Brands are turning to AI and machine learning to answer extremely complex questions, including, “What words should I use to communicate most effectively?” Companies like Persado, Phrasee, Touchstone, and others have turned AI’s data-crunching and pattern-finding power onto email subject lines, as well as other marketing copy.
Oracle Marketing Cloud Consulting has seen high interest in these tools among our clients. However, when our clients have tried them, the results have been mixed.
Among those brands that didn’t find enough success, the most common complaints that we’ve heard are that:
The lift in email performance wasn’t sustainable beyond a few months,
and that the results in general didn’t justify the expense.
If you’re considering these tools, we have some advice to get you on the path to success.
However, first we want to make it clear that these solutions are best suited for brands with large lists. If your list has fewer than 1 million active subscribers, you may not be a great candidate for AI subject line tools, according to Persado.
Scale is critical because these machine learning systems need to explore a ton of language to understand which particular words and patterns of words resonate with your unique audience. Having a large audience allows tests across more segments.
Sending at a high frequency is not required, according to Phrasee and Persado, although it absolutely helps the system learn more quickly. Higher send frequencies equal more test segments over time.
Understanding the importance of scale is key to our first piece of advice, which is...
It’s a common misperception that the goal of a subject line is to get opens. The truth is that the goal of a subject line is to get the subscribers who are most likely to convert to open the email. Unless you’re sending a reengagement email, opens are a secondary metric, at best.
So, optimize your subject lines to drive action as far down the funnel as possible. Ideally, that’s conversions—that is, whatever you’re asking your subscriber to do in a particular email, whether it’s buy a product, register for a webinar, read a blog post, or follow your brand on Instagram.
However, reaching statistical significance on conversions can often take too long, making it so you’re not nimble enough in your testing. Clicks are an acceptable compromise that gets you to statistical significance much more quickly with a metric that correlates to conversions far better than opens does.
The problem is that many brands using AI subject line writing tools are optimizing for opens. Sometimes they’re drawn to opens because they can reach statistical significance much more quickly. But being quickly certain about the influence of an uncertain indicator is not the path to becoming a data-driven marketer.
Optimizing for opens is a contributing factor to the pattern of disappointing performance that many of our clients have reported to us: These tools often produce a boost of engagement for a few months and then results fall back down.
This is because optimizing for opens gears your subject lines toward attracting the curious instead of those who are most likely to convert. Overly clever and curiosity-peaking subject lines generally lead to fatigue and unsubscribes as subscribers are essentially tricked into opening emails that generally turn out to not be of interest to them. On the other hand, optimizing for clicks better aligns subject line content with email content, which better maintains trust and longer-lasting subscriber relationships.
These AI subject line writing tools do not understand your brand, so they will suggest things are off-brand—potentially wildly off-brand. As the protector of your brand, it’s up to you to recognize these off-brand suggestions and reject them.
“One of our travel brands ran their first test last month and the winning subject line started with ‘RE:’ like it was a reply to an email,” says Mark Sambor, Senior Director at Oracle Marketing Cloud. “ They obviously had a level of discomfort with this.”
Bradford Johnson, Senior Director of Strategic Services at Oracle Marketing Cloud Consulting, advises marketers to involve your teammates who own and drive brand copy, messaging, and creative.
“If these tools are run in a silo, you’ll hear from those disgruntled teams about your copy soon enough,” says Johnson. “Work with your vendor and internal partners to find a balance between computer-driven optimizations and preserving your human, on-brand voice.”
Some of these tools give you the ability to add words to a do-not-use list. If offered, use that to steer the AI away from words and tactics you don’t want to use.
At the same time, we recommend keeping a close eye on opportunities to use branded language that your AI isn’t likely to suggest, such as the names of your products, words from your jingle, and other words that have special brand meaning. So, it’s as much about leveraging your brand in your wording as avoiding words that go off brand.
While all the major providers of AI subject line writing tools have accumulated enough real-world experience into their databases to be able to provide good suggestions, it won’t have any experience with your particular subscribers in the beginning. Building up that experience takes time.
Your vendor will likely encourage you to test at a high frequency—and across channels, if they offer that functionality. That’s fine. As we mentioned earlier, more tests equal faster learning.
However, we always recommend testing AI-created subject lines against a control subject line or two that your team creates. It’s probably best to think of these tools as suggesting marketing copy rather than being marketing copywriters themselves.
We firmly believe that AI copywriting tools have a bright future. In time, more competition, lower infrastructure costs, and perhaps different pricing models will change the return on investment calculus that have caused some of our clients to walk away from this technology.
In the meantime, you can maximize your chances of success by:
Optimizing your subject lines for clicks rather than opens
Rejecting off-brand subject lines that you wouldn’t accept from a human and staying true to your brand
Confirming the performance of AI-suggested subject lines by testing against human-created control subject lines
“Testing capabilities like this can drive value for your business,” says Peter Briggs, Director of Strategic Services, Oracle Marketing Cloud Consulting. “You just need to have a plan, internal alignment, and clear understanding for what will equal success.”
Briggs adds, “Sometimes pushing the envelope on your typical UX comfort level is when you can see the biggest impact. Sometimes you have to take risks.”
Need help with AI and analytics tools? Oracle Marketing Cloud Consulting has more than 500 of the leading marketing minds ready to help you to achieve more with the leading marketing cloud, including a Strategic Services team that can help you get the most out of AI and analytics tools.
Chad S. White is the Head of Research at Oracle Marketing Consulting and the author of four editions of Email Marketing Rules and nearly 4,000 posts about digital and email marketing. A former journalist, he’s been featured in more than 100 publications, including The New York Times, The Wall Street Journal, and Advertising Age. Chad was named the ANA's 2018 Email Marketer Thought Leader of the Year. Follow him on LinkedIn, Twitter, and Mastodon.