..- attr(*, "units")= chr "secs"
Looks like the conversion is successful, but now we need to do a quick check for any ride times that lasted for 0 minutes to possibly remove them. These pieces of data are useless, as they don’t give any insight into the data we will be analyzing.
#remove ride times that are 0 minutes
bike_rides <- subset(bike_rides, ride_length > 0)Finally, we need to convert the ride length to a numeric format to be able to do calculations for the analysis we will be doing.
# Convert ride_length to numeric for calculations
bike_rides$ride_length <- as.numeric(as.character(bike_rides$ride_length))
bike_rides <- filter(bike_rides, ride_length > 0)Now we are ready to really start taking a look at the data and analyze it.
First lets see how many rides the casual and member riders did from December to November.
# Begin summarizing and analyzing
bike_rides %>%
group_by(member_casual) %>%
summarise(Amount_of_rides = n())## # A tibble: 2 × 2
## member_casual Amount_of_rides
## <chr> <int>
## 1 casual 2158707
## 2 member 3739381
It seems that there are more members riding than causal riders.Let’s now take a look at the summaries of each groups ride length
# First we look at the general summary of the ride length
summary(bike_rides$ride_length)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 0.01 5.54 9.70 15.48 17.19 1509.37
# Then we compare the member_casual summaries
bike_rides %>%
group_by(member_casual) %>%
summarise(Minimum_ride_length = min(ride_length), Maximum_ride_length = max(ride_length))## # A tibble: 2 × 3
## member_casual Minimum_ride_length Maximum_ride_length
## <chr> <dbl> <dbl>
## 1 casual 0.01 1509.
## 2 member 0.01 1500.
bike_rides %>%
group_by(member_casual) %>%
summarise(Average_ride_length = mean(ride_length), Median_ride_length = median(ride_length))## # A tibble: 2 × 3
## member_casual Average_ride_length Median_ride_length
## <chr> <dbl> <dbl>
## 1 casual 21.1 12
## 2 member 12.2 8.69
Now with those initial looks into the different users, lets start getting into specifics.
# Checking amount of bike rides between Members vs Casuals
bike_rides %>%
group_by(day_of_week, member_casual) %>%
summarise(Amount_of_rides = n()) ## `summarise()` has grouped output by 'day_of_week'. You can override using the
## `.groups` argument.
## # A tibble: 14 × 3
## # Groups: day_of_week [7]
## day_of_week member_casual Amount_of_rides
## <ord> <chr> <int>
## 1 Sunday casual 369935
## 2 Sunday member 418953
## 3 Monday casual 250875
## 4 Monday member 528209
## 5 Tuesday casual 231410
## 6 Tuesday member 568101
## 7 Wednesday casual 271241
## 8 Wednesday member 617738
## 9 Thursday casual 267642
## 10 Thursday member 580997
## 11 Friday casual 320123
## 12 Friday member 537955
## 13 Saturday casual 447481
## 14 Saturday member 487428
bike_rides %>%
group_by(month, member_casual) %>%
summarise(Amount_of_rides = n()) ## `summarise()` has grouped output by 'month'. You can override using the
## `.groups` argument.
## # A tibble: 24 × 3
## # Groups: month [12]
## month member_casual Amount_of_rides
## <ord> <chr> <int>
## 1 December casual 51494
## 2 December member 172285
## 3 January casual 24339
## 4 January member 120149
## 5 February casual 46957
## 6 February member 175861
## 7 March casual 82218
## 8 March member 218967
## 9 April casual 131366
## 10 April member 282992
## # ℹ 14 more rows
# Checking the usage of different bike types between Member & Casual Riders
bike_rides %>%
group_by(rideable_type) %>%
summarise(Amount_of_rides = n())## # A tibble: 3 × 2
## rideable_type Amount_of_rides
## <chr> <int>
## 1 classic_bike 2762271
## 2 electric_bike 2991480
## 3 electric_scooter 144337
bike_rides %>%
group_by(rideable_type, member_casual) %>%
summarise(Amount_of_rides = n())## `summarise()` has grouped output by 'rideable_type'. You can override using the
## `.groups` argument.
## # A tibble: 6 × 3
## # Groups: rideable_type [3]
## rideable_type member_casual Amount_of_rides
## <chr> <chr> <int>
## 1 classic_bike casual 976298
## 2 classic_bike member 1785973
## 3 electric_bike casual 1097194
## 4 electric_bike member 1894286
## 5 electric_scooter casual 85215
## 6 electric_scooter member 59122
After looking at the numbers of some of the data, let now visualize this data for a clearer look
# Changing how the y-axis displays numbers for easier readability
options(scipen=10000)First lets visualize the amount of Casual riders versus Member riders to get a clearer understanding.
# Incorporating percentage to be shown on the pie chart
percent_bike_rides <- bike_rides %>%
group_by(member_casual) %>%
summarise(num_of_rides = n())
percent_bike_rides <- percent_bike_rides %>%
mutate(percent=paste0(round(num_of_rides/sum(num_of_rides)*100,2), "%"))
# casual rider Vs member rider ride lengths pie graph
percent_bike_rides %>%
group_by(member_casual) %>%
ggplot(aes(x="", y = num_of_rides, fill = member_casual)) + geom_col(color = "black") +
geom_text(aes(label = percent), position = position_stack(vjust = .5)) +
coord_polar(theta = "y") +
theme_void()
Now lets look at when each group of riders go out and rent a rideable.
