Customer Testimonial Analysis.Rmd

---
title: "Customer 1st Order Satisfaction Analysis - Aug,2022"
author: "Linh Vo"
date: " Data is collected from 2022-08-01 to 2022-08-31"
output:
  pdf_document: 
    latex_engine: xelatex
    fig_width: 8
    highlight: tango
---

```{r setup environment, include=FALSE}
install.packages("tidyverse")
install.packages("janitor")
install.packages("ggplot2")
install.packages("ggpubr")
install.packages("reactable")
install.packages("kableExtra")
library(knitr)
library(kableExtra)
library(reactable)
library(tidyverse)
library(readr)
library(janitor)
library(dplyr)
library(ggplot2)
library(ggpubr)
```

```{r import data, include=FALSE}
Survey <- read_csv("Survey.csv", na = "empty", show_col_types = FALSE)
head(Survey)
```

```{r cleaning data, include=FALSE}
Survey <- as.data.frame(Survey)
clean_names(Survey)
data <- Survey %>% 
  unite(col = 'cuisine_type', 
        c('Au','Nhat','Han','Viet','Khac'), sep ='', remove = TRUE)
```

## Overall summary  

There are 64 customers attended to the survey. 

```{r overall 1, echo=FALSE}
data %>% summarise(number_of_cust = n_distinct(SessionID))
```

**Customers' portrait:**

Most of our customers (49 out of 64) are serving Vietnamese Cuisine restaurants.  

Veggies seller is stand for "sạp bán rau" which comes from Wet Market channel.  

There are 5 customers who have greater than 1 store (Chain), the remaining is unique store.  

```{r customer portrait, echo=FALSE}
data %>% group_by(cuisine_type) %>% summarise(number_of_cust = n_distinct(SessionID))
data %>% group_by(`Store Type`) %>% summarise(number_of_cust = n_distinct(`Store ID`))
```

### 1. Approaching Channel vs Cuisine Type  

The most effective channel is direct approaching by Sales.  

Comparison between SNS (Zalo, Facebook), Customers seem to know our business more through Facebook Ads.  

```{r approaching channel, echo=FALSE}
data %>% ggplot(mapping = aes(x=`Approaching Channel`, fill = cuisine_type)) +
  geom_bar(stat="count") + coord_flip() + 
  scale_fill_brewer(palette="Spectral") +
  theme_classic() +
  theme(strip.text.x = element_text(size = 8, color = "darkblue", face = "bold"),
        axis.text.x = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.text.y = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        legend.position = "right", legend.title = element_text(size=8, color = "black", face = "bold"),
        legend.text = element_text(size=8, color = "grey20", face = "plain"), 
        legend.key.size = unit(.5, 'cm'),
        plot.title = element_text(size = 10, face = "bold", color = "darkred")) + 
  ggtitle("Number of customers per Approaching Channel")

```

### 2. Tools & Customer Satisfaction vs Cuisine Type  

Most of customers are satisfied with their 1st experience with our business.  

However, there are 13 customers who were not really happy (Neutral) or even felt disappointed.   (Dissatisfied). And most of them are selling Veggies and Vietnamese cuisine.   

```{r Tools, echo=FALSE}
data %>% ggplot(mapping = aes(x=`Tools`,y = cuisine_type, color = cuisine_type)) +
  geom_count(show.legend = FALSE) + coord_flip() + facet_grid(vars(`Customer satisfaction`), scales = 'free') +
  scale_fill_brewer(palette="Spectral") + theme_bw() +
  theme(strip.text.y = element_text(size = 8, color = "darkblue", face = "bold"),
        axis.text.x = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.text.y = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.title.x = element_text(size = 9, color = "black", face = "bold"),
        axis.title.y = element_text(size = 9, color = "black", face = "bold")) +
  labs(x = "Tools", y = "Cuisine Type")
```

### Let's take a look on the 6 measurement that CX used to verify customer satisfaction  

We will group 64 customers into 4 segments depend on their satisfaction above:

-   Dissatisfied
-   Neutral
-   Satisfied
-   Very satisfied

We hope to figure out why they felt unhappy with Kamereo service by going through 6 measurement:

-   Price
-   Product quality
-   Fulfillment rate
-   Product variety
-   Shipping service
-   Customer service

#### 2.1 Price  

As the graph shown, Price has slight correlation with the customer satisfaction.  

Because unhappy customers seem to appear mostly in Dissatisfied group.  

Otherwise, Very satisfied customers felt Very happy with our price.  

