Machine Learning Demystified – An Interview with David Daniels
Today we are welcoming back David Daniels from The Relevancy Group to talk to us about Machine Learning and the game-changing role it can play in personalization. TRG’s latest report, “Machine Learning Demystified: Applying Predictive Analytics to Optimize Personalization,” summarizes how Machine Learning is being leveraged by marketers to unify and personalize customer experiences across channels, even as personalization tactics continue to dominate email marketing.
Zeta: We hear a lot about Machine Learning these days. How you would define it for our readers?
DD: Machine Learning is all about using artificial intelligence technology to learn without programming. For the purposes of our latest report we focused on personalization and how Machine Learning can develop computer programs that can change when exposed to new data.
Zeta: Can you tell us a little about the current landscape and how marketers are currently using personalization tactics?
DD: Sure, today personalization is almost equally split between human and machines. The Relevancy Group research shows that many marketers plan to adopt Personalization Artificial Intelligence (PAI) and have the budgets to do so. However, we also learned that not all marketers have the same affinity for PAI and the differences vary by industry and organization size. Some of the key points we found in our survey were:
Enterprise Adoption Tactics
- Personalization based on real-time – 63%
- Utilization of dynamic content – 62%
- Social retargeting of email resp/non-resp – 58%
Benchmarking Personalization Tactics by Industry
- First name personalization in tech/high tech – 87%
- Human-curated prod personalization in email for retail – 52%
- Human-curated prod personalization in email for financial services – 43%
Zeta: You mentioned marketers having the budget to adopt PAI. What types of investments are being made in Machine Learning?
DD: Marketers are realizing quickly that Machine Learning’s “set-it-and-forget-it” approach is favorable. They see the efficiencies that can be created using PAI and many are setting aside significant dollars to meet that end. The median budget for this technology among the marketers we surveyed was $950,000.
Zeta: Wow, that’s quite a figure. What is the downside to implementing PAI? What are the roadblocks that might prevent marketers from moving forward?
DD: Machine Learning requires human intervention for analysis and fine tuning. Some marketers are challenged to implement PAI due to a lack of human resources that can perform these functions. It’s important for marketers to have the necessary human resources and/or agency service partners to maintain these machine learning investments and maximize their investments.
Zeta: How are marketers leveraging predictive modeling to bolster personalization and increase sales?
DD: Many marketers are overlooking rich behavioral data that has been proven to drive results. Past behavior is the best predictor of future behavior. Our report on The Future of Relevance talks about this link.
Zeta: Let’s get down to the big money question, how much revenue are marketers seeing using PAI vs. human-curated personalization?
DD: Yes, it does always come down to that question, doesn’t it? Well our respondents told us that PAI had monthly revenues that were 8% higher than those doing human curation. This equated to just over $612,000 a month in revenue for PAI email marketers.
Zeta: How would you summarize the Machine Learning marketing landscape currently?
DD: PAI can certainly impact business results online and within email marketing. Successful marketers moving forward will be those who utilize outside agency services to guide and aid in the development and execution of tests and feeding these machines with the proper data. They need to remember that PAI still requires human intervention, oversight and analysis.
Want to learn more? Join us for a webinar Machine Learning Demystified May 10 at 1PM EST
