Employee survey from an auto insurance agency.

Data Source: Job Quitting

Tidying

quitting |> 
  ggplot(aes(y=reason))+
  geom_bar()

quitting |> 
  tabyl(reason) |> 
  arrange(desc(n))

Categories showing less than 10% can be grouped together for simplicity and focus.

quitting2=quitting |>
  mutate(reason_grouped=fct_lump_lowfreq(reason,other_level = 'Other')) |> 
  mutate(reason_grouped=fct_infreq(reason_grouped))
  
quitting2 |> 
  ggplot(aes(reason_grouped,fill=reason_grouped))+
  geom_bar(show.legend = F)+
  scale_fill_brewer(palette = 'Dark2')+
  theme_bw()+
  labs(title='Why People Quit by Count',x='Reasons',y='Count')+
  theme(plot.title = element_text(hjust=.5))+
  expand_limits(y=300)

quitting2 |>
  tabyl(reason_grouped) |> 
  arrange(desc(n)) |> 
  adorn_pct_formatting(digits = 1,affix_sign = T) |> 
  rename(Reason = 1, Count = 2, Percentage = 3) |> 
  gt() |> 
  tab_style(style=list(cell_text(weight = "bold")),locations=cells_column_labels()) |> 
  tab_style(list(cell_fill(color = "#F2F2F2")),locations =cells_body(columns=vars(Reason, Count, Percentage))) |> 
  tab_options(column_labels.font.size =px(14),table.font.size = px(14),table.width = "auto")
Reason Count Percentage
No Advancement 259 23.2%
Dislike Boss 244 21.8%
Salary 230 20.6%
Low Recognition 185 16.5%
Bored 122 10.9%
Other 78 7.0%

Recommendation

If boredom and low recognition gets addressed, then employee retention could increase roughly by 28%. To address boredom, plan and have an office scavenger hunt for a prize, like an Amazon gift card or one free day off from work. To address low recognition, upper management will have to get involved. For example, implement a nomination system for awarding good work ethic and when someone’s name appears enough in this system, then the manager or supervisor (even better if it’s the CEO or someone from the corporate exec team) personally finds the nominee to take a selfie picture with and posts it on the agency’s social media platform.

As for the factors of no advancement, dislike boss, and salary, those need to be addressed by upper management. They will have to make existential changes as to how the agency is led. Out of those three, dislike boss is the most harmful to the agency. One study claims that 75% of American employees’ most stressful aspect of their job is their immediate boss.

Further Research

A 2022 report by McKinsey & Co.:

The comprehensive report can be found HERE.

sessionInfo()
## R version 4.3.1 (2023-06-16 ucrt)
## Platform: x86_64-w64-mingw32/x64 (64-bit)
## Running under: Windows 10 x64 (build 19045)
## 
## Matrix products: default
## 
## 
## locale:
## [1] LC_COLLATE=English_United States.utf8 
## [2] LC_CTYPE=English_United States.utf8   
## [3] LC_MONETARY=English_United States.utf8
## [4] LC_NUMERIC=C                          
## [5] LC_TIME=English_United States.utf8    
## 
## time zone: America/New_York
## tzcode source: internal
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
##  [1] gtsummary_1.7.2    gt_0.9.0           performance_0.10.3 readxl_1.4.2      
##  [5] janitor_2.2.0      lubridate_1.9.2    forcats_1.0.0      stringr_1.5.0     
##  [9] dplyr_1.1.1        purrr_1.0.1        readr_2.1.4        tidyr_1.3.0       
## [13] tibble_3.2.1       tidyverse_2.0.0    report_0.5.7       ggplot2_3.4.1     
## 
## loaded via a namespace (and not attached):
##  [1] sass_0.4.5           utf8_1.2.3           generics_0.1.3      
##  [4] xml2_1.3.4           stringi_1.7.12       hms_1.1.3           
##  [7] digest_0.6.31        magrittr_2.0.3       RColorBrewer_1.1-3  
## [10] evaluate_0.20        grid_4.3.1           timechange_0.2.0    
## [13] fastmap_1.1.1        jsonlite_1.8.4       cellranger_1.1.0    
## [16] broom.helpers_1.13.0 fansi_1.0.4          scales_1.2.1        
## [19] jquerylib_0.1.4      cli_3.6.1            rlang_1.1.0         
## [22] munsell_0.5.0        cachem_1.0.7         yaml_2.3.7          
## [25] withr_2.5.0          tools_4.3.1          tzdb_0.3.0          
## [28] colorspace_2.1-0     vctrs_0.6.1          R6_2.5.1            
## [31] lifecycle_1.0.3      snakecase_0.11.0     insight_0.19.2      
## [34] pkgconfig_2.0.3      bslib_0.4.2          pillar_1.9.0        
## [37] gtable_0.3.3         glue_1.6.2           highr_0.10          
## [40] xfun_0.38            tidyselect_1.2.0     rstudioapi_0.14     
## [43] knitr_1.42           farver_2.1.1         htmltools_0.5.5     
## [46] labeling_0.4.2       rmarkdown_2.23       compiler_4.3.1