There are some customers in Satisfied group felt unhappy with the price, but maybe price is not their main concern when choosing a supplier.  

```{r price, echo=FALSE}
data %>% ggplot(mapping = aes(x=`Price`,y = cuisine_type, color = cuisine_type)) +
  geom_count() + coord_flip() + facet_wrap(vars(`Customer satisfaction`), scales = 'free') +
  scale_fill_brewer(palette="Spectral") + theme_bw() + 
  theme(strip.text.x = element_text(size = 8, color = "darkblue", face = "bold"),
        axis.text.x = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.text.y = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.title.x = element_blank(),
        axis.title.y = element_text(size = 9, color = "darkred", face = "bold"),
        legend.position = "right", legend.title = element_text(size=8, color = "black", face = "bold"),
        legend.text = element_text(size=8, color = "grey20", face = "plain"), 
        legend.key.size = unit(.5, 'cm')) 
```

*n is the count of customers related to each group*

#### 2.2 Product quality  

Customers obviously felt dissatisfied with our product quality.  

Unhappy statement appears in all 4 groups of customer.  

**--\> It probably our product quality were truly not good in the past month.** 

**--\> We need to improve our QC check more strictly to make guarantee to our product quality with customers.**

```{r prodcut quality, echo=FALSE}
data %>% ggplot(mapping = aes(x=`Product quality`,y = cuisine_type, color = cuisine_type)) +
  geom_count() + coord_flip() +facet_wrap(vars(`Customer satisfaction`), scales = 'free') +
  scale_fill_brewer(palette="Spectral") + theme_bw() + 
  theme(strip.text.x = element_text(size = 8, color = "darkblue", face = "bold"),
        axis.text.x = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.text.y = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.title.x = element_blank(),
        axis.title.y = element_text(size = 9, color = "darkred", face = "bold"),
        legend.position = "right", legend.title = element_text(size=8, color = "black", face = "bold"),
        legend.text = element_text(size=8, color = "grey20", face = "plain"), 
        legend.key.size = unit(.5, 'cm')) 
```

#### 2.3 Fulfillment rate  

Almost customers felt satisfied with our fulfillment.  

There is only one customer in Viet group felt unhappy but overall, they still felt satisfied with our service.  

```{r fulfillment rate, echo=FALSE}
data %>% ggplot(mapping = aes(x=`Fulfillment rate`,y = cuisine_type, color = cuisine_type)) +
  geom_count() + coord_flip() + facet_wrap(vars(`Customer satisfaction`), scales = 'free') +
  scale_fill_brewer(palette="Spectral") + theme_bw() + 
  theme(strip.text.x = element_text(size = 8, color = "darkblue", face = "bold"),
        axis.text.x = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.text.y = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.title.x = element_blank(),
        axis.title.y = element_text(size = 9, color = "darkred", face = "bold"),
        legend.position = "right", legend.title = element_text(size=8, color = "black", face = "bold"),
        legend.text = element_text(size=8, color = "grey20", face = "plain"), 
        legend.key.size = unit(.5, 'cm')) 
```

  
#### 2.4 Product variety  

Almost customers felt satisfied with our product variety.  

There are 2 customers who felt unhappy. Their feedback for that is they need more FMGC products (toilet paper/tissue) and want to buy trứng vịt lộn item.  

```{r product variety, echo=FALSE}
data %>% ggplot(mapping = aes(x=`Product variety`,y = cuisine_type, color = cuisine_type)) +
  geom_count() + coord_flip() + facet_wrap(vars(`Customer satisfaction`), scales = 'free') +
  scale_fill_brewer(palette="Spectral") + theme_bw() + 
  theme(strip.text.x = element_text(size = 8, color = "darkblue", face = "bold"),
        axis.text.x = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.text.y = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.title.x = element_blank(),
        axis.title.y = element_text(size = 9, color = "darkred", face = "bold"),
        legend.position = "right", legend.title = element_text(size=8, color = "black", face = "bold"),
        legend.text = element_text(size=8, color = "grey20", face = "plain"), 
        legend.key.size = unit(.5, 'cm')) 
```

#### 2.5 Customer service  

So great!  

All of our customers are happy with our Customer Service.  

```{r customer service, echo=FALSE}
data %>% ggplot(mapping = aes(x=`Customer service`,y = cuisine_type, color = cuisine_type)) +
  geom_count() + coord_flip() + facet_wrap(vars(`Customer satisfaction`), scales = 'free') +
  scale_fill_brewer(palette="Spectral") + theme_bw() + 
  theme(strip.text.x = element_text(size = 8, color = "darkblue", face = "bold"),
        axis.text.x = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.text.y = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.title.x = element_blank(),
        axis.title.y = element_text(size = 9, color = "darkred", face = "bold"),
        legend.position = "right", legend.title = element_text(size=8, color = "black", face = "bold"),
        legend.text = element_text(size=8, color = "grey20", face = "plain"), 
        legend.key.size = unit(.5, 'cm')) 
```

#### 2.6 Shipping service  

Two in a row!  

