How AI Helps Email Marketing

If your business has been looking for ways to start using artificial intelligence (AI) in your marketing, one channel you’ve likely considered is email.

Most businesses rely heavily on email as part of their sales and marketing strategy, and as Google phases out third-party cookies, the need for a strong contact database and an effective email program will only intensify.

Email tends to be a central channel for our clients, and it’s important for us as well, so it was top of mind as we considered new ways to incorporate AI into our toolkit.

After doing our research and comparing options, we recently onboarded an AI-based email marketing optimization tool called Seventh Sense, which uses machine learning to personalize email delivery times and, ultimately, increase open and click rates.

If you’re interested in a manageable, cost-effective technology that can get you started with AI-based email marketing, read on.

The Case for Optimizing Send Times with AI

I’m not sure how many times I’ve searched for the “best days/times to send email for [insert industry],” but it’s too many. Do the search yourself and you’ll find the answers to be an unhelpful mix of “it depends,” “Tuesdays at 10am” or “definitely NOT Tuesdays at 10am.”

Good stuff!

Advice on suggested send times will ultimately hedge its bets with some obvious common sense: the best way to optimize for send time is by conducting tests and analyzing your contacts’ engagement to see when they tend to open and click emails.

This is true, but there are a few problems with this, namely:

  • Humans aren’t great at this kind of analysis
  • You have other time-sensitive stuff to do
  • Your database is constantly changing
  • There’s always the remote possibility that your database is not one homogenous block of people who all like reading their emails on the same days at the same times

So, what’s an email marketer to do?

Enter Artificial Intelligence

While humans may not be great at it, testing, analyzing and optimizing email engagement using data is exactly the type of task AI is suited for.

If we could improve email performance while freeing up the time that would go to manually testing and analyzing engagement trends for more strategic work, it’d be a clear win. And we could stop doing searches for the latest on the best days/times to send emails.

Of course, send time optimization is just one application of many within the AI-based email marketing category. For example, there are tools like Phrasee and Persado that use AI to craft and test different email subject lines and copy variants based on engagement data. The list is growing fast.

While we’ve got our eyes on some other tools in the market, for now we’re keeping the copywriting in human hands while we test AI on send time optimization.

About Seventh Sense

Seventh Sense struck us as a good fit because, in addition to having the features we were looking for and overwhelmingly positive customer reviews, the platform is built to optimize email delivery for HubSpot and Marketo. As a HubSpot agency partner, many of our clients use the platform for marketing and sales emails.

By integrating with HubSpot, Seventh Sense analyzes your contacts and their engagement to personalize and predict when they are most likely to read your email. Using machine learning, the tool continually tests and analyzes engagement data so that over time, open and click rates increases, your sender score improves and you get more results from your email campaigns.

(Seventh Sense dashboard view of monthly audience engagement)

There are two primary ways in which we are using this tool: workflows and batch emails.

In the workflows (automated emails), we simply add the Seventh Sense nurture send time optimization between a delay and email send, so that the email send time will be optimized (this can either be a 24-hour or 7-day window).

For batch (mass) emails, we pick the window within which we want the email to go out and have Seventh Sense tell HubSpot when to send the emails. That’s it.

(Seventh Sense dashboard view of contact engagement by day and time)

While this type of testing might seem similar to A/B testing in some ways, there’s a big difference.

Where A/B testing is seeing how two variants perform against each other when shown to users at random, Seventh Sense is using machine learning to analyze user data and predict when your contacts will be most likely to engage – and testing is happening on the individual contact level rather than against randomized segments.

(By the way, if it’d be helpful to have a crash course to get you up to speed on AI and machine learning, I recommend checking out The Ultimate Beginner’s Guide to AI in Marketing from Marketing AI Institute.)

As of the writing of this post, we’re still in the early days of using Seventh Sense, so we’re not yet a case study. If all goes well, I hope to be writing a follow-up post soon with a nice up-and-to-the-right graph showing how much AI has improved our email performance.

Determining What’s Right for You

For businesses sending a high volume of emails, the benefits of an AI tool like Seventh Sense are clear. Even an incremental improvement in engagement over time could easily move the needle and be worth the price of admission.

One thing to keep in mind is that machine learning needs data to learn – the more the better. If your business email program starts and ends with a monthly newsletter to a few hundred contacts, it’s unlikely that you’ll see much lift from this technology.

Another consideration is how engaged your contacts are with your emails. If there is a significant amount of historical engagement data to draw from, machine learning tools will have more to work with. If engagement across your database is generally low, it will rely more on a predictive, test-and-see approach – in which case the runway to improvement will likely be longer.

If you use HubSpot to do regular email marketing and are interested in learning more about Seventh Sense or AI-based email in general, drop us a line.