This fall, Searchlogic celebrated Advertising Week by packing its bags and flying to New York for the Google Partner Summit. From peeking into the future or voice-based artificial intelligence to getting the down-low on new data-driven CRM strategies, the summit packed in an impressive wealth of actionable advice. One of the main topics was machine learning: where it stands now, what it’s being used for, and how it might unfold in both the near and distant future.
What is machine learning?
If you need a quick refresher, machine learning teaches programs to teach themselves so they can identify data patterns and continue improving their skills over time. It’s at the core of revolutionary medical advancements like genomic mapping. It’s what you see in those neural net-generated Google DeepDream images and what powers Google’s Teachable Machine. It’s a huge and promising field, and we’ve only riding the first wave.
A quick machine learning definition from Wired: “ … the hottest paradigm of the software world: using learning algorithms (‘learners’) and tons of data to ‘teach’ software to accomplish its tasks.” There was too much on the topic at the Google Partner Summit to cover exhaustively, but here are a few highlights that really got us thinking.
Marketing the machine learning way
Whether it’s in-house or agency, most marketing teams haven’t fully embraced a machine learning-first approach. Once adequate tools are widely available, marketing will never look the same. Some commonly-used tools and tactics are already leveraging machine learning to great effect, like AdWords, content analysis, and automated data visualization.
In the future, we’re likely to see machine learning automating much of the function of a marketing team. Voice assistants and bots will rapidly replace human marketing labor. Marketing teams will use it to predict churn and demonstrate ROI more precisely than ever before. It’ll also help bring more personalization to each and every customer interaction
Machine learning and voice AI get personal
One of the biggest impacts of machine learning on marketing is the integration of customer service and voice-controlled AI services like Alexa. Machine learning means voice assistants are learning all about their users — everything from their work schedules to their unique senses of humor, offering brands the novel ability to get to know their customers on a one-on-one level.
This means huge things for customer acquisition and retention. But sometimes learning about a customer also means gaining the ability to intervene in dangerous situations and possibly even save lives. For example, brands like Amazon now have to consider how they’ll have Alexa respond when she identifies worrying patterns from her users. They say more than 50 percent of Alexa interactions aren’t about getting a piece of information — they’re about relating to her for entertainment, companionship, and even emotional support. If she thinks someone is in danger of harming themselves, she’ll offer words of encouragement along with the number of a crisis line.
Marketing job security depends on solid strategy
So even though machine learning might answer a customer’s question or offer other high-level types of support, your marketing job isn’t doomed. Ad tech is becoming more and more automated — agencies and directors must show insight to keep their jobs.
There’s still no stand-in for the precision and creativity needed to come up with stellar strategy. Campaign execution and customer service in the assistance age means marketers had better learn to hone in on it. Marketers can thrive by doing strategy while machine learning takes care of the rest. Shifting from execution to strategy makes marketers more future-proof. Regardless of where it takes us next, machine learning is metabolizing staggering amounts of data into actionable insights for marketers faster than ever before.
How will you use it to level up this year?
