Actioning Individual Customer Context With Artificial Intelligence – Forbes
Today’s customers interact with products or brands at various stages, from brand awareness to comparison to purchase to product support and loyalty. Most marketers are focused on awareness and maybe a little on experience during comparison and loyalty. They rarely get deeper, either because they do not have access to the full customer experience cycle, because there has not been a way to handle all of it, or because they are not incentivized to do so,
In recent years, marketing has started to delve into customer support to standardize customer experience. However, machine learning (ML) and artificial intelligence (AI) have started to provide tools that can offer much deeper insight and actions. So, what are the true and far-reaching opportunities for marketers with AI?
Imagine being able to serve every customer or potential customer at any touch point in the cycle as if there were a live salesperson there who instantly knew that person’s history with the brand — the summary of clickstream, purchase history, loyalty, support encounters and sentiment. Imagine that salesperson’s efficiency if they could instantly have at their disposal a best offer or reward, or information to nudge the customer to engage or repeat engage with the brand.
So far, we have seen data analytics and maybe some ML used to analyze clickstreams and some digital visitor behavior to make recommendations for cross-selling and up-selling. Much more recently, we have seen chatbots used to scale customer support, as well as do voice-based sentiment analysis with tools based in some form of ML, AI, natural language processing and speech recognition. While these have been steps in the right direction, they are very narrow. They are point solutions based on isolated data pools. They take no steps toward either determining broader context or finding the best opportunities for marketers and sales folks to increase acquisition and retention. Catering to customer segment is old-fashioned. With AI, we can cater to individuals based on extended or immediate history with the brand.
The current capabilities of ML and AI allow not only the complete context of the customer to be gathered instantly, but also to serve the experience most likely to drive customers further down the funnel. The following key elements are required to exploit this phenomenal opportunity that is now available to marketing and sales.
- To enable capture of complete customer context: Data from the following different pools needs to be joined and integrated into a single customer 360 view or hub: store and online purchase data, website and mobile clickstream, social behavior, support encounters, sentiment analysis, demographics and loyalty data. Joining data across anonymous and authenticated customer sessions is still not easy, but there are workarounds. Data so integrated provides a complete immediate and extended history of any customer.
- To create the right experience for each context: Whether created manually or generated via automated methods, content and experiences that cater to customer intent and context at various points in the customer life cycle have to be developed. Everything from appropriately keyworded variations of product or service descriptions and videos to granular reward schemes for loyalty or for support scenarios to different purchase experiences that surprise and delight must be generated and kept at the ready.
- To serve the best experience each time: Previously, data analysis and even ML was used to determine hidden customer segmentation and to predict product or service demand. Marketing and sales then manually arranged experiences to be served, after some A/B and even multivariate testing. Now AI can be used to run acquisition and retention maximization models. AI can run multivariate testing using the data and the experiences developed in the previous steps. Based on training gained from such testing (at scale or in smaller focus areas), AI can instantly determine the most likely experience that would cause new customer acquisition or repeat sale and retention.
All of the above should only be done in stages. Companies have spent years attempting to pull together customer 360 databases. Start by integrating online clickstreams with demographics and purchase and support data. Bring in-store purchase, sentiment analysis and social data at a later stage. With the same objective of leveraging online first, create content and experiences in digital formats before taking on in-store experiences or field sales agent empowerment. Retention models to achieve or increase customer lifetime value are much more lucrative and less infested with stakeholders. Honing and running retention models before new customer acquisition models is a great way to prove out AI capabilities with the fewest unnecessary concerns and the least amount of interference.
Relinquishing manual or even rules-based recommendation is not an easy cultural shift, but already, results with ML and AI outperform anything marketing and sales have otherwise done.
- Best Western used AI-based tips and tricks for various destinations to help travelers plan and realized a 48% increase in traffic to Best Western Locations.
- The Humane Society used ML messaging and realized an 86% lift in shelter pet adoption by reaching friends and cohabitants in the relationship graphs of those who had been identified visiting a shelter and showing any intentions to adopt.
Marketers reacting in retrospect to segmentation analysis is old-school and far too late. The new tech-savvy marketer wants to wield individual customer context and experiences nimbly, in real time, to catch the most new customers and achieve the greatest customer lifetime value. The race is on. Brands that win this race will set themselves so far ahead of the rest that they will not be caught for years.