How many times have you used ChatGPT yet? (Notice how I didn’t ask if you’ve used it). What about DALL-E2? Or Midjourney? Every little aspect of marketing can now be optimized, improved, and speed up thanks to Artificial Intelligence.
But let’s not talk about every aspect of marketing. There’s probably an AI for that. Let’s talk about something we’re good at: email marketing strategies.
If you’re a true AI enthusiast, you didn’t miss the pun two sentences before and immediately thought about our holly mecca of new tools: theresanaiforthat.com. In the image below, there are all the solutions we currently have for email writing.
That’s right! Just writing. And as we know, building effective email campaigns is much more than just crafting a smart copy. There is audience segmentation, offer personalization, template optimization, A/B testing, and all kinds of good stuff AI may or may not have already taken over.
That’s why you’re here, isn’t it? To learn exactly how AI can help in your email marketing efforts that stretch beyond writing. So buckle up, and let’s get to it!
How does AI even work?
First of all, it’s important to state that when we talk about AI in email marketing we actually mean a couple of different technologies: machine learning (ML), computer vision, natural language processing (NLP), deep learning, or even email marketing automation.
Related read: Marketing automation for B2B companies Guide
AI tech is trained to analyze large datasets and work with big data to make predictions and provide recommendations that significantly boost personalization and segmentation of our digital marketing campaigns.
Natural language generation AI solutions (like the infamous ChatGPT) compute vast amounts of data and generate responses based on whatever’s been fed to them. So the quality of the output you get directly depends on the depth of the data you’ve been able to provide it with to begin with.
ChatGPT, for one, was trained in 2021, so for now, don’t expect it to talk about the post-Covid societies or the war in Ukraine.
More advanced machine learning systems learn how to improve their predictions and decisions over time. Others are able to gather the most recent or even real-time data to optimize their performance further.
AI in email marketing
The latter is particularly important in digital marketing and email campaigns. We want our automated messages to sound as real and natural as human communication would, so it’s important for the AI tool to have the latest information about the user or the recipient on the other end.
But how to get that information? Email databases already have quite sufficient data regarding your contact list. If you take it a step further and let AI crawl your website and gather zero-party data on user behavior throughout their customer journey, you should be able to leverage AI to create much more personalized emails.
The current state of AI email marketing
The future is here, and it has been for a while. You might be surprised, but a lot of processes that keep your email marketing campaigns running on a daily basis are already AI-based.
AI can already be most email marketers’ faithful personal assistant!
Here are the ones you probably interact with on a daily basis, just to name a few:
- spam detection
- inbox delivery & filtering
- smart compose
- optimized send times
- product recommendations
- subject line generation
Now, let’s take a closer look at each of these ways AI tools make our email marketing experience better.
The spam folder has existed since the 1990s. But it had little to do with AI. Remember how some parts of AI share many similarities with automation? That’s just another example of that!
Today, AI spam filters are present in virtually every email client, diligently scanning each incoming message for any questionable content and identifying risky sending patterns. Their job is to prevent invalid email addresses from entering your inbox, ensuring you only receive the most relevant messages.
How does Gmail decide which message from Uber Eats deserves to land in the “Promotions” folder and which needs to go to “Spam” to never see the light of day? With sentiment analysis, that still has a long way to go.
But such a basic application of AI-powered software for email validation already significantly improves our sentiment towards email deliverability and user experience with email clients, such as Gmail.
Filter by category
Gmail (as well as other providers) now takes on even more responsibility in managing our inboxes and even sorting our mail for us.
The algorithm still relies on the user to assign specific personalized labels and quite often mistakes important emails for “Everything else”, but 9 out of 10 times it gets it just right, saving us precious minutes in the day.
Google’s AI even takes a step further by automatically sorting your inbox into “Primary”, “Social”, and “Promotions” emails. It may not help email marketers maximize inbox delivery for their email campaigns, but it surely helps the end user!
You may not be impressed by Google’s Smart Compose if you spend several hours a week playing around Grammarly’s AI, but for the rest of us, it’s a pretty big deal!
