A scientist’s guide to email language
by Dr Neil Yager
There’s not a secret winning formula for email marketing success – but injecting some science into the process is a sure-fire way to give your campaign performance a boost. Using technology to help shape your email marketing campaigns can increase results in many ways – and as Phrasee’s Chief Scientist I’ve seen first-hand how powerful it can be, particularly when it comes to the use of language.
The language you use in your emails may seem like just one small piece of the overall puzzle, but the impact of using the right language is not to be underestimated. A huge 69% of email recipients report having marked a marketing email as “spam” based solely on its subject line. This language could lose more than two thirds of your potential customers before they’ve even opened the email.
Our AI-powered copywriting has created email subject lines that have reached more than 300 million people across the world. That’s a lot of people, and a lot of data that highlights what works well, and not so well. Here are my key outtakes on the language that creates successful email marketing campaigns time and time again.
Treats, not tricks
Misleading or spammy subject lines can sometimes lead to a short-term boost in open rates. Lines like “Free beer!” might catch your eye and get you to open an email, but if you’re not really offering free beer, you fall at the first hurdle. Sure, your open rates will be great, but people won’t go further than that.
It is easy to get people to open an email using misleading language. But in the long run, this strategy is a bad idea. A good email needs to compel the recipients not only to open it but for them to be in the right frame of mind when viewing the email body. Those who are tricked into opening an email are unlikely to convert. But it goes further than that – open rates will drop over time. There are only so many times you can trick someone into opening an email – they’re unlikely to do so again and this can also lead to higher unsubscribe rates.
In a similar vein, it’s unfortunately still common to see marketing emails that rely on language that incites negative emotions to try and sell more. Whether it’s making someone feel anxious, guilty or like they’re missing out, intentionally triggering these types of emotions is another tactic that’s going to quickly turn off customers.
In fact, research from Phrasee and Vitreous World found that consumers are turned off by marketing that uses unethical approaches, such as using high-pressure or anxiety-inducing language to encourage sales. Almost seven in ten (68%) consumers across the UK and the US said they would not buy from a brand that used negative emotions in their marketing. That’s proof that this can hit not only your email open rates but longer-term customer loyalty, and your bottom line.
Test, learn – and test again
Multivariate testing is already best practice in email marketing, to identify the subject line and language that will perform most effectively. It’s important to make this a long-term strategy as well as using it to ensure short-term performance for a given campaign.
Conducting repeated multivariate tests over a longer period of time produces much more robust performance data than any single subject line test can. And a winning email subject line won’t necessarily work the second time around, so it’s important to keep language fresh to keep customers interested – and keep testing.
Plus, we also know that every audience is different. A strategy that works well with one email list (e.g. using curiosity) may not work well with another. Therefore, it is important to run multivariate tests and include a wide variety of different words, phrases, syntactic structures and sentiments. Incorporating more linguistic variance into the subject lines you test on your audience is the only effective way to find out what’s going to boost your campaign performance and make your brand stand out from the crowd.
To help brands measure the linguistic diversity of their email subject lines, Phrasee have built a tool, which gives brands ‘a diversity score’ based on their current output, and the option of comparing this with their competitors.
By analysing historical multivariate test data, you can learn what works best. At Phrasee we use deep learning that looks at thousands of linguistic features, so we know what language your audiences respond to most.
Dr Neil Yager PhD is Co-founder and Chief Scientist at Phrasee
Trying to get ahead of the competition?
Check out our Learning Hub for everything you’ll need to know!