The Future of Personalization with AI and Machine Learning

the future of personalization

Cookie-cutter customer experiences just won’t cut it. And online personalization, once seen as a competitive differentiator, is now a strategic imperative.

Consider this: nine out of 10 marketers say their customers expect individualized experiences – and, according to Gartner, by the end of this year, organizations that have fully invested in online personalization will outsell those that haven’t by more than 30%.

Even though personalization is becoming more of the norm, we’ve all had those eye-roll-inducing experiences where a brand that should “know” us completely misses the mark – like, for example, when your bank encourages you to apply for a credit card you already have. Or your favorite retailer emails you about a hot sale in a category you’ve never shown interest in (e.g., baby clothes if you’re not a parent, or lawn care if you live in a high-rise apartment). In your work life, you’ve likely gone to a company’s website, only to be greeted with the same video shown to you before…for an industry that isn’t even relevant to what you do.

What’s the outcome in situations like these? It’s a frustrated customer. With a wide array of options, consumers today can (and do!) take their business to places where they feel recognized, appreciated and valued as an individual.

So companies have to do better.

The Netflix Effect

Consumers have been conditioned to want and expect personalized experiences from all the businesses they interact with, thanks in part to “The Netflix Effect” – where companies such as Netflix, Amazon and Spotify have established a baseline for what personalization means. Consumers think: “If Netflix recognizes me from transaction to transaction and follows my activity from mobile app to the web, why can’t my bank – for example – do the same?” It’s a fair question!

There’s often a misconception among marketers, though, that personalization is something only the Netflixes of the world can do effectively…and can afford to do. This simply isn’t true. Cost and technical barriers to personalization are lower than ever.

Personalization in the Age of Machine Learning

At Digital Growth Unleashed, I’m looking forward to talking about how artificial intelligence – and in particular, machine learning – has changed the game for personalization in recent years. It’s a topic I am passionate about; in fact, I devoted my new book to exploring this topic in detail.

When we discuss machine learning in the context of personalization, it means using computers to process vast amounts of data, in milliseconds, to make the best decision about what to show each person. Machine learning puts the vision of “The One to One Future,” which renowned customer experience experts Don Peppers and Martha Rogers, Ph.D., predicted in 1993, truly (and finally!) within reach – accessible to businesses of all sizes and across industries.

With machine learning, true personalization – also called individualization – can be done at scale. Marketers can evolve from the traditional and manual rule-based approach to personalization (e.g. “If a person falls into Segment A, then show him Experience X”), which is most effective at targeting groups of people, to using machine-learning-based algorithms and predictive analytics to present the most relevant experience to each and every visitor.

With the right platform, algorithms can be created, customized and managed by marketers and other business users – no need to cede control to a “black box,” and no coding or IT required. And, importantly, one well-tuned machine learning algorithm can do the work of thousands of manual rules.

Getting Started

I often advise companies that are just beginning to tackle personalization to start small initially. Have a clear strategy in mind, and, at the onset, go after what’s possible right now. You’ll see a difference and can add more complex initiatives incrementally.

Applying personalization across all relevant channels – so customers are recognized and can pick up where they left off –  should be the goal. To do so, companies need to be able to: 1) track an individual’s behavior across different channels; 2) merge that information with pertinent customer data from other systems; 3) automatically interpret the data to determine affinities and intent; 4) house everything in a central place – creating a single, unified profile for each person; and 5) act on all the data in real time.

The most important element – and an essential part of all five steps – is, of course, data. Your personalization strategy is only as effective as the data informing it. Even the most advanced algorithms can’t work their magic if they have incorrect, inadequate or outdated data. But I do not recommend that you wait to start personalizing until you have cleaned up all your enterprise data. If you do, you will be waiting a very long time. Start with your digital channels. It’s easy to leverage a next-generation personalization platform to start bringing in deep, contextual, real-time, accurate behavioral data from your sites, apps and email. You can use this data in your personalization platform to create terrific results. Then leverage this success to start bringing in other enterprise data sources one by one, cleaning them as you go.

I look forward to discussing this more at Digital Growth Unleashed in a few months and explaining how, with machine learning, we can all raise the bar on personalization – doing away with those frustrating one-size-fits-all approaches and providing superior customer experiences.

About the Author

Karl Wirth is the CEO and co-founder of Evergage, the real-time personalization platform company, and author of the book “One-to-One Personalization in the Age of Machine Learning.” Check out his session at Digital Growth Unleashed, “Raising the Bar on Personalized Experiences with Machine Learning and AI.”

Conversion Conference – Now Digital Growth Unleashed!

Las Vegas – May 16-17, 2018.