How I Use AI in Email Marketing With My Clients

Over the course of the past year, I've been exploring how different artificial intelligence (AI) tools stack up against each other—and what kind of content quality I can expect from each one.

My initial testing shows that there's some promise in using generative AI to create short-form content, like personalized email marketing messages. But it's not right for every project—here's how I decide when and how to rely on AI.

• Testing AI performance
• Discussing AI with clients
• Using AI for client work
• Editing AI content
• Tracking AI content performance
• Continuing to use AI in email marketing

Testing AI performance

Before incorporating AI into any client projects, I used my own agency, DigitalME, as a test case. My team and I started by sending out a newsletter entirely written by an AI-powered tool.

I wanted to make sure it would wind up in their inboxes—not the bin or the spam folder—just like any other email. And I didn't tell any of my subscribers that I was testing AI. Instead, I saved that revelation for the very last line of the email.

We sent our message to 4,500 contacts, resulting in a 65% open rate. That's a 15-20% increase over our normal metrics—so right away, I knew I was on to something good.

After the big reveal at the end of the email—the fact we created it with the help of AI—I included a call to action that let readers know they could get in touch with us if they wanted to explore how to use AI in their email marketing efforts.

Discussing AI with clients

If a client comes to me with interest in using AI, I'm happy to chat with them about it right away. Otherwise, I tend to wait to bring it up. Despite the promising results of my test, I'm still not defaulting to AI with every client account.

AI generated content still isn't a complete replacement for human-crafted messages. And I prefer to take a human-first approach to my clients' projects. Doing so enables me and my team to understand every bit of a client's campaign and ensure their voice and tone is bang-on.

I'll typically only suggest AI to my clients if their budget doesn't support the cost of engaging my team to write their email copy. In that case, I'll offer AI-assisted copywriting as an option.

A handwritten email will run a client $400. Opting for AI-assisted email content brings the pricing down to $250. This can make working with us more feasible for some clients—and they still benefit from my team's expertise

Using AI for client work

If you're interested in using AI in your or your clients' email marketing, be sure you've spent time working with different tools. You don’t need niche AI email marketing tools, either. You can use an AI chat tool and your regular email provider.

My agency tested both ChatGPT and Google Gemini. Personally, I'm keen on using Gemini. It's the most responsive to how I naturally write, and it's given me some high quality outputs. Plus, I like that Gemini is part of my team's Google Workspace accounts.

But choosing a tool you like is only part of the process—you've got to train the AI on voice and tone, too. This can get to be a bit complex if you're working with multiple clients. I make sure that I have different chat threads for each of my clients, and I train each one with human-written examples of what I want.

My team and I have also had to invest time and resources into honing our prompt engineering process. We've found that short, direct prompts work best for generating relevant content—and it helps to do some of the initial editing and revising right in the AI chat. This further trains the tool on what we do and don't want in our outputs.

Editing AI content

We don't stop the process there, though. Once we've got a workable draft out of the AI tool, we still make sure it's edited by either myself or a member of my team. This allows us to catch any errors and make the copy feel more like a human has written it, not an algorithm.

The trick is to find a balance between comprehensive editing and total rewriting, as that would defeat the whole purpose of using AI in the first place. Consider these best practices when editing AI content:

• Look for big errors, like misspelled names and poor punctuation
• Adjust format (AI doesn't always put the line breaks in the same way you or I would, for example)
• Make slight tweaks to bring the content more in line with the client's style and improve natural language
• Ensure the email subject line and call to action (CTA) are strong

It's essential to do this so that the content sounds human. Otherwise, you run the risk of turning off recipients—and even getting your clients' emails labeled as spam.

Tracking AI content performance

Once I send out a client's AI-written email, I track it just like any other email marketing campaign: by monitoring click-through rates, conversion rates, email deliverability, and more. I want my clients to get a great return on their investment, whether they paid my team to hand write their content or they opted for AI.

And audiences are responding well to AI content. So far, my clients are typically pleased with the results compared to their financial investment in each campaign.

But I make sure to educate them that AI technology isn't a magic solution. Everyone involved in an email marketing campaign—my team and our clients alike—needs to follow best practices for responsible sending. You also still have to pay attention to things like subject lines, send times, target audience segmentation, and more.

At this time, I’m only using AI for content creation. All segmentation, strategy, and building of workflows is done by my team.

Continuing to use AI in email marketing

While I'm all-in on exploring AI in email marketing right now, I'll stop using these tools at any point if they're no longer benefiting my clients.

Google and Yahoo have announced new security and marketing authentication standards for anyone sending bulk emails in 2024. The goal is to reduce spam and improve email security—something I'm in favor of. Nobody likes a spam email full of dodgy links.

I don't see the use of AI becoming an issue, if it's just used to write email copy (and there's still a human in the mix every step of the way). But I'll be keeping an eye on these security developments.

And it's one of the reasons I recommend that anyone interested in growing an email list hire a professional who's skilled in that form of marketing. It's my job to keep an eye on developments like these security changes and make sure that we're creating and sending emails that convert.

The average business owner is so busy with other tasks that they might miss an update related to email marketing—and accidentally damage their sending domain in the process.

Incorporate AI into your email marketing strategy

Not all AI tools are created equal—and you may actually be able to get a better product out of a free app (like Gemini) than a paid SaaS tool that says it's dedicated to email marketing copy. You just need to know how to engage with the tool correctly, or work with someone who does!

And even if you're familiar with using AI writing tools, it's a good idea to consult with someone familiar with the requirements that, if violated, could get your emails flagged as spam.

Visit my profile to get in touch and learn more about DigitalME email marketing services. If you're interested in exploring how to leverage AI for faster, more successful email marketing, I'm ready to chat.