Tech Triumph: Study Shows Effect Of AI On Email

Email deliverability rates are soaring. That’s the hype. The truth is that they are on the way down, according to Optimizing Email Deliverability With AI, a study by the Relevancy Group, sponsored by Return Path. 

The average delivery rate has fallen from 89.7% in 2015 to 87.8% ithis year, according to the study, which was written by David Daniels and Nicholas Einstein. Of course, that’s still slightly higher than the 86.97% reported in 2017.  

One reason is that “the filters and algorithms that inbox providers often leverage to protect recipients often result in false positives,” the authors state. 

How do you avoid having 13% of your audience cut off from news about products and services? By sending relevant emails driven by artificial intelligence.

Consumers demand it — a study found that 51% want emails that feature relevant products, and 41% want emails to show products that they viewed online, according to the study. And 52% will ignore or opt out of irrelevant emails.



Happily, 96% of marketers are confident that AI will improve the customer experience, although only 54% are now employing it. And those who use it soon see results, the authors contend.

For example, open and click-through rates average around two points higher for AI senders versus those who rely on human intervention, the study shows. Conversion rates are also stronger, although the rate of increase is not as high.

In addition, delivery rates are a point higher for AI senders. And they pull an average order value of $145.08 compared to $138 for non-AI email.

Finally, AI produces 41% higher monthly revenue from fewer emails — the average send for users is 35 million, versus 36 million for brands that are not yet up on AI.

How do you get on board this train?

The authors recommends a process called AMP — Analyze, Model, Predict. 

  • Analyze — This helps firms make sense of the mountain of data that accrues from an email marketing send—one billion points of data for firms that blast out over 36 million a month.
  • Model — This process relies machine learning algorithms to classify and process data into segments. This would take hundreds of work hours to achieve if done by humans. 
  • Predict —  AI allows marketers to make meaningful, actionable predictions, the authors assert.

What’s the price tag for all this? The mean budget for personalization and recommendation technologies is $928,350, the Relevancy Group found in a June survey.


Next story loading loading..