Artificial Intelligence and Machine Learning make marketing magic

Marketers want to understand their customers: if you know what customers want, you can give it to them. They’re more likely to stick around, engage, buy things, visit frequently, and say nice things about you.  

Customers want to be understood by marketers. We don’t want our time wasted with offers and messaging we have no interest in. Don’t show me a promo for spicy food when I literally never buy it; learn my coffee order and use that to personalize an upsell offer just for me, and we’re in business. Even better: learn my order and have my coffee ready and waiting for me at the Queen St store on Wednesday at 8am.  

This level of personalization is only possible if you know enough about me. You’ll know where I shop, what time I get to Queen St, what I order and on which days. You know I can’t turn down a donut, and you know that showing me the chili lunch special will get no traction. Bonus points if you also know that today it’s raining and traffic is bad, so you hold my coffee back 15 minutes. 

Yes, it’s all marketing, but when consumers don’t know what’s going on behind the scenes (and it’s done properly!) it’s also kind of magic. 

 

analytics studio spotlight.png

Personalization needs a lot of data 

The more personal you want to get, the more data you’re going to need.  

You can’t get inside your customers’ heads by relying on one piece of information about them. Consider purchase history, which is a reasonable place to start. If you’re a chocolatier it certainly helps to know if someone prefers milk chocolate over dark. But you also need to know if they are more likely to have chocolate in the morning or in the afternoon. And if they have it with coffee but never with tea. And if they don’t buy chocolate on hot days, because then they prefer something less melty. And you should know that they typically go for the 3-1 special and only buy the expensive boxes in August because it’s mom’s birthday. And that they stop in at one location on weekdays but a completely different location on weekends.  

For many marketers the logistics of this level of personalization may seem daunting; you need to be able to take in a vast array of data, and you need to be able to use that information (preferably in real time) to create an engaging experience for each individual customer. 

 

A brief history of marketing personalization 

Once upon a time, marketing was personalized because marketing was personal. Shopkeepers knew their customers by name, knew what they bought, and knew when they were likely to need to come in to stock up. Which means they could recommend products, give advice on alternatives or additional items, and just generally give customers exactly what they needed, when they needed it. 

Now marketers are dealing with far more customers, across more regions: Amazon talks to 197m people a month; McDonald’s serves over 69 million customers daily. You’re probably going to need a bit of assistance remembering all those names – let alone the rest of their information. 

Using data to create marketing that speaks directly to customers is nothing new; marketers just have a lot more information to work with than ever before. And thanks to the likes of Amazon and Netflix, customers have come to expect that marketers will give them relevant content: whether it’s related items or recommended viewing.  

Luckily, as customers have come to expect more personalized marketing, they’ve also become more willing to let marketers have access to their information: if I want to have my coffee ready and waiting for me at 8am, I’m going to have to let you know where I am at 8am. The important thing is that marketers ensure they’re collecting only the data necessary to create the experience, and that they actually use the data they collect. 

 

Using Artificial Intelligence and Machine Learning- powered platforms to create magic 

Because marketers are now working with hundreds of millions of data points a day, AI and ML- powered platforms are increasingly being used to do the heavy lifting; the sheer scale of data makes it impossible to handle manually. 

Instead of trying to update spreadsheets and get reports out of the data team, marketers use intelligent platforms that generate insights and make real-time marketing recommendations based on the data they take in. You don’t have to watch the platform work (and in many cases may not even need to train it), you just let it handle the data for you. Most of this happens automagically, as the platform recognizes trends and shifts in the data and adjusts its recommendations accordingly. 

And it pays off. Gartner predicts that AI-powered personalization based on customer intent will help lift digital business profits by up to 15% by 2020. BCG found that personalized shopping experiences makes people 110% more likely to buy additional items and 40% more likely to spend more. They’re also more likely to recommend the business to others. 

But wait, there’s more! 

 

analytics studio dashboard.png

From intelligence, to insight, to action 

Marketers can use AI and ML to refine personalized marketing based on a constant stream of customer data, which is great. But what’s more important is where that takes you next: what does the data help you to understand about consumer behavior? How can you use those insights to make something awesome happen? And how can AI and ML help?  

An AI-enabled analytics platform should be able to learn what matters most to you: based on your data, how you use the insights, the reports you typically run and any other information that feeds its algorithm. It’ll not only report back on your data; it’ll also be able to alert you to critical information you didn’t know you needed. Maybe something you didn’t know you could look for.  

Is there a sudden spike in loyalty card registrations? Is today’s redemption activity outside the established pattern of behavior? What’s caused it and what does it mean for marketing? Is one piece of content under-performing others in a specific region? If you double down on an unexpectedly successful offer, what will the impact be on your bottom line? If you have all these insights delivered to you, then the hard data work is already done. No more spreadsheets, no more relying on data scientists to unlock the information behind the numbers. Marketing’s job becomes easier and making data-driven marketing decisions becomes a lot faster. 

And because the platform’s working with a wealth of customer and contextual data, it improves customer experience along the journey: using all its smarts on all your data to give people what they need, before they even know they need it. Like magic!