The promise of artificial intelligence has put stars in many marketers’ eyes, but it’s not always clear what’s hype and what’s real. Contributor Kyle Henderick lays out some tactics that can be employed today.
Not only is artificial intelligence (AI) one of the hottest topics in the media today, it’s also one of the biggest points of discussion among email marketers. From robots to hyper-personalization, the potential marketing implications for AI are exciting to discuss.
But many marketers may be in the dark about how they can leverage AI in a practical sense today, and vendors have been hungry to try to take advantage of it by selling pipe dreams of no more work to do.
Let’s get real. You’re not going to replace your marketing and sales team with robots any time soon, and any AI tools will require proper data, investment and additional resources. That said, there are a few relatively easy ways for you to use AI to improve the effectiveness of your email marketing campaigns.
Machine learning can be used to automate and optimize many aspects of your email marketing program, such as content, send times and frequency. In turn, you can be confident that you’re moving the needle to maximize engagement and conversions. For example, AI could ideally predict the content and offer that are most likely to drive an individual customer to convert, along with the best time of day and the frequency that is most likely to keep them engaged.
While one subscriber might be more likely to make a purchase from subject line A, imagery B and promotion C, another subscriber might be more likely to convert from an email that’s the complete opposite.
Humans simply cannot replicate this degree of constant learning and optimize at the same pace. It would take a significant amount of A/B testing to eventually determine the most effective content, frequency and offers for each individual subscriber, and that’s not even dealing with creating content to support it! The right AI tools have the potential to do this in seconds, freeing up email marketers’ time to think more strategically.
How AI can optimize three key areas of email marketing programs
As a marketer, you’ve probably spent extensive time and money determining the right content and promotions to offer in email campaigns, along with the right send times and cadence. It can be tricky to determine the combination of these factors that will maximize conversions. That’s where AI can help the most.
AI has the potential to optimize three key areas of any email marketing campaign.
1. Email content
Content is an obvious part of email marketing strategy that could benefit from automation. That could be anything from the body copy to the subject lines to the images within the layout. Traditionally, marketers test combinations of each of these to determine what performs best, but the process is prone to human error/bias and is quite time-consuming. That’s where AI can help.
Not only does AI technology enable you to learn more quickly what combination of content will perform best (and therefore maximize revenue sooner), but it will also alleviate a lot of the time you and your team spend working on A/B testing, and it will allow you to expand to greater variations of testing elements. On top of that, the right tool can pick the winning combination for each segment during delivery to start serving up the right content right away and drive incremental revenue.
2. Send times and frequency
It’s common for marketers to rely on broad assumptions about the best time of day to send marketing emails. While this might be effective for some of your subscribers, it doesn’t apply to every single individual on your list. At the end of the day, subscribers are human beings, not statistics. Each one has different habits and preferences for checking and engaging with email.
AI could automate the ideal send time and day for each subscriber based on his or her engagement history. It can take the analytics modeling and turn it up in real time — which would be very tough for humans to replicate because there simply isn’t enough time. Machine learning technology can understand preferences for large lists of subscribers quickly to determine the time and day to send emails that will maximize engagement and conversions.
The same idea applies to the frequency and cadence of the emails you send to each subscriber. While one subscriber might engage more if you increase the frequency of your check-ins, others might feel overwhelmed and unsubscribe.
Often, brands miss out on revenue opportunities by giving subscribers the wrong promotion or offer. While some subscribers might convert with a lower offer (for example 25 percent off versus 40 percent), others might only engage with an offer when it cuts the price by a certain dollar amount. It’s easy to see how the wrong offer could lead to a loss of revenue for you, but it’s almost impossible to accurately predict the right offer using humans alone.
The right tool can collect and learn from consumer purchase history data over time and determine which types of offers perform best for each subscriber. This might be a combination of a percent off and free shipping, a certain dollar amount off or nothing at all. Whatever it is, AI technology can help you drive more revenue from your promotional campaigns.
Start small and build
Lastly, while all of the above can help your program, it’s important to remember much of this AI buzz is still in the ideation phase; taking on everything at once never works, and if it sounds too easy, it’s probably a trap. Even some of the use cases mentioned above are still just ideas. However, there are some tools that are already paving the way for AI as it relates to email marketing.
While the concept of AI is exciting, and also a little scary, it’s important for marketers to start small. Make sure you have the right technologies in place now to easily implement AI tools when they become more accessible down the road. While you do, keep up the hard work using the email marketing best practices you know and love to keep your subscribers engaged and loyal over the long term.
by Kyle Henderick
source: MARTECH TODAY