B2C customer experience has been undergoing an evolution over the last decade. While your phone was out, you probably looked at ten other brands in the meanwhile, and e-vendors might have even approached you with a “Looking for a new phone?” banner during your idle browsing. No company who is serious about selling phones will ignore such a scenario. In a highly competitive environment, your customer is likely to switch at the blink of an eye if the vacuum for a sale exists at that point. Now, consider a similar scenario with a different product. Let’s consider the story of a builder, who leased a dragline excavator for completing time-bound delivery of a commercial construction project. Soon after the project started, the excavator malfunctions, and the project comes to a halt. We don’t know what caused the failure, and neither do the engineers working on the project. The contractor calls the dealership and informs them about the issue – if the project gets further delayed, there are big losses on the line. And no one likes that.
In the situation we just discussed, who should reach out to whom first? Should it be your dealer reaching out to the manufacturing company’s support center? Companies that are truly customer-centric today, are already transforming their first step – identifying the problem for their customer, while informing and apologising to the right stakeholder for an unexpected failure. How? If the dragline excavator had an IoT device that detected critical failures and reported them to engineering support teams right away. This would cut the time taken for an outsourced call center to report the obvious to your support teams. Knowing that a product failed is necessary, but not sufficient to fix the problem. Knowing why holds the real value here.
Enter AI - an intelligent capability working in the middle of enterprise software and communication channels that could interpret the output from the IoT device (error codes, log files etc). Then weave in other data inputs, typically lying in disparate systems, such as location, machine model, serial number, geography and product specific troubleshooting manuals, and customer records. The best course of action – reaching out to the dealership, informing them of their machine’s failure, an expected cause and timeline of service and an apology – could have been done proactively.
There is one powerful constant amidst a hundred variables in the customer service process, particularly for enterprise customer support. Issues repeat themselves, in whole or in parts. With the right type of machine learning models, it is possible to judge the similarities across issues and support tickets when a machine encounters failure. AI picks on miniscule patterns between in a complex network of data points – those that are impossible to pick manually. If a smart capability could tap into these use cases, customer support would be a dream.
AI makes use of highly complex models to generate insights from a wide variety of data. In other words, it does the digging and grunt work. Indeed, it can help you reach out with the right information, without making three customer support executives navigate through five different software while playing hide-and-seek with countless files in countless directories. Empowering the customer means taking care of them in such difficult situations with the right information.