Testing is a top priority for email marketers. Without it, we’d just be relying on intuition and assumption. At CustomInk, where groups can design their own t-shirts for occasions like family reunions or school events, we face both an advantage and a challenge when it comes to testing our email marketing programs.
Here's our opportunity: a lot of people come to our site to design t-shirts, and as part of the process, they save their designs under an email address and can sign up to receive email reminders. On the one hand, this gives us the ability to reach prospective customers, but on the other, we have to work hard to earn these soon-to-be customers' trust without alienating them pre-purchase.
Needless to say, we've done a lot of testing to identify the most effective way to help these customers follow through and order their customized products. We have to identify why the customer put off placing an order. To get at the reason, our testing must go beyond measuring the effectiveness of a single email subject line and delve deeper into the content of the email itself.
Here's what we've learned:
Our biggest "Aha!" moment came when we sent prospective customers emails with the simple question: "How can we help?" We then gave them a series of options about what kind of information they needed. Did they want to know more about pricing or the tools available to make really creative designs? We had a pretty good idea of common questions based on discussions with our customer support team and other research. Once the emails were sent, we took note of the information prospective customers want more of and let that inform our messaging going forward.
We learned that at certain points in the experience, price was most important to these customers while at others, people wanted to know more about the product offering and tips on creating a great design. People are looking for different information at different times in the process and testing helps us reveal what this information is and when it's best to present it.
To send or not to send?
We are also very careful about not inundating our potential customers with emails. To figure out the right quantity and the right messaging, we run tests where we will stop sending a specific email for a few weeks. If we see that there are no changes to the number of orders completed, we assume that the holdout email wasn't effective. We're happy to remove it from our cycle of emails because we want each of our emails to be valuable. We're trying to enhance their experience by providing valuable information to them at the right time.
We typically schedule the emails far enough apart that we can tell whether a specific email worked or if another one earlier or later in the track did better. After about a week, we can tell which email led a customer to buy.
Timing is everything
The question about when is the right time to send an email is an important one. When a CustomInk visitor saves her design, she immediately gets an email with a link to that design so she can access it easily and quickly. We then wait a few days before sending our first follow-up email, and then we send an email every few days after that. We have tested various frequencies and learned that some customers are more responsive to frequent, intense emails while others preferred more spaced-out communication, so testing showed us just when to send emails to engage the most people.
For example, if a customer saves a design on a Monday and then gets an email about it on the following Monday (as opposed to the Sunday after she saved the item), is there a trigger there because it's the exact same day of the week?
At CustomInk, we've come to appreciate how important the testing of our email campaigns is to our business. But we understand too that it's an evolving process. As consumer preferences quickly change, what works today doesn't necessarily work tomorrow. We aim to be the first to know.
Allyson Ayers is Email Marketing Manager at CustomInk, an online retailer that enables customers to design and order custom apparel such as t-shirts and sweatshirts.