The thing is not even about the microseconds won from pressing Tab over completing outgoing messages on your own. Instead, it’s about matching your tone of voice, suggesting the right polite sign-off, or reminding you that you might want to ask the recipient how their day is going before sharing your one-of-a-kind offer with them.
For email marketers
But what if you’re in marketing and need to talk to hundreds or thousands of people at the same time? Is AI still helpful?
Well, it’s getting there.
Optimized email send times
For now, you’re probably best familiar with optimized sending times and email scheduling as crucial elements of your email marketing strategy.
Basically, instead of guessing or googling “The best times to send my product newsletter”, you pass it on to AI that analyzes your open rates over time and establishes the perfect timing for the email to be deployed.
Simple yet highly effective.
Email product recommendations
We all know how product recommendations work, right? You went online doing competitor’s analysis, browsed some brands, a sneaky cookie or another bot recorded that information, and now you have competitors’ ads following your every click.
With email, things are a bit different. Less invasive than digital ads or mobile push notifications, and harder to miss than a stealthy web push.
Here, a business will be only reaching out to those contacts who have already explicitly agreed to receive communications from them. Then, recommended products would be based on the learning model embedded in the website’s structure, thus dealing with invaluable zero-party data, not a third-party hand-off.
This way, embedding AI product recommendations into an email might be even more effective than simply showcasing it in the browser!
Subject line generation
Last but definitely not least, you can always use AI to help you write email subject lines. After all, what could be more beautiful than the possibility of passing on the responsibility of creating a converting subject line to a robot optimized for delivery?
No more writer’s block. No more hours spent trying to desperately choose between “🙌” and “⚡️” at the beginning of your statement. No more drama!
AI is already actively used by most email marketing platforms and you can even test it now with any paid GetResponse plan – how cool is that?
How to use AI for crafting better emails
And it only gets cooler from here! Here are 5 new, better ways for you to implement AI into your email marketing strategy today and 5 awesome AI tools that will make it a breeze.
AI for email writing
It’s not a big secret that artificial intelligence is becoming more and more helpful when it comes to copywriting in email marketing. And it goes way beyond its ability to write catchy subject lines!
When it comes to email copywriting, AI can help you with:
- Overcoming writer’s block
- Producing high-quality summaries of the long-form content
- Tying your newsletter together with an umbrella topic
- Creating catchy subject lines, preview messages, and CTAs
- Proofreading the copy you’ve written
- Converting blog posts into newsletter sections
Highlighted tool: copy.ai
copy.ai is perfect for teams who are happy to pass on most of their copywriting tasks to the AI. From emails and social media posts all the way to full-on blog posts, the tool promises to deliver “premium results in seconds”. Or so it says.
Generating content here does take a while. Of course, not as long as preparing an email marketing campaign from scratch would, but still longer than working with an average prompt in ChatGPT, so that can get a bit unnerving. Especially so, if you’re not guaranteed to get the result you’ve been hoping for:
I used a suggested prompt and still received an error upon exit
AI-powered tools for inbox management
I, for one, am terrible at managing my emails. I have 4 providers, 775 notifications on my private Gmail, and whooping 4376 unread messages on my work Outlook. But am I doing something to fix that? Not quite.
Gmail and Outlook already try to address the issue to an extent, but we all know that there’s no limit to perfection, especially when one’s goal is as ambitious as reaching inbox zero every day. That’s why AI-powered inbox management is a must-have for busy marketers.
When it comes to inbox management, AI can help with:
- assigning automatic tags to emails
- bringing all your emails into a single dashboard
- creating shared inboxes that actually get opened
- engaging your team in cooperation
- managing customer accounts and sensitive data
- saving time on reading, composing, and sending emails
Highlighted tool: Front
Front uses AI to help teams work more efficiently when it comes to email management. It improves customer experience by minimizing the response time and allows teams to work on all of their inboxes through a single dashboard.
This is a good AI tool that will prove invaluable when it comes to shared communications and streamlining your CX processes, yet it does little to help you actually achieve the promised “Inbox Zero”. So if you’re looking for a solution that will sort through your spammy emails and cherry-pick the worthy ones – this is not the right tool for you.
