How to take advantage of predictive technology for an improved customer experience
A summary of Christopher Nash’ talk at WebTomorrow 2015 on ‘Creating Lifetime Customers… by Using Machine Learning and Predictive Personalization’. Millions of personal interactions and transactions are made on the web every single second and minute. The fact is, Data never sleeps. One way of looking at that is to say the web is still a booming platform of human interactions. Another more mercantile way to look at it is that there are millions of customer experiences every minute.
Often however, the experiences are of low quality today: you visit a website and you’re disengaged because it doesn’t deliver what you’re looking for. You often feel ill-targeted because you’re prompted to buy stuff you do not need. You regularly feel disconnected from the conversations, tweets and updates that take over your social media newsfeeds. In short: today, millions of “dis-customered” experiences are around.
So what about the web tomorrow then? Can we keep on living in this “dis-customered” situation or had we better anticipate the needs of customers, in the moment, based on predictive technologies? Actually, the Holy Grail is: know what the customer knows before they know it.
Tweetaway: Create relevant experiences by predicting your customer’s next best action insit.es/1E8Wyga by @HakimZemni #mrx #newmr #webtomorrow
The next best action
Customers reveal data all the time, in the moment. Brands must create relevant experiences by using this data and by anticipating what the next best action for the customer is, says Chris. So, how do you reach that ‘next best action’? Chris argues two technologies will help brands define the ‘next best action’ for customers, improving the customer experience all together: predictive personalization and machine learning.
Predictive personalization starts from the idea that there are x number of personae and x number of product categories for each online interaction. After a few interactions made by an online visitor, matches are made; as the experience evolves, the profiling can become more distinct based on the intent of the visitor. In order to be truly predictive, you want to store that single view of the customer and, through machine learning, start building your algorithm for better predictions. That’s where cohort analysis comes in: what are common denominators and what do other people do in similar situations? And that’s how the future of the web looks brighter and less “dis-customered”, according to Chris.
Tweetaway: Big data promises more relevance; but we’re definitely not there yet insit.es/1E8Wyga by @HakimZemni #mrx #newmr #webtomorrow
The final thought went to making a self-assessment for your company and pinpointing your customer experience performance today. On his website, he created a self-assessment test for businesses to check where you are today, in order to be better prepared for tomorrow.
So yes, big data is still promising the world that it will help the Internet become more relevant and less “dis-customered”, but we’re definitely not there yet. It seems there still are a few decades of “dis-customered” experiences ahead of us. In the meantime you can improve your customer experiences by inviting your customers into your boardroom, really taking in their feedback and even more importantly: acting on it.