Don’t Confuse Common Practice with Best Practice
There is a massive need to share results and best practices, since this is still an area which is not widely understood. As such, I welcome all research into shopping cart abandonment and recovery emails.
So it was with some eager anticipation that I read Epsilon’s Abandoned shopping cart email strategies report. Unfortunately, when analyzing their results for the best time to send shopping cart recovery emails, Epsilon confused common practice with best practice:
“Typically 48 to 72 hours is optimal for abandoned shopping cart emails; anything sooner may seem too intrusive. Follow-up emails beyond the three-day mark may be too late to convert a potential buyer.”
This is pure speculation – in fact Epsilon’s own research doesn’t support this. What the research shows is that there is a distribution on send times for abandoned shopping cart emails ranging from the same day to more than a week. Most retailers who send abandoned shopping cart emails do so between 48 and 72 hours.
While this is the most common send time, it doesn’t make it the best practice. Best practice is a term often mis-used, so here is an example of what I mean:
“Just because lemmings are jumping off a cliff, it doesn’t mean that we should all do the same.”
Let me clearly state that sending a real time shopping cart abandonment email is ‘best practice.’
In this case, best practice is based on multiple independent studies including one from Forrester, two separate MIT studies and a Marketing Experiments test. Each of these studies have concluded that real time responses work better.
SeeWhy, using our A/B testing module, has shown on multiple tests for customers that the same email will perform significantly better when sent in real time. Open rates, click-throughs and conversion rates are all higher when shopping cart abandonment emails are sent in real time. Using data from more than 60,000 recovered shopping carts, we know that if abandoners are going to come back and buy after remarketing two thirds will do so in the first few hours following abandonment.
So here’s my simple summary of why real time shopping cart recovery emails work better:
1. Recover more abandoned shopping carts
The guys at MIT found that 90% of ecommerce leads go cold in just one hour, so do not be surprised to recover significantly more revenue by emailing immediately following an abandonment. We all know that leads go cold quickly, so it’s pretty obvious that if you follow up 24 hours later or more (as more than two thirds do in the Epsilon study) this isn’t going to work as well as a quicker follow up.
It works better because the potential purchase is still ‘front of mind.’ For most consumers, shopping for many categories of goods, especially discretionary purchases, is not a mechanical buying process. There’s emotion involved, and emotions are very powerful forces in overcoming objections and driving conversions. But the emotion drains quickly, so timing is critical in stoking the emotion.
2. Stay in step
The single biggest reason why more ecommerce companies don’t do shopping cart recovery emails is because if the challenges of staying in step with the customer. If the customer comes back and buys on your website before your recovery email has gone out, then your campaign is now out of step. You are just about to send a recovery email to someone who has just purchased. If you include a promotion, then you have now increased the stakes – customers will complain if they have purchased at the full price, but are then subsequently offered a discount. The single biggest cause of getting out of step is because of batch based data transfer between your web analytics or ecommerce platform and your email engine.
Staying in step is one of the key reasons that most batch based shopping cart recovery email campaigns are only a single step and do not contain promotions. Sending additional emails after the first increases your risk of getting out of step, and a promotion can cause customer complaints.
Yet research shows that the two biggest causes of shopping cart abandonment are (1) price and (2) timing. Batch based campaigns struggle to address the two biggest causes of shopping cart abandonment: the price issue (addressed with a promotion) or the timing issue (addressed with a sequence of emails over time).
Real time emails have a second hidden benefit: in order to be able to send real time emails, you need to be monitoring visitor activity on your website in real time. This real time monitoring enables your recovery campaign to stay in step with customers, and therefore enables you to run both promotions and multi-step campaigns.
3. Better service – Brand support
A website that loads quickly, conveys professionalism. Putting a phone number on the site, reassures the customer. Answering the phone promptly, communicates good service. Following up on an aborted purchase days later….what message does this send about your brand, and your service?
An immediate service based message demonstrates all three: professionalism, reassurance, and good service.
The key to making this work is the tone and the approach of the email. If this is done well, then visitors like and appreciate the emails. Every real time remarketing email opened, is opened on average 2.2 times, illustrating that visitors are keeping the emails as an aide memoire, and using the direct link back to the cart for convenience. Unsubscribe rates for real time emails are also lower – no doubt because they are more relevant.
When the customer is still considering purchasing the abandoned item from you, and weighing it up against other purchases from elsewhere, do you want to follow up immediately, reminding them about your site, the item they almost purchased and all the values that your brand stands for? Or do you want to be the other guy, who follows up several days later, sends a subliminal ‘poor service’ message and ultimately misses the sale, nine times out of ten?
So I hope you understand that this wasn’t an idle claim that sending shopping cart recovery emails in real time is best practice. It simply works better. We have yet to find a situation where this isn’t the case, including high value sites (B-to-B and B-to-C) but if you think you have one, then we’d be happy to do an A/B test on your site.