AI-powered technology for customer feedback analysis
62% of customers would prefer handling their communication via email as opposed to other channels. That makes email a logical choice for sending feedback requests, especially since the average response rate for an email survey is 24.8%. So where does AI come in?
You won’t be surprised that it gathers customer data better than any human could hope to. So, when it comes to customer feedback analysis, AI can help you with:
- easily processing qualitative responses
- grouping survey results into predefined categories (e.g. “positive” and “negative”)
- analyzing large volumes of data in a short time
- preparing survey questions for a specific group of recipients
- designing and sending automated messages based on the feedback received
- prioritizing solutions based on the customer feedback
Highlighted tool: BetterFeedback
BetterFeedback is a qualitative customer feedback analysis tool that uses AI to discover what your customers “think, want, and have issues with”.
It tackles the most vital challenges when it comes to analyzing complex customer feedback reports, including but not limited to running automated sentiment analysis, discovering trends across several channels, and grouping the feedback from multiple channels in a single space.
BetterFeedback is still a new solution that works on improving their functionalities. Yet, one of the most crucial aspects of feedback analysis – advanced reporting – is still WIP. At the same time, the tool comes with a free 7-day trial but no free-forever version which might be disheartening for some users, and not enough time to experiment with all the features’ fullest potential.
AI tools for A/B testing
If you meddle with email marketing, you must have seen this one coming! After all, split testing is about as inseparable from email best practices, as writing strong subject lines.
But it also tends to get quite boring and repetitive over time. Testing whether putting “⚡️” at the end of your subject line would convert better than “👉” or no emoji at all email after email can get very tiresome really quick. Especially, since there are 3664 emojis you could potentially experiment with!
And that is not to mention other important bits of your email marketing strategy you should be A/B testing, like, you know, copy, header images, send time, etc. Fortunately, now you can use AI to make all the hard decisions for you!
When it comes to split-testing, AI can help you with:
- going beyond “A/B” and testing C, D, E, and many more options with ease
- running more tests at once
- personalizing split tests per segment, not per channel
- applying sophisticated targeting
- optimizing copy variants for better conversions
Highlighted tool: abtesting.ai
OK, to be fair abtesting.ai is not limited to email alone, but how is that a bad thing?
The platform is dedicated to optimizing your A/B testing efforts across multiple channels for a variety of purposes, and they’re not afraid of using AI to maximize their effectiveness for all the crucial elements of your email (or a landing page), be it a subject line, copy, or a catchy CTA.
Limitations: there is simply more to email than there is to landings, I’m afraid. It’s more sophisticated, more touch-point-y, and more, well, direct. So if you want to optimize emails and emails alone, this tool may not be efficient enough.
Artificial intelligence for predictive analytics
You could say that all predictive analytics is dependent on AI to an extent, and you wouldn’t be wrong. The problem is the statistical application of the model makes it far too technical for most email marketers.
At the same time, you could achieve great things with your email marketing campaigns if you could estimate the recipients’ reactions and draw actionable insights grounded in data analysis. That’s where AI comes in.
When it comes to predictive analytics for email marketing, AI could help you with:
- estimating and preventing churn and unsubscribing
- increasing the likelihood of conversion/purchase
- defining the best timing for sending communications
- devising customer win-back strategies with data-driven insights
- preparing future campaigns based on real-time metrics
- sending personalized emails thanks to smart segmentation
Now, if you read carefully, you must have noticed that predictive analytics, in a way, is present across every possible application of AI in email marketing. In fact, it’s the most important component all your campaigns powered up by artificial intelligence should have in common.
That’s why there won’t be a single tool recommendation (well, kinda).
Instead, a short word of caution: none of us wants to end up overwhelmed by dozens of digital platforms, SaaS products, and annual subscriptions to send a simple email. So, find a single AI-friendly platform that covers most of your email marketing needs and start from here.
For example, we are about to launch a complete AI Email Creator at GetResponse. A fast and easy way to support our users on every crucial step of launching their email marketing campaigns – craft converting email subject lines (is already live!), complete email layout and design, natural language generation for smart copywriting based on your input, A/B testing, and more!
Do we expect it to change the way AI in email marketing is seen? Perhaps.
