Advertisement
  1. SEJ
  2.  ⋅ 
  3. Content

How AI & Dynamic Content Are Reinventing Email Marketing

AI and dynamic content are helping email marketing become even more effective. And the possibilities for personalization go far beyond subject lines.

How AI & Dynamic Content Are Reinventing Email Marketing

AI is undoubtedly one of the biggest marketing buzzwords today and, combined with dynamic content, it’s helping one longtime brand asset – email – become even more effective.

Just like digital out-of-home is helping billboards remain relevant and brands like Ikea are using AR to reinvent the humble catalog, marketers can now tap into a myriad of data sources to cater messages to subscribers at the precise time and location of open.

That’s in part because the majority of consumers use smartphones as their primary device for checking email. Adobe’s 2016 Real-Time Marketing Insights Study also found 77 percent of respondents were planning to use real-time technologies that can help improve personalization efforts with more relevant data. (The 2017 study found smartphones remain the most commonly used device to check email.)

Email Marketing Remains a Powerful Brand Asset

According to email marketing firm Adestra, for example, 72.5 percent of consumers pick email as their preferred communication channel with businesses.

What’s more, a blog post from email experience platform Movable Ink said email is successful with consumers of all ages.

And, according to data from research-based marketing firm Ascend2, more than half of marketers rank email as an effective marketing channel – behind only brand websites.

This is surely, at least in part, because email has a median ROI of 122 percent, which is four times higher than other channels, according to email marketing platform GetResponse. The GetResponse study also noted email makes it easy for subscribers to share marketing messages and for brands to track engagement and conversions, as well as to test and optimize campaigns.

And yet, according to email marketing company MailChimp, the average open rate for emails in nearly 50 industries is 21.8 percent, and respondents to Adobe’s 2016 study found less than 25 percent of brand offers via email are interesting enough to open.

And that’s why dynamic email, in particular, offers promise – it takes personalization one step further by not only targeting specific messages to specific consumers but by also waiting until the precise moment of open to pull in other relevant data points to determine what content is served.

Research from email and cross-channel marketing firm Yes Lifecycle Marketing found personalized subject lines generate a 50 percent higher open rate, but only 2 percent of email subject lines employ personalization. And only 1.2 percent of all emails sent in Q2 2017 were personalized based on other factors, such as loyalty program status, browse behavior or purchase history.

Moving Beyond Personalization

But the possibilities for email personalization go far beyond subject lines.

“When an email is being sent to a consumer, what is being sent is the HTML codes that describe the text of the email and the layout of what it looks like, but…when it gets to [the email client], only then [does Apple Mail or Gmail] ask the server what image to show to the consumer,” said Liad Agmon, CEO of personalization technology stack Dynamic Yield. “If companies like us or Movable Ink can intercept that request, we know it came from you because we know your email address and location and every contextual data [point] that we need in real time to generate an image or set of images that are being sent back to the email program. That’s how the technology works. This is what enables us to do dynamic email.”

And that, for example, means LaFourchette, which Bridgette Darling, product marketing manager of personalization management tool Adobe Campaign, described as sort of like the French version of restaurant reservation tool OpenTable, can use the time of open to determine which restaurants still have availability and present those options to recipients while they can still get a table.

However, these dynamic email players are able to tap into much more than just time of open.

Dynamic Yield, for example, collects information about how consumers arrived on the site, what products they looked at and what they bought, which enables marketers to create audiences. These audiences can be as simple as “users in New York” or they can incorporate more complex behavioral data such as “viewed three product pages but did not complete a purchase.”

And, according to a rep, when an email recipient is identified within Dynamic Yield, the platform can determine if that user belongs to an audience in order to receive a dynamic experience.

“Have you looked at skirts, did you make any purchases [at Urban Outfitters] using an Urban Outfitters credit card? These signals are important,” Agmon added. “Are you reading the sports section or the fashion section of the New York Times? It’s a combination of every activity you do online…[and] we and companies like us [determine] out of the options, what is the best [content] to serve based on the signals we’ve collected over time.”

So if, say, Urban Outfitters has a sample sale in New York, the retailer can include the most relevant products based on a user’s purchase history – and update those products in real time so that the consumer will see products that are still in stock if he or she opens the email during the sample sale, Agmon said.

