Evaluating AI Tools: Key Aspects Email Marketers Should Consider – Litmus


It’s 2023, and artificial intelligence (AI) is everywhere. From generative AI tools like ChatGPT to coding assistants like GitHub Copilot, the potential of AI is top of mind for almost everyone in the digital landscape. However, evaluating its use requires a judicious approach, to say the least.

Read on as we dive into key aspects for email marketers to consider when evaluating the use of AI in their workflows.

Where AI shines

AI is great for tasks that follow patterns and rules. For example, ChatGPT can quickly generate text based on a specified set of inputs, and coding assistants like GitHub Copilot can easily provide autocomplete suggestions while you code. In short, AI tools excel at tasks that are more repetitive in nature, giving marketers time back to focus on more complex tasks.

Now, what exactly are those manual, repetitive tasks in the email workflow? Our 2023 State of Email Workflows Report found the top two biggest bottlenecks in email production cycles to be 1) collecting feedback and 2) content creation.

How exactly can AI help with these two pain points? Given the apprehension many leaders have about AI-generated content, it’s unlikely that AI will assist in gathering feedback and approval soon. But as for content creation, there’s significant promise: it’s how AI is used the most, according to our 2023 State of Email Design Report.

When it comes to content creation, AI can help streamline:

  • Idea generation such as brainstorming and sparking ideas for content topics
  • Email copywriting based on predefined parameters
  • Editing and proofreading to help refine grammar, tone, and style (like Grammarly)
  • Brainstorming sessions for concepts and campaigns
  • Visual content creation such as imagery and video
  • Personalization that can tailor content to subscribers, e.g. predictive recommendations

Where AI falls short

As much as AI offers promise, it also faces significant barriers. Despite how quickly generative AI tools work, they can struggle with things that need a more nuanced, human approach.

For example, if AI is being used to generate code for an email, ensuring its functionality across different settings (like Dark Mode) or making sure it meets accessibility standards requires human supervision.

Another example is with AI-generated imagery; biases inherent in the models might create unfair outcomes, perpetuating bias and discrimination.

That’s just the tip of the iceberg: marketers should also consider privacy when using generative AI chatbots. These chatbots collect information, and the line can be hazy with how that data is used and shared (one of the major reasons why Italy banned OpenAI).

Source: OpenAI Privacy Policy

5 questions to ask when evaluating AI for email marketing

Looking to use AI to help create content for your email campaigns? Here are five questions for you to consider.

Is the email copy accurate?

This is a no-brainer, but if you’re using AI to assist with writing email copy, ensure that any statements made are accurate. Misinformation is one the major pitfalls for AI, so take extra caution when using it to help you write.

Are there any biases’ present in the email imagery?

Are you using generative AI to create your email hero image? Make sure it’s free from bias and  discrimination so as to not perpetuate harmful stereotypes.

Does the code render well?

If you’re using code generated from a coding assistant, make sure it renders well. There are over 300,000 ways an email can render and the last thing any email marketer wants is to deliver a poor subscriber experience. Plus, bad code can lead to the spam folder, too.

1f4a1 - Evaluating AI Tools: Key Aspects Email Marketers Should Consider - LitmusTip: Test your emails with a tool like Litmus, which makes it easy to preview how your code will render across clients and devices.

Is your email optimized for Dark Mode?

Another thing to watch out for in AI-generated email code is Dark Mode, which may not factor in the intricacies that come with coding for Dark Mode emails. Emails that aren’t Dark Mode optimized might have indiscernible text (e.g. dark text on a dark background) leading to a poor subscriber experience. If you’re not optimizing your emails for Dark Mode, you’re potentially neglecting 35.4%1 of your subscribers—so don’t leave them in the dark.

Is your email accessible?

Making emails accessible means ensuring that everyone, no matter their disabilities or any assistive devices they use, can easily understand and receive your message. If you’re using AI-generated code, ensure it meets accessibility standards, whether it’s for color blindness, readability, ALT text, and more. This goes a long way in creating inclusive emails for everyone.

How will you use AI?

Today, in a world where technology is able to act as a partner, not just a tool, we’re committed to helping you navigate and embrace artificial intelligence (AI) with integrity and confidence.

Try it for free

1Our Dark Mode data is based only on the Apple Mail app for iOS 13, macOS 13, iOS 14, macOS 14, iPadOS 14, iOS 15, iPadOS 15, and Apple Mail privacy-impacted devices.

The post Evaluating AI Tools: Key Aspects Email Marketers Should Consider appeared first on Litmus.

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

We Know You Better!
Subscribe To Our Newsletter
Be the first to get latest updates and
exclusive content straight to your email inbox.
Yes, I want to receive updates
No Thanks!
close-link

Subscribe to our newsletter

Sign-up to get the latest marketing tips straight to your inbox.
SUBSCRIBE!
Give it a try, you can unsubscribe anytime.