# casual rider vs member rider day of week popularity bar graph
bike_rides %>%
group_by(member_casual, day_of_week) %>%
summarise(num_of_rides = n()) %>%
arrange(member_casual, day_of_week) %>%
ggplot(aes(x = day_of_week, y = num_of_rides, fill = member_casual)) +
geom_col(position = "dodge") + labs(title = "Total Number of Rides: Casual vs Member Riders by Day",
x = "Week Day", y = "Number of Rides", fill='User Type')## `summarise()` has grouped output by 'member_casual'. You can override using the
## `.groups` argument.
# casual rider vs member rider month popularity bar graph
bike_rides %>%
group_by(member_casual, month) %>%
summarise(num_of_rides = n()) %>%
arrange(member_casual, month) %>%
ggplot(aes(x = month, y = num_of_rides, fill = member_casual)) +
geom_col(position = "dodge") + labs(title = "Total Number of Rides: Casual vs Member Riders by Month",
x = "Month", y = "Number of Rides", fill='User Type') +
theme(axis.text.x = element_text(angle = 45, hjust = 1))## `summarise()` has grouped output by 'member_casual'. You can override using the
## `.groups` argument.
We also should check on their average ride times to get a general idea about each type of riders habits on how long they ride for.
# Average ride time per day of week calculation
avg_ride_time_daily <- bike_rides %>%
group_by(member_casual, day_of_week) %>%
summarise(avg_ride_time = mean(ride_length))## `summarise()` has grouped output by 'member_casual'. You can override using the
## `.groups` argument.
# Average ride time per day of week calculation
avg_ride_time_daily %>%
group_by(member_casual, day_of_week) %>%
ggplot() +
geom_line(aes(x=day_of_week, y=avg_ride_time, group = member_casual, colour = member_casual)) +
labs(title = "Average ride time: Casual vs Member Riders", colour='User type') +
xlab("Day of Week") +
ylab("Average Ride Time (Minutes)")
# Average ride time per month calculation monthly
avg_ride_time_monthly <- bike_rides %>%
group_by(member_casual, month) %>%
summarise(avg_ride_time = mean(ride_length))## `summarise()` has grouped output by 'member_casual'. You can override using the
## `.groups` argument.
# Average ride time line graph montly
avg_ride_time_monthly %>%
group_by(member_casual, month) %>%
ggplot() +
geom_line(aes(x=month, y=avg_ride_time, group = member_casual, colour = member_casual)) +
labs(title = "Average ride time: Casual vs Member Riders", colour='User type') +
xlab("Month") +
ylab("Average Ride Time (Minutes)") +
theme(axis.text.x = element_text(angle = 45, hjust = 1))
Finally we’ll check which of the rideables are most popular between casual and member riders.
# Comparing the popularity between rideables divided between member and casual riders.
bike_rides %>%
group_by(rideable_type, member_casual) %>%
summarise(num_of_rides = n()) %>%
ggplot(aes(x = member_casual, y = num_of_rides, fill = rideable_type)) +
geom_col(position = "dodge") + labs(title = "Total Number of Rides: Casual vs Member Riders by Rideable",
x = "Month", y = "Number of Rides", fill='Rideable Type')## `summarise()` has grouped output by 'rideable_type'. You can override using the
## `.groups` argument.
After taking a look at the data and some visualizations, there are some conclusions we can make about said data.
Most users of Cyclistic’s services are already members of the program with 63.4% of users from December 2023 to November 2024 being said members, while only 36.6% are casual riders using Cyclistic’s services.
Casual riders were much more likely to rent a rideable on the weekend with Saturday being the most popular day to do so with Sunday not too far behind. Member riders were the opposite, as they are more likely to ride during the week with Wednesday being the most popular day for them to rent a rideable.
Both Casual and Member riders were much more likely to use a rideable from May to October. This means that from late spring to early fall more people were using Cyclistic’s services, with September being the most popular month the rent for both groups.
Casual riders had a much higher overall average ride time during the week with it spiking to almost 25 minutes on the weekend. On the other hand, members have a much shorter average ride time, with them not even breaking 15 minutes.
Casual Riders also had the highest average ride times between the months of May and July. Meanwhile, Members would have their highest Average ride times from May to August, but they were still much shorter on average compared to casual riders.
Both Casual and Member riders prefer to use electric bikes over the 2 other options available.
Casual riders are much more likely to use Cyclistic’s services for leisure because of them most likely riding during weekends and summer and having an overall higher average ride time, the general peak of outside leisure activities.
Member riders are more likely to use Cyclistic’s services for to get to places for work/school due to them riding more often on weekdays, and having overall shorter average ride time.
Electric bikes are the most popular rideable due to their convenience and ability to get to places much quick and easier than traditional bikes.
After analyzing the data provided, these are my 3 top recommendations to help convert casual riders into becoming members:
- Start an advertising campaign between the months of May and September to help get the word out about Cyclistic. This can be done with an online advertising campaign through social media, and with fliers/posters on the stations themselves.
Late Spring to Early Fall are the most popular times to rent a bicycle, so targeting that time frame will yield the greatest possible chance for people wanting to use Cyclistic’s services. Also advertising online and near the stations can bring awareness to Cyclistic’s services.
- Possible alongside the previously mentioned advertising campaign, start a program that gives a discount for becoming a member for at minimum 1 month to help get the casual users into Cyclistic’s ecosystem.
Making products and services cheaper is always more appealing to the average customer and makes it appealing for any customer to start using Cyclistic’s services.
- When buying/upgrading rideables, prioritize classic and electric bikes as those rideables are much more popular than electric scooters. Either buy/upgrade the rideables with the ratio of 2:2:1 or 3:3:1 for electric bikes, classic bikes, and electric scooters respectively. The ratio that is overall better cost wise should be choose, but data on the cost of these processes much be collected and analyzed to determine the best ratio.
Riders, both members and casual, much prefer bicycles compared to scooter, which should be taken into account when deciding to buy/upgrade rideables.