Go Beyond Statistics: How Do Real Email Marketers Use Gen AI

Cover image for an article on how real email marketers use AI email generation
28 April, 2025 • ... • 1 views
Diana Kussainova
by Diana Kussainova

You’ve probably seen a million reports showing how popular AI is for email marketing. But do email marketers generate entire campaigns? Do they review the AI output before sending? 

As avid fans of new technology (and an email service provider with a great AI assistant), we decided to get to the bottom of the topic and find out! We asked real email marketers about their AI use cases and added our tips.

Generative AI for email marketing: What’s the deal?

According to statistics, 69% of marketers are already using AI tools in their work. From specialized platforms to AI in email service providers and chatbots, AI-based email marketing has been on the radar for a while now. 

Over half of Selzy users are sure that AI can improve email newsletters. Some of its use cases include email copywriting and image generation, enhanced segmentation, and hyper-personalization. Recently, it has become possible to create a complete campaign — from the HTML template to the body copy — from a single prompt, in an AI email generator.

You may feel like you’re behind on this trend or wonder to what extent it is true. Plus, however convincing, statistics lack personality, so it’s hard to apply them to your own work. And if you’re like us, you’re probably curious: “Do professionals really send fully AI-generated emails?”

We’ve gathered more than 20 expert opinions on the topic from marketers working in various industries and are ready to share!

How real email marketers (don’t) use Gen AI in their work

Overall, most of the experts shared that they use AI chatbots and dedicated tools for subject lines, A/B testing, and brainstorming. The work Gen AI does for them almost entirely concerns copy. That said, every email marketer pointed out that they review the output. This means editing the copy to make it more relevant to the brand and more personal.

Here is how Mimi Nguyen, the founder of Cafely, explains her process:

Mimi Nguyen's headshot
Mimi Nguyen

Founder of Cafely

I do use AI to help generate emails, BUT with strict oversight! I mostly use it to brainstorm subject line variations, draft initial content structure, or repurpose blog content into email copy. I always rewrite whatever output it generates before pushing through with the next step (design, publishing, or distribution). 

Besides not wanting to make our content look obviously AI, I see tone and brand voice as very critical. Especially in campaigns like nurturing campaigns or high-conversion product launches.

Some experts noted that AI can be especially helpful when it comes to certain email types. For example, Raviraj Hegde, Head of Growth at Donorbox, believes AI drafts for newsletters and onboarding sequences can be a good starting point:

Raviraj Hegde's headshot
Raviraj Hegde

Head of Growth at Donorbox

Where AI has been useful is in planning out and drafting initial iterations of onboarding sequences or newsletter copy. I use tools like ChatGPT for initial drafts or Copy.ai to test out different headline styles. It accelerates the process, but every draft gets passed manually. Nothing leaves without a human touch.

Simone Byrd, Content Specialist at Online Optimism, shared the same sentiment but also made it clear that AI can’t nail every company newsletter. Its output can vary depending on the industry the business is in: 

Simone Byrd's headshot
Simone Byrd

Content Specialist at Online Optimism

I find AI to be the most helpful in creating short newsletter emails for clients who send regular emails to their audiences. This works best when you have a straightforward subject matter that the LLM can easily write about. For example, at our agency, we have a client that offers training for commercial truck drivers, and another client that offers transcranial therapy for depression treatment, and I find that Claude generates better copy for me when I prompt it to write about the truck driving content.

Before we get into specifics (and answer your burning question), check out other insights we’ve gained from these and other experts:

  • Besides ChatGPT, email marketers also use Claude, Jasper, Copy.ai, and tools that come with their email service providers of choice.
  • Some AI-generated copy performed better than human-written, some worse. But an interesting detail is that AI can help improve the open rate, but it gets difficult with click-throughs.
  • Email marketers strongly believe that their audience can sense when email copy was produced by AI. Some even had customer feedback about it!

Do email marketers generate entire emails?

From what we’ve gathered, a part of email marketers experimented with a fully AI-generated email at least once! Although experts mostly mentioned copy, there was one striking case from Priyanka Prajapati, Digital Marketer at BrainSpate. The team not only generated several copy variations for their campaigns but also created images. Check out the detailed explanation, complete with results:

Priyanka Prajapati's headshot
Priyanka Prajapati

Digital Marketer at BrainSpate

We used AI pretty heavily for a B2C fashion brand’s seasonal sale campaign. Used Jasper + Notion AI to draft the emails (including subject lines and CTAs) and Midjourney for some quirky product visuals with a surreal aesthetic to match the brand tone.

But the key wasn’t just “generate and send.” The workflow was: AI drafts — human tweaks — tone-match layer — behavioral segmentation tweaks — manual overrides for VIPs.

