AI for Email Marketing

Optimizing delivery times was my first experience applying artificial intelligence to email marketing. The concept is to send an email when a recipient is likely to engage based on her historical opens and clicks — a form of machine learning.

Apple’s iOS privacy changes have made open rates less reliable, but it used to be a solid strategy. Many email platforms offered send-time optimization, including Mailchimp, Iterable, and IBM Watson Campaign (now Acoustic).

AI continues to disrupt email marketing. Platforms such as Grammarly, ChatSonic, ChatGPT, Jasper, and more can generate quality, personalized content quickly.

For example, I prompted ChatGPT to “write a paragraph of email marketing copy promoting my retail brand’s new 20% discount and include a mention of the socially responsible program we’re introducing to donate some of our proceeds to the community.”

Here’s the response:

Exciting news! Get ready to save big on all your favorite products at our retail brand. For a limited time, enjoy a 20% discount on all items in-store and online. Plus, with every purchase you make, you’re not only treating yourself, but also giving back to the community. We’re proud to announce our new socially responsible program where a portion of our proceeds will be donated to local non-profit organizations. Shop now and feel good knowing you’re making a positive impact while saving on top-quality products.

Impressive! The response took less than 3 seconds. It needs a bit of rewriting but it’s a decent foundation.

ChatGPT responded to a prompt for marketing copy in less than 3 seconds. Click image to enlarge.

ChatSonic’s response to the identical prompt is even better, with added details such as a discount code and informal phrasing.

Welcome to [Retail Brand], the place to find all the best products to make your life easier. We’re excited to announce that for a limited time, you can enjoy a 20% discount on our entire selection — just use code DISCOUNT20 at checkout. Plus, to make the deal even sweeter, we’ve launched a new socially responsible program. For every purchase you make, a percentage of the proceeds will be donated directly to local charities. So, do your part and shop with us today!

ChatSonic’s response added details such as a discount code and informal phrasing. Click image to enlarge.

Pros and Cons

AI is a powerful tool, but it’s not perfect. The best results combine AI output with human input. Here are my pros and cons for using AI for email marketing.

Pros

• Efficiency. AI can analyze data and generate quality copy much faster than a human.

• Personalization. AI can personalize content for each recipient.

• Optimization. AI can test variations of copy and optimize campaigns for better engagement and conversions.

• Scalability. AI can analyze and automate huge email volumes, helping large businesses and otherwise freeing up time for other tasks.

Cons

• Lack of creativity. AI-generated content can lack the creative flair of humans.

• Limited understanding of context. AI does not (yet) have a human understanding of language, tone, and context.

• Errors. AI-generated copy contains errors and omissions.

• Dependence. Businesses that rely on AI may struggle to adapt if the technology becomes outdated, restricted, or unavailable.

Future of AI

Neil Patel, the author and influential marketer, believes Google will “crush” ChatGPT. In a LinkedIn post, Patel wrote, “Once Google begins to take their gathered data and develop/train AI tools of their own, chatGPT won’t be able to compete.”

Perhaps Patel is alluding to DeepMind, a subsidiary of Google’s parent Alphabet and a top global artificial intelligence lab. (DeepMind’s CEO recently addressed the platform’s ambitions.)

Regardless, AI-generated content in 2023 is only as good as the data it’s trained on. AI is a boon to marketers, but human review is critical to avoid errors and differentiate a brand. It’s the human element that makes email marketing successful.