It’s been 3 months since we released Mailchimp’s Personalized Product Recommendations tool (affectionately dubbed P-Rex around MC HQ). In that time, 26.3 million people have received a total of 88.3 million recommended products.
Those campaigns have made a total of $1.4 million USD in revenue, and $170,000 of that revenue—more than 12%—came from people who bought or clicked on recommended items.
We love seeing small businesses succeed, so these numbers make us super happy. They don’t, however, tell the whole story. Do campaigns with personalized product recommendations actually perform better than those without? Let’s take a look.
For starters, we put together a group of users who had recently sent campaigns both with and without personalized product recommendations. We then compared campaigns with and without recommendations for each user, and quantified the average differences in revenue, click rates, and unsubscribe rates. Our analysis focused on regular campaigns, which are the most common type of campaign (as opposed to automation workflows or A/B split campaigns).
Here are a few things we learned:
Small businesses get an average of 47.8 cents for every person who opens a campaign without custom recommendations. Campaigns with recommendations got an average of 62.6 cents per open. That’s a 31% increase in revenue.
We also found that 14.4% of recipients clicked on campaigns without recommendations, and 14.7% of recipients clicked on campaigns with personalized product recommendations. That’s a 2% increase in the number of people who clicked.
Finally, 1.8% of openers unsubscribed from campaigns without recommendations, but only 1.6% of openers unsubscribed when a campaign contained recommendations. That’s an 8% drop in unsubscribes.
George Rodriguez of Rodrigo Cigars is a small-business owner who has a lot of experience with email marketing. In the past 3 months, he’s sent multiple successful campaigns with personalized product recommendations. We asked him for his thoughts on the process and results.
“I was using another third-party app to duct tape together recommendations using custom HTML, which was not easy to maintain nor was it responsive to mobile," George says. "Now, I get dynamic personalized product recommendations with a simple drag and drop into any email, and I can use the same great-looking responsive email templates my customers enjoy reading on any device.”
After using personalized product recommendations, George noted that his click-through rate “increased dramatically, and so did conversions.”
"I love how it gives new customers the best sellers and existing customers recommendations based on purchases," George elaborates. When asked if he had any tips for using the feature, he shared his simple-but-effective strategy: "Just dropping it into any communication with your customers will give your audience the urge to click through and shop."
Personalized product recommendations have been a huge success for our customers, and we’re going to continue our research to make the tool even better. In fact, we’ve recently made a few behind-the-scenes updates:
More focus on recent purchases: All purchases that a customer has made impact their recommendations, but recent purchases are now more important than older purchases.
Smarter top sellers: We used to use a year’s worth of data to calculate top sellers, but we recently ran a simulation and found that one month’s worth of data did a better job.
Other general improvements: Additional tweaks have been made to our process to ensure that you get higher quality recommendations in a more timely manner.
These small changes will improve the quality of our personalized product recommendations, and we’ll continue to research and refine the tool as we learn more in the months ahead. In the meantime, we’ve seen amazing results from personalized product recommendations in both automation and regular campaigns, and we look forward to seeing more.