Conditions like weather can also be used.

“So the value there for the marketing team that [is] sending [the email is they] don’t have to guess where you’re going to be and craft such an email in advance,” Agmon said.

Vivek Sharma, CEO of Movable Ink, said a customer like travel company Orbitz uses dynamic email to pinpoint consumers’ starting cities and where they are likely to go and present the best available flights.

“Prices change quickly. Inventory passes quickly. And [it’s a subpar experience] if you’re finding that [a particular flight] is no longer available to you,” he added.

And, per Sharma, United Airlines sends a triggered email before a flight that provides a live seating chart, which allows travelers to see what Economy Plus seats are still available and to pre-purchase WiFi.

Agmon said dynamic email providers also know whether a recipient is currently traveling.

“Let’s say Urban Outfitters has a great sale in New York,” Agmon said. “If I know you’re traveling, it doesn’t make sense to send [the email] now. It’s clutter and won’t help to do a transaction, so understanding whether to send a specific email and when is the first part in really creating a personalized email program.”

Other applications include sports scores, which can be updated in real time, as well as breaking news updates – like, say, if the New York Times sends a news alert at 11 a.m., but the subscriber doesn’t open the message until 4 p.m., the recipient can still see the most updated version of the story, Agmon said.

Darling said package delivery status is another example as it allows consumers to see where their orders are without having to click through to the USPS, UPS, or FedEx sites.

Per Agmon, dynamic email also allows marketers to do better A/B testing. In other words, instead of the marketing team making educated guesses about the right images and creative, an algorithm can predict which creative a given user should see based on every data point that is known about that user.

“Machine learning is helping marketers in the selection of creative and the value for you as a consumer is trying to predict your taste and what is more relevant to you and getting the optimal experience in a way that couldn’t be done before,” Agmon added.

Per James Armstrong, marketing manager for KVH Media Group, using open-time personalization and geo-targeting, clients have seen click-through rates go up about 10 percent.

In a case study, travel company Travelocity, too, said dynamic email yielded open rates 40 percent higher than the typical marketing email and click-through rates that were 15 percent higher.

The Email Subject Line Challenge

One challenge with dynamic email, however, is that subject lines have to be decided at the time of send.

Marketers can still A/B test different versions or tap into natural language processing and machine learning to predict the best options, but once the decision has been made to offer a specific subject line to a specific user, it cannot be changed after the email is sent.

So if a consumer isn’t intrigued enough by the subject line to open the email, he or she won’t see the dynamic content at all.

“You can do personalization in email subject lines, but you have to do it while sending the email and not when the customer opens it,” Agmon said.

Darling concurred there is still no way to update subject lines.

“So when you’re creating a subject line, you have to make sure whatever image you’re putting forth to the customer…will still resonate,” she said.

However, Darling noted Adobe Campaign has a predictive subject line tool that allows marketers to use an algorithm to analyze subject lines and historical data to find the most appropriate words.

Companies like language optimization software firm Phrasee are also using algorithms to generate subject lines, as well as body copy and calls to action, based on “hundreds of emotions, sentiments, and phrases [to predict] what your audience will respond to” and, in turn, drive more opens, clicks and conversions.

In fact, Phrasee says it writes better subject lines and body copy than humans by using machine learning to understand what works best for a given brand and developing language algorithms tailored to each client.

Similarly, AI-generated marketing language company Persado has a Pro Email product that it says uses AI to generate better-performing email subject lines through emotional language and that cater to brand voice to generate high engagement.

Future Email Marketing Advances

As AI advances, it’s likely we’ll see innovation in other areas as well.

In fact, at Yext’s recent Onward17 event, Paul Roetzer, founder of the content hub Marketing Artificial Intelligence Institute and CEO of PR 20/20, said he suspects AI will also eventually enable marketers to determine the best send times for individual recipients as well.

In the meantime, Sharma said he sees dynamic content evolving beyond email to include display ads, websites, and social ads.

“It provides a cohesive experience and adapts to wherever the customer happens to be,” he said.

More Email Marketing Resources Here:

Category Content
ADVERTISEMENT
Lisa Lacy Freelance Writer

Lisa Lacy is a reporter who covers digital and search marketing. She has written for CMO.com, TechCrunch, ClickZ, Search Engine ...