Five Email Split-Tests that Really Move the Needle
Email split-testing is an essential part of the campaign development process as you learn what to say to get your shoppers to open, click and even convert.
Learn how to set up effective split-tests in our blog post “Be Better than the Benchmark”
Day of Week
According to our 2016 retail email benchmark whitepaper, Saturday and Sunday have the lowest email volume, with 12% and 13% of total weekly volume being sent on those days, respectively.
Retail email volume has traditionally been lower on weekends for two reasons. First, in the past, the eCommerce and store channels were separate and email was used solely to drive online sales. These online shoppers were typically on their computers Monday through Friday while at work so emails were sent during those hours. And secondly, marketers sent the messages while they, too, were at work instead of on the weekend when they were at home.
This just isn’t the case anymore. Shoppers are always connected through their phones and tablets and it has been proven that emails contribute greatly to in-store sales. And with the ability to schedule a strategic series of messages, emails can be set in advance and deployed at the right time.
You should definitely try sending email on Saturday and Sunday, as the conversion rates stay about the same and there are fewer emails to compete with in the inbox:
We worked with many of our retailing clients to figure out the optimal day to send simply by split-testing messages on different days. In this example, we tested Saturday versus Sunday with the winner determined by the conversion rate. Sunday’s conversion rate was 10% higher than Saturday’s but it’s important to note that Sunday’s message also had a 27% higher ROI.
If you are currently only sending one or two emails per week, we’d recommend identifying the four days of the week that have the highest performance rates and sending those days. No time to add more messages to your workload? Try augmenting your deployment schedule with a Recurring Automated Campaign. Learn more here.
Time of Day
Just as important as what day to send is what time of day to send...especially if you’re already sending four or more days of the week. Optimizing messages to be delivered at the right time of the day will greatly increase sales and revenue.
We have performed many time of day tests with our clients in order to determine if their messages perform better in the morning or afternoon. And what we found may surprise you. Different types of messages work better at different times, stressing the fact that it is so important to test all of your messages.
In this example, the winner was determined by the amount of revenue per email and the message sent at 10:00 am brought in more than twice as much revenue as the one sent at 4:00 pm. However, the 4:00 pm message had an open rate that was 3.5% higher than the morning message and a 4.4% higher read rate than the earlier message, showing that the audience was more engaged in the afternoon.
Personalized Recommendations Based on Browse and Purchase Behavior
If you are still hand curating product recommendations or recommending the same products to every shopper – such as top sellers or new markdowns – you will see a big increase in sales and revenue by personalizing those product recommendations based on each shopper’s online behavior and past purchase history.
Personalized product recommendations have been proven to increase conversion rates more than 20% - proving that the shoppers are seeing merchandise they are interested in greatly increases revenue. In this example, personalized recommendations brought in twice as much revenue per email as the email that promoted static top selling merchandise.
Everyone loves to find a great deal, but you could be giving away too much – or not enough – of a good thing. Split-testing offers will absolutely have a positive impact on your bottom line as you will determine exactly how much of a discount you need to give in order to get customers to buy and you’ll stop giving away 20% off when 5% will do.
Split tests to try: a dollar amount off versus a percentage, a free shipping threshold, a free gift versus a dollar amount off if a maximum amount is spent – get creative! The more you test, especially before the holidays, the better your margins will be.
In the example below, a client tested offering free shipping at $50 or $125. As you can see, the $50 free shipping threshold had a conversion rate nearly two times higher than the $125 threshold. But other metrics, such as click rates, were much closer. This proves that shoppers were interested in the products but the higher threshold to receive free shipping could have gone into their decisions not to purchase.
We get a lot of questions about whether or not prices should be included in emails – especially for product recommendations. The answer: it depends. The goal of the email should be to entice shoppers to go to the site – you don’t want customers making purchasing decisions in the inbox.
Split-testing prices in emails is definitely a test worth running as it will help you make better business decisions that lead directly to more conversions and higher revenue.
In the example, you can see that adding the price in the email doubled the conversion rate. But you can see that the email without the prices had approximately 40% more clicks, proving that shoppers were interested in the products being promoted.
What other email aspects have you tested that led to increased sales? Let us know in the comments!