No negative feedback for our Shipping service.  

```{r shipping, echo=FALSE}
data %>% ggplot(mapping = aes(x=`Shipping service`,y = cuisine_type, color = cuisine_type)) +
  geom_count() + coord_flip() + facet_wrap(vars(`Customer satisfaction`), scales = 'free') +
  scale_fill_brewer(palette="Spectral") + theme_bw() + 
  theme(strip.text.x = element_text(size = 8, color = "darkblue", face = "bold"),
        axis.text.x = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.text.y = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.title.x = element_blank(),
        axis.title.y = element_text(size = 9, color = "darkred", face = "bold"),
        legend.position = "right", legend.title = element_text(size=8, color = "black", face = "bold"),
        legend.text = element_text(size=8, color = "grey20", face = "plain"), 
        legend.key.size = unit(.5, 'cm')) 
```

  
### With their positive and negative satisfaction, will they place second order? and even increase their volume?  

#### Second order?  

As the graph below shown, even customers felt satisfied, they still not place second orders.  

```{r second order, echo=FALSE}
data %>% ggplot(mapping = aes(x=`Will you order second time?`, fill = cuisine_type)) +
  geom_bar() + facet_grid(vars(`Customer satisfaction`), scales = 'free') +
  scale_fill_brewer(palette="Spectral") + theme_light() + 
  theme(strip.text.y = element_text(size = 8, color = "darkblue", face = "bold"),
        axis.text.x = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.text.y = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        legend.position = "right", legend.title = element_text(size=8, color = "black", face = "bold"),
        legend.text = element_text(size=8, color = "grey20", face = "plain"), 
        legend.key.size = unit(.5, 'cm')) + ggtitle('Will you order second time?') 
```
  
Let's review their comments to understand their point of view:  

Their concerns are all about price and product quality.  

```{r comments , echo=FALSE}
data_fb_no <- data %>% filter(`Will you order second time?` == 'No',
                              `Customer satisfaction` %in% c('Satisfied','Neutral')) %>%
  select(`Store ID`, `Store Name`,`Customer satisfaction`,`Will you order second time?`,`Comments` )
data_fb_no <- arrange(data_fb_no, `Customer satisfaction`)
knitr::kable(data_fb_no, format="markdown")
```
  
#### Increasing volume?  

There are 52 out of 64 customers who will place second order with us.  

Let's check if they will increase their volume in the chart below:  

```{r increasing volume , echo=FALSE}
data_fb_yes <- data %>% filter(`Will you order second time?` == 'Yes')
data_fb_yes %>% ggplot(mapping = aes(x=`Will you increase your volume?`, fill = cuisine_type)) +
  geom_bar()  +
  scale_fill_brewer(palette="Spectral") + theme_light() + 
  theme(strip.text.x = element_text(size = 8, color = "darkblue", face = "bold"),
        axis.text.x = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.text.y = element_text(color = "grey20",
                                   size = 6, hjust = .5, vjust = .5, face = "plain"),
        axis.title.x = element_blank(),
        axis.title.y = element_blank(),
        legend.position = "right", legend.title = element_text(size=8, color = "black", face = "bold"),
        legend.text = element_text(size=8, color = "grey20", face = "plain"), 
        legend.key.size = unit(.5, 'cm')) + ggtitle('Will you increase your volume?')
```

Oh, There are 19 customer who will not increase their volume.  

Let's review their comments:  

```{r feedback, layout="l-body-outset", echo = FALSE}
data_fb_yes_no <- data_fb_yes %>% filter(`Will you increase your volume?` == 'No') %>%
  select(`Store ID`, `Store Name`,`Customer satisfaction`,`Will you increase your volume?`,`Comments` )
knitr::kable(data_fb_yes_no, align ='lllll') %>%
  kable_styling(full_width =T, font_size = 9)
```


  
And again, the most common reasons are price and product quality.  
  
    
### Conclusion  

In a nutshell, most of our customers were happy with their first experience at Kamereo.  

However, to improve and increase our user retention, we must take more concentration on our price and product quality.  

**Next Action:**  

For Category Management team, as our CGO stated in Sales Sep strategy, we will increase personalized offer price via Sales rather than applying mass promotion.
For Product Quality issue, we will revamp and enhance the quality control with our suppliers and storage condition.