While we were working alongside many early-stage ecommerce businesses some time ago, there was one pattern that really stood out. The email campaigns of these companies were gorgeous, filled with great-looking AI-generated subject lines, dynamic product blocks and perfectly timed sending “schedules” … but they all saw an increase in their unsubscribe rate and saw their revenue stagnate.

At first glance, it looked paradoxical. These companies were using great marketing technology, right? And they followed the best practice checklists to every letter, didn’t they?

It wasn’t the tools that caused this — it was the assumptions these email marketers made while using them.

They viewed “AI in email marketing” as just another tactical shortcut; they viewed “AI” as just an easy way to automate the creation of email copy, personalize an email at scale or achieve higher open rates through some sort of algorithmic guesswork.

Peter Drucker was ahead of his time when he stated that, “It is the customer that determines what a business is.” This continues to be true today, but businesses have found ways to honor it through new methods.

AI should not replace our judgment, but should be used to increase customer relevance in a manner that could not be achieved by manual segmentation. The risk of AI does not come from using it but from using it without a strategy.

The “spray-and-pray” trap, upgraded

The “spray-and-pray” philosophy of digital marketing (i.e., sending out the same basic email to an entire mailing list and hoping for a small number of responses) required no understanding of the customer or what would connect them emotionally.

But personalization without insight is just noise with a first name field.

In fact, an analysis of 40 billion emails found that campaigns that used automatic triggering based on customer behavior would result in greater than 50% open rates as well as significantly higher lifetime value of the customer.

Yet, many companies’ use of AI is as simple as changing “Hi [First Name]” to “Hi [First Name], We Miss You!” without understanding the context behind that message and what will make it more likely for a customer to make a purchase.

This is what separates successful and unsuccessful campaigns. The success of AI marketing is dependent on having a clear strategy regarding the intent of the message.

• What specific problem does my product/brand solve?

• What is in it for me?

• How can I make my life easier, more fulfilling or less stressful than it is currently?

Without understanding these concepts, even the best algorithm will result in sub-optimal campaigns.

Automating campaigns is just one step in the evolution of marketing. Looking ahead, we have to understand how to develop predictive marketing, allowing the consumer’s needs to dictate how to communicate.

From automation to anticipation

Using AI in email traffic is not as much about getting faster — it is much more about understanding better!

For example, send-time optimization (sending the email at the best time based on the likelihood of the recipient to open it) can seem like a technical aspect of an email system. When you dig deeper, however, you begin to understand how it reflects an overall cultural shift in email marketing from demanding attention on your scheduled time frame to meeting customers at the place that they are at the time that you send them your email.

Similarly, recommendation engines that use predictive analytics to suggest products based on a customer’s immediate need are valuable not only because they “predict” what a customer may want. Studies show that personalized product recommendations — often powered by AI — can also increase conversion rates by an average of 22.66%, demonstrating the commercial value of anticipatory suggestions!

The results were boosted with marketing messages such as:

“You may want to consider investing in a rain jacket for your next trail run.”

By accurately predicting a customer’s need for a product, businesses have significantly driven an increase in sales with limited marketing spend.

That’s empathy, scaled.

The real risk is losing the human thread

Although AI is capable of creating a subject line, it does not have the capacity to capture the nuanced humor of an entrepreneur who has experienced the same challenges as his or her client.

Although it can analyze the tone of an email, AI cannot express this meaning in the same way as an individual.

There are situations where artificial intelligence has artificially swapped a formal sign-off of “Best Regards” with one that was intended to sound “cool and trendy” — for example, “Stay Lit Fam.”

Those campaigns resonate with the very audience that was intended to be targeted because the brand’s voice was not created with a natural reflection of an authentic relationship with customers.

When it comes to effective email marketing strategies, the most successful marketers use AI technology as a supportive tool to assist them with email writing; in other words, AI takes on the role of co-pilot and not pilot when creating a campaign.

Email marketers begin their campaigns by reviewing the data collection from interviews, support tickets and user behavior, and provide that information to artificial intelligence.

Even better, email marketers will add in emotional context, such as handwritten PS notes and their failed product stories, to provide authenticity to their campaigns. For example: “We noticed you’ve been quiet — everything okay?”

That’s the kind of content that builds trust in an age of AI spam.

Start small, think strategically

You can start small and build your tech stacks to achieve your business goals with strategic testing.

1. Audit one campaign to see how your AI is utilizing personalization based on users’ actions, or whether you’re only using location or other demographics, to send the personalized message.

2. Humanize the customer experience by including a line that only a human could have written based on the customer experience; for example, using something like a customer review or shared customer experience.

3. Measure how many people click on your emails and how many of them become repeat purchasers or provide feedback, such as returning to inquire about additional products.

In my opinion, focus on proving your value before you scale your business. Validate whether your messaging is being received before you automate your communications on a larger scale.

Email marketing isn’t dying. Irrelevant email is.

Marketers who are successful will not be the ones who use the “best AI” — they will be the ones who understand that behind every inbox is a human being.

And that’s something no algorithm can fake.