Will it make the lives of hundreds of email marketers easier? Absolutely!
Author’s note: The tool is currently in beta and should be released in the nearest future, stay tuned!
Instead, you can already tap into the power of AI with our highly customizable plan – GetResponse MAX.
But is this galore of AI-powered solutions, ChatGPT craze, and our overall desire to do more with less that is so prevalent in B2B marketers ready to satisfy the market’s need?
Current challenges and limitations of AI
Of course, not everything’s colors and rainbows in the rapid adoption of artificial intelligence. In fact, there are quite a few pitfalls we, digital marketers, might be a little too eager to jump into.
But let’s stick to email campaigns as our main area of focus.
As you recall from the beginning of this post, an AI model is only as good as the data you feed it. And that’s where the problems begin.
Access to quality data needed to train an artificial intelligence model is often expensive, that’s a given. But for building a strong email marketing strategy, not only does your data need to be, well, good, but it also needs to be relevant.
That’s why the best way to train your own AI model for email purposes is to already have an active website with steady monthly traffic and enough email subscribers for the AI to learn from.
It might be a piece of cake for a large corporation, but if you’re a digital agency or a B2B SaaS solution, chances are you can’t get enough data. Here, you can always jump on the out-of-the-box AI solutions (like product recommendations embedded into your marketing automation platform) and still achieve impressive results.
“But you can always buy data!” you say. Well, the golden rule of succeeding in email marketing is simple: never purchase your contact lists.
AI could be an exception here for as long as you use the database just to train your model and later move it to the real audience. Yet, there’s another catch: AI is an up-and-coming trend that hasn’t been well-regulated just yet, but the changes are coming.
If you google “AI privacy,” you’d get more than 4 billion search results (when I started this article last week, the number was at 3.8 billion).
AI depends on data and a significant chunk of that data is highly sensitive personal information a user might be willing to trade in exchange for an ebook, but wouldn’t be particularly happy about being fed to a robot.
California’s Consumer Privacy Act (CCPA) in the U.S. and the General Data Protection Regulation (GDPR) in European Union have some AI-related clauses, but none of them clearly pave the path toward ethical applications of Artificial Intelligence and machine learning.
So be mindful of that.
Bias & misconceptions about artificial intelligence
There’s a natural bias towards AI as something new and unfamiliar. It appears to be candy in a shiny wrapper, but who knows what’s really underneath? For the generations growing up with The Matrix and Westworld (back when it was still good), this trend may seem scarier rather than promising.
In fact, more people are worried (38%) about AI rather than excited (15%) about it (MediaPost):
That’s why before using artificial intelligence in digital marketing to its fullest, we first need to understand how it works and why it makes decisions the way it does. Explainable AI (XAI) might be able to help there, but we’re far from the majority of organizations adopting this model.
The loss of human touch
Automation has already taken its toll on email marketing’s overall humanity.
No personalization best practice can really make up for dozens of “Hey Anna, I represent Product X and thought you’d be the right person in GetResponse to chat about my awesome marketing product” emails I receive on a weekly basis on a customer side of things. Or, perhaps, AI would generate a more appealing copy here.
No power in the world will persuade me that the “Happy Birthday” email my internet provider diligently sent at 1 AM my local time was done out of concern for my personal well-being.
And that’s OK, as long as I know there’s a person behind that marketing automation platform who thought about giving me a 5% discount on a day when I’m likely to be in a good mood. That the decision has been made by a human being striving to connect with another human being, albeit for commercial reasons (but hey, we are marketers, we have to be commercial-oriented!).
The text could be mechanical and didn’t reflect on more than some first-party data I willingly provided, but it wasn’t generated by a robot. And once AI takes over the rest of the email marketing automation, I’m not sure there will be enough humanity left in our already transactional relationships.
But, perhaps, that is what the future of (email) marketing should be. Keeping business to business and person to person – outside of the Outlook accounts and LinkedIn profiles, saving the personal for the people we consider the closest, and never having to take the work home ever again.
Only time will tell. Meanwhile, let’s have fun with what we’ve got and see how AI will change email marketing in the nearest future.