While reporting a recent story, I wanted to send customized versions of the same email to dozens of potential sources. It was a classic use case for mail merge and, since I was already working with a spreadsheet that included prospective sources’ names and emails in R, I wondered: Could I run a mail merge without ever leaving RStudio?

It turns out the answer to this question is a resounding “yes” – and here’s how I did it.

1: Configure Gmail.

This was the trickiest part. I started by setting up two-factor authentication and then created an app password. I’ll use the password in the final step to send emails from R.

Next, to avoid including my app password in published scripts, I stashed it in my .Renviron file, saving it as the environmental variable GMAIL_APP_PW. If you get tripped up looking for your .Renviron, you can always pull it with usethis::edit_r_environ().

2: Format emails.

From there, all I had to do was read in a csv file with names and emails, create an email template and use string substitution to customize each message.

# Libraries
library(tidyverse)
library(curl)
library(here)

# Parameters
file_in <- here("updates/mail-merge/example.csv")
pw <- Sys.getenv("GMAIL_APP_PW") 
sender <- "adipierro@edsource.org"

# Code

## Read in file of recipients
recipients <- read_csv(file_in)

## Write an email template
template <- "
From: 'Amy DiPierro' <adipierro@edsource.org>
To: '{first_name} {last_name}' <{email}>
Subject: Hello, {first_name}!

Dear {first_name},

How's it going at {employer}?

Cheers,
Amy
"

# Use string formatting to customize each email.
message <- recipients %>% str_glue_data(template)

# Extract list of recipients' emails
emails <- recipients %>% pull(email)

Here’s the original data I used to send test emails to myself:

first_name	last_name	email	employer
Amy	DiPierro	adipierro@edsource.org	EdSource
Mya	DePriori	adipierro@edsource.org	CodeUser

And here’s an example of how the email template looks after string formatting:

From: 'Amy DiPierro' <adipierro@edsource.org>
To: 'Mya DePriori' <adipierro@edsource.org>
Subject: Hello, Mya!

Dear Mya,

How's it going at CodeUser?

Cheers,
Amy

3: Send emails!

Finally, I used the send_mail function from R’s curl library to email each customized message to the right recipient. In the code below, I wrap send_mail in the purrr function map2, which loops through the list of customized messages and emails simultaneously.

# Send all emails
map2(
  .x = message,
  .y = emails,
  .f = ~ send_mail(
    mail_from = sender,
    mail_rcpt = .y,
    message = .x,
    smtp_server = "smtps://smtp.gmail.com",
    username = sender,
    password  = pw,
    verbose = FALSE
  )
)

And that’s it. Within a few moments, my test emails hit my inbox.

Found a bug? Got a comment, criticism or suggestion for improvement? Drop me a note: adipierro.edsource.org.

--- title: "Code: A neat mail merge trick in R" date: "2025-07-26" categories: [code] --- While reporting a recent story, I wanted to send customized versions of the same email to dozens of potential sources. It was a classic use case for [mail merge](https://support.google.com/mail/answer/12921167?hl=en) and, since I was already working with a spreadsheet that included prospective sources' names and emails in R, I wondered: Could I run a mail merge without ever leaving RStudio? It turns out the answer to this question is a resounding "yes" -- and here's how I did it. # 1: Configure Gmail. This was the trickiest part. I started by setting up [two-factor authentication](https://support.google.com/accounts/answer/185839?hl=en&co=GENIE.Platform%3DDesktop) and then [created an app password](https://support.google.com/mail/answer/185833?hl=en). I'll use the password in the final step to send emails from R. Next, to avoid including my app password in published scripts, I stashed it in my .Renviron file, saving it as the environmental variable `GMAIL_APP_PW`. If you get tripped up looking for your .Renviron, you can always pull it with `usethis::edit_r_environ()`. # 2: Format emails. From there, all I had to do was read in a csv file with names and emails, create an email template and use string substitution to customize each message. ```{r} #| message: false #| warning: false # Libraries library(tidyverse) library(curl) library(here) # Parameters file_in <- here("updates/mail-merge/example.csv") pw <- Sys.getenv("GMAIL_APP_PW") sender <- "adipierro@edsource.org" # Code ## Read in file of recipients recipients <- read_csv(file_in) ## Write an email template template <- " From: 'Amy DiPierro' <adipierro@edsource.org> To: '{first_name} {last_name}' <{email}> Subject: Hello, {first_name}! Dear {first_name}, How's it going at {employer}? Cheers, Amy " # Use string formatting to customize each email. message <- recipients %>% str_glue_data(template) # Extract list of recipients' emails emails <- recipients %>% pull(email) ``` Here's the original data I used to send test emails to myself: ```{r} #| message: false #| warning: false #| echo: false recipients %>% knitr::kable() ``` And here's an example of how the email template looks after string formatting: ```{r} #| message: false #| warning: false #| echo: false message[2] ``` # 3: Send emails! Finally, I used the `send_mail` function from R's `curl` library to email each customized message to the right recipient. In the code below, I wrap `send_mail` in the `purrr` function `map2`, which loops through the list of customized messages and emails simultaneously. ```{r} #| message: false #| warning: false #| results: false # Send all emails map2( .x = message, .y = emails, .f = ~ send_mail( mail_from = sender, mail_rcpt = .y, message = .x, smtp_server = "smtps://smtp.gmail.com", username = sender, password = pw, verbose = FALSE ) ) ``` And that's it. Within a few moments, my test emails hit my inbox. Found a bug? Got a comment, criticism or suggestion for improvement? Drop me a note: `adipierro.edsource.org`.