While the most crucial parts of the new user experience occur in the product, email is still a valuable channel for onboarding. Emails help welcome new users, get them back into the product during periods of low engagement, and convert them into customers.
A well-orchestrated email campaign is especially important for products that want to achieve high free trial-to-paid conversion rates and maintain a lean sales organization.
This blog post will dive into how one such organization, CloudApp, crafted a data-driven and personal onboarding email series that hit all the right notes and made a difference on the bottom line.
When Scott Smith, CloudApp’s VP of Sales and all-around good guy, first turned his focus to his user onboarding emails, he had no real direction to follow. Considering that his version-one onboarding emails basically all just said, “here are three things to do, let me know if you want a demo,” his emails have come a long way.
CloudApp’s onboarding emails now have layers of data manipulation in order to prioritize and personalize each outreach. This helps them automate part of the sales process for users that only need a self-service product, which reduces overhead while scaling their small team of two people (they act more like a team of 10).
Scott didn’t build it all overnight. He talked to marketer friends, read blog posts, and listened to users who responded to his onboarding emails to improve each one. Through the process, Scott added layers of complexity, from machine learning to Clearbit data, all over the course of 13 months. But far from feeling robotic, the emails are so well personalized that they continue to get warm responses. “We’re seeing growth in our email replies from people who feel the outreach is real, not automated,” says Scott. All this contributes to faster onboarding with fewer complications for the sales team.
Getting users past the first onboarding milestone
The first and only thing Scott and the CloudApp team care about for new users is that they download their desktop app. That’s the basis of their product, and without the desktop app, people can’t truly use the product.
So they use a series of emails—five, to be exact—to push people to install. CloudApp’s app data lives in Segment, which tells Customer.io which users have not yet installed the desktop version. Customer.io then triggers the email series, stopping it when a user installs.
Here’s an example of an early message in the series. The first call to action is “Download the app”:
Evaluating the likelihood of buying
After a user installs the CloudApp desktop app, CloudApp helps them discover features and use the product. The more active they are, the more likely they are to upgrade to the paid product.
So the overarching onboarding flow passes through three phases:
Get users to install the desktop app.
Help them use the tool and discover free features.
Talk to them about the features they’d get with the paid version.
CloudApp has a sales team that can talk to leads about whether upgrading is right for them. But to keep the team lean and effective, they should really be talking only to qualified leads, such as people who could get their team at work into a CloudApp team plan. Scott set up a way to qualify users and route them into different onboarding experiences with different amounts of sales and support attention.
MadKudu, a machine learning tool, scores CloudApp’s new users automatically. MadKudu first evaluates CloudApp’s successful customers and then looks for look-a-likes among new users, based on in-app behavior and Clearbit firmographic data like job title, company industry, and company size. Then they score each new user and put them into tiers that indicate how likely they are to become successful customers.
This score is pushed from MadKudu back into Segment, where Scott assigns users to segments that will receive different targeted emails.
Users who qualify as “likely to buy” will get emails from a CloudApp account manager offering demos. This makes sure the sales team doesn’t get unqualified appointments on their calendar and helps them stay lean. Here’s an example of a short, warm introductory email in the onboarding series from a sales teammate, Kat:
Users who don’t qualify get similar emails, but from the customer support team and without the offer of a demo. These users are often people who signed up with a Gmail or a iCloud account, indicating that they’re individual and not business users, or without associated Clearbit profiles. The emails are structured templates that immediately help a user self-serve and get started on their own, rather than striking up a conversation like the sales emails do.
Over the next few days, the user will continue to receive tips in the series.
Give users dynamic help
Sometimes, users got “stuck” during onboarding in the sense that they stopped short of trying certain CloudApp features. Scott implemented emails that are triggered in Customer.io based on what events a user has done and hasn’t done, and helps them try a specific feature or reach a certain milestone, moving them along the funnel.
Here’s an example of an email that pushes users onward from their first CloudApp “drop” to the next step: annotating an image.
In the next example, the email shows the user how to upload files to CloudApp for sharing.
These emails were slightly different, depending on whether the user was a self-serve account or a sales-qualified account. For example, this outreach email references the user’s potential difficulty in adopting a new tool into their workflow:
Other emails would ask if sales-qualified leads wanted to sign up for a team plan. This example is automatically personalized for the recipient, referencing the “3 people” at their company who are already using CloudApp:
These emails were an effective way to automatically help people get the most out of CloudApp without sending their whole product documentation to everyone at once.
Win back expired trial users
All of CloudApp’s paid plans offered a 14-day trial window, and CloudApp encouraged trial users to upgrade to a full-time, paid plan before the two weeks were up. But looking at MadKudu data, Scott saw that 50% of CloudApp’s paid users had converted after their trial had ended. He realized then that it was just as important and fruitful to communicate with users about upgrading after they had technically “churned” from the trial. He immediately set up a series of three automated emails with a casual tone to win them back.
These were triggered based on the “trialExpired” event in Customer.io. Here’s an example of one of the later notes in the series:
“Now, when I look at the number of responses per day that Kat receives, it’s substantially higher because of these win-back emails,” says Scott. “We now see pre-expiration and post-expiration as part of the same pipeline.”