Can AI Simplify the Email Workflow? – Litmus

In the 2023 State of Email Workflows Report, we discovered marketing teams spend on average two weeks or more to create a single email. That’s from the campaign inception to the campaign delivery—they’re spending two weeks on everything that goes into a great email campaign: strategy, copy, design, code, setup, segmentation, testing, approvals, and sending.

As the promise of artificial intelligence (AI) is top of mind for everyone right now, our main focus at Litmus is optimizing the email workflow. This raises the fundamental question: 

Where will AI have the greatest impact on streamlining the email workflow, so marketers can spend less time on manual, tedious tasks—and more time creating amazing campaigns?

Email marketing vs. email writing: We’re talking about more than copy

First, I think it’s really important that we distinguish the difference between email marketing and email writing.

Using generative AI to write your business emails, sales emails, and cold emails is a valid tactic. You can train generative AI to write like you do and to learn the value in what you’re selling—and even help you personalize the emails easily. The options are endless and the potential is huge.

Dear Reader, this article is not about that kind of email.

We’re here to talk about how AI will impact email marketing. I’m talking about the one-to-many communications your marketing team sends to your customers and subscriber every day. I’m talking about the channel that wins the award every year for the strongest ROI of any channel ($36 for every $1 spent). Compare that to a cold email, where response rates alone are well below 10% on average.

Yes, these emails include copy (and are sometimes even plain text), but they are meant to deliver brand communications to an opted-in subscriber who has consented to receiving emails from your brand. 

They need to have:

• A compelling and engaging design that aligns with your brand experience
• Valuable content that a subscriber or customer can use or benefit from
• Clear insights on the back end to help you measure what’s working

This is not to say that AI is not already changing email marketing. It’s just that these things are different and the areas of potential impact are also going to be different.

So what’s taking so long? 

We asked marketers in our State of Email Workflows survey about where they’re spending the most time. Let’s take the most time consuming areas to see what AI opportunities will look like.

Feedback and approvals

Teams report that the most time consuming part of sending an email is the feedback and approvals stage of the send. Anyone who has ever tried to get an email out the door and needs to get approval from multiple stakeholders can relate.

It goes something like this: You send a test email and forward it to the group who needs to review it. You think you’re asking for a last pass at grammar edits to give you some peace of mind. Inevitably someone in the reviewing group takes that moment to question the approach, the image, the call-to-action (CTA) button color, the audience—or even the whole idea.

(Note to marketing leaders here: let’s either empower our teams to own the final product or be timely in our feedback.)

How can we speed this up? We recommend taking an Agile approach, a strong creative brief, and very defined roles within the proofing stage. (And Litmus Proof takes those email threads out of rotation altogether.)

Is AI going to have an impact on this? We’d love it to—but considering the apprehension around AI-generated copy, the fear from higher-ups might actually lead them to getting *more* involved in the short term. To ensure the multitude of boxes are checked correctly for the perfect email send, humans need to be involved in this process. 

Content creation

This one gets a pass as we believe it’s really where the time should be spent. Inboxes are crowded and email is as close as we get to our customers, prospects, and communities. Here at Litmus, we advocate that this is the space for your best creative thinking and to give yourself time to create the most engaging campaign.

Is AI going to have an impact here? Yes, indeed. This is arguably what most of the buzz is about. 

Generative AI can speed up many areas here:

• Idea brainstorming
• Generating subject line options
• Copywriting and editing
• Image generation
• Video content creation

And don’t forget about generative AI’s partner in crime—assistive AI that encompass tools like Grammarly to support content creators while they’re doing their best work. 

Design

Of course, we’re big believers that your email design should be some of your best work, built to drive your most compelling and thoughtful campaigns. It’s insulting to subscribers everywhere when brands spend the big bucks and the big resources trying to pull in new audiences when loyal subscribers are stuck with an afterthought like “template 42” with a logo upload.

Is AI going to have an impact here? Yes—it’s just that the jury is still out on how.

Generative AI is moving quickly on the design front, but when it comes to being on-brand and effective, you’ll want to pay as much attention to design here as you are to the copy to ensure your brand stands out in the inbox. We’ve all seen how AI has struggled with the human hand.

Machine learning and predictive analytics will play an increasingly important role in how campaigns are designed for personalization and that strategy will be core to the design.

Coding emails

A strong workflow process and project management tool will allow this phase to be as quick as possible at this stage. And having an email developer and email designer who have crafted a highly templated, modular email design system with re-usable components will also make things not only faster for the team but easier. 

I’d way rather our email developer Carin Slater spend her time testing new ways to innovate and personalize emails than custom coding every single email design. (And luckily Carin would rather be doing that too.)

Will AI speed this up?

The jury’s still out on this one. While there has been some experimentation with generative AI tools in email code creation, AI misses the mark by neglecting to take into account of things like Dark Mode, the need for accessible code, and potentially 300,000+ ways an email could render. A dedicated email developer will always take on these considerations.  

QA testing emails

Testing is infinitely easier than it was 10 years ago (something we’re particularly proud of). And that’s largely due to the manual nature of testing in the early days.

While those days are over, email rendering environments are still a black box with no unified standards. Email clients are still changing how emails are rendered without any warning or notice.

Will AI play a role?

Absolutely. We envision a world where apps are infinitely more helpful, guiding you through natural language processing to solve for why emails aren’t showing up with the design you perfected. But in the mean time, with complex emails that include high levels of personalization or live content, you’ll want a set of human eyes testing every facet of your emails to ensure you never send a broken email.

The email workflow is complex and nuanced. Everyone does email differently. Yet,  there are commonalities in how we do email that could be made better, easier, and less stressful with some help from AI. Our ongoing exploration and accumulation of insights are illuminating what those possibilities could look like.