Why? What has changed?
In the past, advertising (today we would say marketing) was all about creativity, intuition, research and… lots of whiskey, if “Mad Men” is anything to go by.
Now it’s still a creative job but much more automated and data-oriented than is commonly thought.
What about intuition? It’s not that simple. A study from Netflix shows that algorithms are much better than people in predicting which show will be watched next. Its authors called it the “intuition failure”. It’s worth paying attention to because typically a viewer loses interest after perhaps 60 to 90 seconds of trying to find a video to watch. At the same time, Netflix estimates its algorithms produce $1 billion a year in value from customer retention.
“Tomorrow, there is no need to waste 50% of your advertising budget, because we will know who is interested in our messaging and we will pay them to listen. AI will tell us who we need to talk to, when and how with what message and information”, says Mark Mueller-Eberstein, CEO & Founder at Adgetec. You’ll agree that he’s right when you realize that 35% of Amazon.com’s revenue is generated by its recommendation engine.
Another example is Target, the second-largest discount store retailer in the US. After focusing on analyzing customer data their revenues grew from $44 billion to $67 billion. All of that wouldn’t be possible without machine learning algorithms. What’s important, you don’t have to be a technological giant like Amazon or Netflix to take advantage of them. All you need to do is to integrate AI-driven software with your service.
But what exactly can you do with artificial intelligence software?
Imagine you have an e-commerce business and you’re able to predict if someone who has entered your website is willing to buy, or what and when he or she will buy. Sounds like magic, doesn’t it? But it’s real and possible, thanks to machine learning algorithms.
Prediction mechanisms work when you collect customer data and use to continuously create segments. Algorithms recognize repetitions in behavioral patterns of particular customer segments and are able to predict the most probable behavior based on previous similar scenarios. At Synerise, we’re working hard on them, and our Predict module is accurate up to 95%.
Another exciting application of machine learning in marketing is dynamic content, commonly used in newsletters, landing pages and on the websites. It allows you to create one campaign with detaills that change automatically according to predefined conditions, including:
- demographic data
- past behavior of the client
Dynamic content gets results. According to a report by the Annuitas Group, leads that are nurtured with targeted content increase sales opportunities by up to 20%.
Have you heard of chatbots? Very often when you enter a company’s website, you will see a little chat icon encouraging you to ask a question or start a conversation. Usually, there is a sales rep on the other side, but more and more often it’s… a bot. And if it’s well designed, you may not even recognize it.
Bots can work 24 hours a day, seven days a week, during Christmas Eve and other holidays. They can contact an enormous number of clients at the same time. They’re also connected to the database, CRM and get all the information necessary to solve the problem efficiently.
All the applications mentioned above are just the beginning. Marketers use image recognition software, smart display campaigns, churn prevention, customer lifetime value forecasting, and more. Marketing has become predictable. Intuition is a desirable quality in a marketer, but what can be better than effective tools to support it?