And we didn’t just send one version. We had 3 tonal variations — playful, urgent, and sarcastic — and mapped them to past click behaviors using Klaviyo. 

Result? 18% higher clickthrough rate vs. last year’s same promo. And a 27% increase in returning customers within 14 days. All because we let AI handle 70% of the heavy lifting, so we could go crazy on micro-personalization.

A more common experience is different, though. Marketers who spent time generating an entire email campaign hadn’t had impressive results. Here is how Josh Neuman, the founder of Chummy Tees, relays his experience:

Josh Neuman's headshot
Josh Neuman

Founder of Chummy Tees

I’ve run a couple fully AI-drafted emails just to test — like a “cold day, warm tee” type of campaign in early fall. We sent it to a small segment, and it did okay. But it was missing a bit of soul. Since then, I have never sent anything out without editing it myself. I don’t want it sounding like a machine. My customers don’t either.

Others, like Arthur Favier, the founder and CEO at Oppizi, even had negative customer feedback and ultimately decided to be more cautious with AI email generation:

Arthur Favier's headshot
Arthur Favier

Founder and CEO at Oppizi

I’ve run a couple of full email flows using AI for initial drafts just to experiment — particularly for a re-engagement campaign we did last year. We generated a three-email sequence using ChatGPT, then revised it to match our tone and goals. Open rates were average, click-throughs were okay, but it didn’t outperform our usual manually written content. Still, it gave us a few decent angles we wouldn’t have thought of on our own.

There was a time we used Jasper to draft part of a campaign focused on event invites. The copy sounded a bit generic, and it didn’t connect with the audience as well as we expected. Engagement dropped, and a few people actually replied, saying it felt impersonal. That was a clear sign it missed the mark. Since then, we’ve been more cautious.

And Sam Speller, the founder of Kenko Matcha, even challenged AI to outperform humans and had worse results:

Sam Speller's headshot
Sam Speller

Founder of Kenko Matcha

One of our most frustrating experiences testing AI email generation was with Jasper AI for our seasonal matcha subscription campaign. We wrote long prompts describing our organic, stone-ground matcha and its health impacts, and the generator kept spitting out copy that was vague and salesy, with words like “revolutionary green tea powder” and “miracle health results” that could not have been further from our artisanal brand identity. And worse still, the AI hallucinated claims about “clinically proven weight loss effects,” which we had to delete to allay compliance risks. Despite many rounds of refinement, the output typically still needed heavy editing 80% of the time, cancelling out the time-saving proposition. 

When we A/B tested AI-generated emails versus those created by humans, we found the latter had a 35% higher open rate and 50% more clicks. The AI versions (which we had spent 12 hours tinkering with) were no better than our old templates.

AI email generation debates

So far, we’ve only talked about typical experiences and general opinions. There are, however, at least two topics email marketers debate when it comes to AI.

Bland vs Emotional. Some experts believe that AI can’t understand emotional nuance and produces bland, even soulless copy. On the other hand, email marketers also use AI chatbots to analyze customer sentiments or compose nuanced messages communicating delays and other challenging situations.

Quick vs Time-consuming. One perspective on AI tools that many email marketers share is that they can really speed up the process, especially when it comes to first drafts. Another perspective, however, is to also count the time spent editing and revising the AI output, which then often exceeds the time a real person would have completed the work in.

Where’s the truth? You’ll have to try and see for yourself since every business is different, and the quality of AI’s work depends on the tools, training, and prompts.

Unique and inspiring AI use cases

What’s often missing in Gen AI discussions is the fact that this technology is extremely versatile. There are dozens of options on how you can adapt it to your business needs and workflows, yes, even the all-mighty non-specialized chatbots.

We’ve found two examples of creative and practical approaches that can bring out AI’s strengths. Gerry Wallace, Managing Director at Greenline, a company building custom outdoor spaces, shared how AI’s ability to role-play helps in email marketing. Often, marketers tend to assume that AI can only act as a writer, but it can actually be a reader, too!

Gerry Wallace's headshot
Gerry Wallace

Managing Director at Greenline

We don’t rely on AI to come up with email content. That’s still a human job, and our tone, positioning, and voice are too specific to outsource. Where AI does shine is acting like our audience to spot what’s missing from our content or strategy.

After we draft a campaign, we’ll feed the full copy into a model and prompt it to “be the customer” where it can ask questions, spot where things are unclear, or point out what’s missing.

It’s been a good thought experiment, and sometimes it hits the nail on the head. Yes, we may not always use its inputs or agree with the justification, but it’s definitely something with potential. And I think that’s the value, clarity, not creativity.

