Can automation and AI help create better emails?
- Time to rethink the humble email...
Forty-one percent of employees think that email is critical to the overall success of a company, according to the Litmus 2021 State of Email Report. But email is not the same today as it was a few years ago, and it's time to put aside the basics and start thinking about how to best leverage email across the entire customer lifecycle.
The emails companies send
In the Litus report, the two most popular emails are the welcome email and the newsletter. The welcome email welcomes a person to the start of their relationship with the brand, and the newsletter works hard to keep subscribers coming back again and again.
But there are plenty of other types of email, including product emails and sales/promotion emails - these round out the top four types cited as most created.
All these email types need to be personalized or better targeted to the person on the receiving end. Yes, we're talking about personalization again because generic emails will not cut it. But we're talking about a lot more than personalizing the greeting: "Hi <<insert name here>>, How are you enjoying your work at <<insert company name here>>."
By the way, I've seen Robert Rose refer to that as being "familiar," not being "personal." And a Validity webinar called it "tokenization." They are both right.
The report lists the top personalization factors as:
- Profile data
- Customer segments
- Past email Interaction
In the Validity webinar, Kate Adams, SVP of Marketing at Validity, offered better ways to personalize email, which is somewhat easier for B2C. For example, if someone was shopping on a site and put something in their cart but didn't check out, or they spent a chunk of time looking at a particular product, an automated email can be sent that includes this information. Or, if they did purchase, they get a personalized email showing the product and offering related items.
How can you do this on the B2B side? If a person has downloaded a gated asset, you know who they are and customize your emails to offer related information. And if they also spend time on your website looking at pages, you can build an automated workflow that delivers personalized emails based on this information.
The key is you personalized the email according to what you know about the person receiving it. And that does require you to know something about them. But can we take it even further?
Optimize with the help of AI
Testing subject lines often happens, but it's usually a marketer creating the A and B options and running the test. Yes, you can run your subject line through a tool like CoSchedule's headline analyzer, and you can get a score that shows how well it uses power words and emotion words, if it's the right length, the right reading level, the right sentiment, and so on. But that's not personalization – that's optimization. To go further, you need different tools.
One example is to integrate a tool like Phrasee into your email platform. Phrasee is an AI-powered copywriting platform that can help you determine the best content for your emails (and other copy):
When you enter a brief, Phrasee generates thousands of potential copy ideas - whether it's a headline, a subject line, or something else. It then takes all those copy variants and puts them through their Deep Learning Engine to rank the best ones in order of predicted performance. It then shows a sample from that ranking for the user to choose from.
Phrasee and MessageGears, a B2C messaging platform, recently announced their integration. You have to be a customer of both platforms, but it allows marketers using MessageGears to pull in tested messaging from the Phrasee platform without moving between the two. Using Phrasee's AI, email subject lines can be better targeted or personalized.
And that's only one example (Albert is still around as well). Many AI tools can help you create more personalized email experiences, some integrated or native to your email platform, others you'll need to use separately.
Ask your favorite AI to help
When we look at AI tools that help improve targeting or personalization, the list starts to get a little long (there are many up and coming).
A couple I have looked at:
- Jasper.ai - formerly called Jarvis, I looked at this tool for a client, and it had so many features that it was almost overwhelming.
- Rasa.io - I came across this one from a blog on the Marketing AI Institute. Rasa helps you send personalized newsletters. Not segmented, but personalized to an individual.
- Persado - I've written about Persado; it's an enterprise-level platform that can help you optimize your copy for all types of short-form content.
The thing about most of these tools is that they work on a campaign level. They analyze emails and content and subjects lines from a segment or group of people and offer back the best messaging for the group.
The big question about tools like Phrasee, Persado, and Albert is, do they work for individual emails? Or for only segments? Can you create tests in these tools that test an individual's preference?
For example, if I have received 20 emails from a brand, can they tell which subject lines appealed most to me and then provide the next email in line with what I like? Can they examine the text of the emails sent and tell what I read or clicked on, compare that to what I have done on their website, what products I may already own, what support issues I have had, and then offer additional ideas for new emails - specific to me? And then do that for the thousands of other people on the brand's email list?
AI works because it analyzes lots of data. But one person doesn't produce that much data to analyze. Maybe it's enough to figure out the next best step for one person, but that's not what most AI tools do.
Brands do better at email when they think strategically about email. Privacy is one thing they need to pull into that strategy early. But equally important - or maybe even more critical - is designing an email program that thinks about the person on the receiving end as an actual person.
Smaller companies will do better at this because they have smaller sets of contacts to work with. Rasa.io seems to do this with its personalized newsletters. But we need to see it with all marketing emails. That's where we need to focus our email strategies and technology.