How We Use Automated Emails to Increase Trial Conversions

A few months ago we changed our trial from unlimited to 14 days. This caused a problem with the welcome campaign emails. Unlike before, they needed to be in sync with where people were in their trial.

We recently fixed this using's triggered campaigns. Setting up an email campaign was a stab in the dark for us, so this is a quick guide to getting set up for newcomers. The campaigns haven't been running for long enough to tell how effective they are, although my instinct says some communication is better than none.

Why email

Email is still an important way we communicate with users. For new sign ups, it's a direct line to the users outside of our platform and a vital part of onboarding. For older users, it allows us to tell them about important new features.

Throughout our platform, email also lets us test things faster. While we were working on our new set up guide, we used emails to direct customers to features. We then tracked engagement and sign-ups to see which were key to upgrades. As emails have less design and engineering costs up front, we are able to test a lot faster than an A/B test on our site.

What the emails look like

Our main goals were to:

  1. Establish communication with new customers
  2. Lead them in the direction of features that will most likely lead them to upgrade
  3. Remind them that the trial expires
  • We also don't want to be annoying

Since writing an email marketing campaign was new to us, we did some research by subscribing to a bunch of other companies' trial emails. In the end, we decided to model ours after Squarespace's campaign. This meant four emails:

(1) Welcome email: A simple email that pretty much says "Hi! You made an account. We noticed. Hooray!" This email exists mostly to let people know we're a real company. It also gives new users a point of contact as they get started.

(2) How to use your trial: This email is delivered three days into the trial. It suggests using three features we've noticed most users engage with before upgrading.

(3) 3 days left in trial: Before this email, a lot of people would delete their project after the trial because they didn't remember to try it out. This gives people a warning and recommends a few easy ways to delve into ReadMe in a short time.

(4) 24 hours left: This is a final warning to users that they are running out of time. We hope it at least reminds people that they created a project. At best it gets them to upgrade. We also let people ask for more time by responding to any of these emails.

We use liquid tags in to personalize emails. Each customer has certain attributes in our database. When we call these data points they allow us to insert names, dates, and project information. We can also send them in links to direct customers to the exact page we want inside of a desired project.

How we set it up in

Each of these emails is sent to a segment of users created in Because the emails are designed to go with different portions of the trial, they sometimes need to be at different times for different customers (if we change the trial end date). This means the standard "wait __ days" between emails didn't work for us. Instead, we created trigger-based campaigns for each email.

The first email is triggered with an event called "Signed Up". For the subsequent emails, the status of a project is checked every day at 10am through a cron job that calculates where each user is in their trial. If they are 3 days in, have 3 days left, or have 24 hours left we send an event to with that information. They then are added to the segment for that event, and the email that goes with that segment is created.

In the Trigger-based emails, customers are sent an email the first time that they enter a segment. This means customers can only receive each email once. While our users can have multiple projects in each account, we only send them this email campaign once for now.


In the future we'll use this platform for more targeted drip-campaigns and test out's "newsletter" format to target emails to specific users without a trigger.

Because we weren't sending emails before, it's hard to compare this form of triggered email to past data. However, we have seen the rate of upgrades increase so far, and we'll continue to refine our emails to increase conversions.

Stay In-Touch

Want to hear from us about APIs, documentation, DX and what's new at ReadMe?