When it comes to building customer relationships, nothing comes close to personalization. Eight in 10 consumers are more likely to purchase from brands that offer personalized experiences, Epsilon research shows. No wonder personalization is a priority for so many of today’s organizations.

To make personalization possible, however, companies have to sift through stacks of data. Uber alone has collected over 100 petabytes of customer data — far too much for human analysts to handle in a timely manner. That’s why it and others have turned to AI technologies like machine learning.

By using machines to analyze how, when, and what customers shop for, companies can predict on a person-by-person basis what customers want. Here’s how:

1. Individualize digital experiences.  
   Until the rising of ML, A/B testing was a manual process that took hours to deliver actual insights. AI models, however, can use online behavioral signals to determine the best experience to serve up in real time. And because AI systems become smarter as more data is fed into them, they actually get better at predicting outcomes over time.
   Financial services company HSBC, for example, recently tested AI-driven dynamic content against static content on its mobile app’s home page. The personalized results drastically outperformed their static peers in terms of click-through rates to product pages.

2. Optimize emails.
   We live in an attention economy. Like never before, marketers are vying for consumers’ mental space. Nowhere is that more evident than in email inboxes. The average office worker receives 121 emails per day, each of which is carefully crafted for opens and click throughs.
   To compete, marketers are turning to AI to analyze email engagement data, predict open rates, and minimize churn. One brand that hit it out of the park with personalized emails is Virgin Holidays. Just by individualizing its messaging, the travel company increased its email opens by 70% and clicks by 65%. As a result, Virgin Holidays raised its revenue by 49%, or about £17.3 million.
3. Make ads more effective.
   Not long ago, advertising was a guessing game. Billboards, television, radio, and print ads were one-size-fits-all, forcing marketers to predict what would resonate with the widest slice of their audience. Today, the combination of digital media and AI have made advertising into a science. By 2020, eMarketer predicts that more than 86% of digital display ads will be bought programmatically. In other words, almost nine in 10 of them will soon be tailored to each viewer.
   To understand why companies are rushing toward programmatic ads, consider how Missing People used them to maximize a small budget. After receiving a £10 million donation, Missing People used programmatic ads to make location-specific appeals in its audience members’ social media feeds. In doing so, it raised its response rates from 50% to 70%.
   
4. Broaden customer bases.
   AI can help brands better connect with their customers, but it can also help them identify new ones. Known as lookalike modeling, the process involves comparing demographic and psychographic traits of existing customers to predict which other consumers are most likely to convert. On the flip side, AI can make smart recommendations around audience suppression, saving brands money by encouraging them to skip over consumers they’re unlikely to win over.
   Using an AI-powered data management platform, Princess Cruises employed lookalike modeling to identify high-value customer segments. By considering the interests and experiences prior cruisers, Princess Cruises helped its marketers spot prime consumer targets.

5. Customize shopping experiences.
Largely thanks to Amazon, consumers have come to expect customized online shopping experiences. Around 35% of the e-commerce giant’s revenue is tied to its recommendation engine. Gartner predicts that by 2020, digital companies will boost their profits by as much as 15% by using AI to predict customer intent.

But AI can do more than suggest similar products. British fashion retailer Asos helps shoppers choose their perfect size by analyzing which items shoppers keep and which they return. Asos’s Fit Assistant is powered by Fit Analytics, which uses machine learning to make 250 million sizing recommendations per month, in over 20 languages, and across more than 17,000 brands.

AI may be a collection of technologies, but it’s key to a more human customer experience. Personalization is built on the back of big data, which all but impossible to analyze by hand. Instead, put your trust in AI. You — and your customers — won’t be disappointed.

Shannon Hamilton

Shannon is a Product Manager for Adobe Target. She is responsible for the machine learning / AI capabilities within Target, including ML-based personalization and ML-enhanced testing. Prior to Adobe, Shannon worked in management consulting, building customer experience strategies for clients across North America. Shannon holds an MBA from the Kellogg School of Management at Northwestern University, where she specialized in data science and marketing.