Another great idea comes from Kelly Redican, Digital Marketing Manager at A.M. Custom Clothing. The team uses AI to update last year’s seasonal email campaigns that are featured in their marketing calendar

Kelly Redican's headshot
Kelly Redican

Digital Marketing Manager at A.M. Custom Clothing

With the nature of our business, our Marketing calendar remains pretty static each year, so there are a number of email campaigns that we continue to send annually. We have incorporated ChatGPT into these campaigns by creating a Newsletter Reviser prompt within the interface. The prompt understands our tone of voice, the nature of the writing, and who we are addressing. We take an email from a previous year and use the prompt to help rewrite the email, using up-to-date information and offering alternative angles.

This benefits our company as we save time and resources, and what could take us half a day can now be completed within 30 minutes or so. We are also able to bulk schedule campaigns ahead of time in a minimal amount of time.

Is using Gen AI for email marketing easy?

Yes! Well, yes, if you have the right tools and use them smartly. 

Reading the expert opinions we’ve gathered for this piece, we actually discovered a big knowledge gap. Although it’s been over 2 years since ChatGPT’s launch, and AI technology has made big leaps since then, most email marketers still use Gen AI only to create copy! 

First of all, Gen AI can do much more than that. And second, it can be even easier. With Selzy, you don’t have to switch to a different tool if you need to use AI. Our platform comes with a built-in AI assistant that can analyze your email campaign as a fellow marketer or psychologist, generate copy and images, and even make a complete HTML email campaign from scratch. It’s not only easy, it’s also fast, and, dare we say it, addictive.

For the upcoming Star Wars Day, we’ve decided to prompt Selzy’s AI to create a May the 4th campaign for a made-up clothing brand called Galaxy. Of course, this is the perfect opportunity to use segmentation!

Here are our prompts for the dark and the light versions:

  1. Create a holiday email marketing campaign for a clothing brand called the Galaxy about a May 4th (Star Wars Day) sale. Use black and red colors, but make the email contrasting enough. Add images of dark clothing. In the copy, write something friendly about joining the dark side. Add recommended product blocks.
  2. Create a holiday email marketing campaign for a clothing brand called the Galaxy about a May 4th (Star Wars Day) sale. Use white and blue colors, but make the email contrasting enough. Add images of light clothing. In the copy, write something friendly about joining the Jedi. Add recommended product blocks.

And here are the campaigns themselves. Which one do you prefer?

An email with red and black colors featuring the banner text “Join the Dark Side” and two product recommendations at the bottom
Source: Selzy
The same dark-themed email shown in Selzy’s email editor: the email is on the left, while the AI assistant interface is on the
Source: Selzy
An email with white and blue colors featuring the banner text “Join the Jedi” and three product recommendations at the bottom
Source: Selzy
The same light-themed email shown in Selzy’s email editor: the email is on the left, while the AI assistant interface is on the right
Source: Selzy

Selzy’s AI can be useful for ideation, first campaign drafts, and quick visualization. The best part is — it’s so easy, even your colleagues from other departments can do it!

Quick tips on getting the Gen AI for email marketing right

As with any other technology, AI is only as effective as those who use it are skilled. So here are some of our tips to help you reap good results:

  • Don’t just copy and paste AI’s output. Edit it, adapt it, humanize it.
  • Consider the time spent working with AI vs without it. Remember that it needs to actually benefit you.
  • Use AI for drafting, campaign analysis, and revising human-written copy. 
  • Find AI tools that work with your other tech stack and give the best results. Dedicate time to make them align with your brand’s tone of voice, unique value proposition, etc.
  • Try using AI for A/B testing and segmentation.
  • Consider the ethics of using AI and its environmental impact. Make sure it complements your business’s positioning and values.

Last but not least, we strongly advise you to be transparent with your customers about your use of AI. Here’s how Phil Portman, founder and CEO of Textdrip, approaches it:

Phil Portman's headshot
Phil Portman

Founder and CEO of Textdrip

If I use AI heavily in crafting an email, I include a disclaimer (e.g., “This message was assisted by AI”) to maintain transparency. It’s similar to how we’ve long used “This is an automated message” in out-of-office replies. When the AI’s role is minor — like grammar touchups or rephrasing — I don’t feel a disclaimer is needed.

28 April, 2025
#AI
Article by
Diana Kussainova
Writer, editor, and a nomad. Creating structured, approachable texts and helping others make their copies clearer. Learning and growing along the way. Interested in digital communications, UX writing, design. Can be spotted either in a bookshop, a local coffee place, or at Sephora. Otherwise probably traveling. Or moving yet again.
Visit Diana's
Latest Articles
Selzy Selzy Selzy Selzy