shimmery help sign hanging from front of building
Automating frustrating and time consuming processes can help customers and workers alike PHOTO: Adrian Scottow

Anyone who's opened a brokerage account, managed an issue with the tax collector or was admitted to a hospital knows the frustration of the waiting game.

The level of complexity in regulated industries such as finance, healthcare and government requires human judgment to make even typical processes operate correctly. The operative word here is “correctly.” In some industries, getting something right is more important than getting it done fast. 

For companies in these markets, incorrect process flow can result in fines, lawsuits or in the case of healthcare, deaths.

Unfortunately, while humans are the best way to make judgment calls, they can be slow. Slowness impacts customer experience negatively. Waiting is a source of stress and frustration, the very opposite of what one wants in customer experience.

Speeding Human Processes With Machine Learning

Augmenting human decision making and judgment is a classic application of machine learning. In this context, machine learning aids knowledge workers and customers alike. 

First, it helps encapsulate the decisions of others into an easy to access experience. Machine learning is just that — learning. Systems such as Google Cloud AI, IBM Watson or Microsoft Azure Machine Learning are used to track previously made decisions, sift through complex information and build heuristic models of success and failure. These models can then be used to guide others, in effect sharing expertise. 

This, in turn, allows humans to move quickly through tasks that can hold up a process.

Machine learning can also impact customer-focused processes by reducing errors. Humans are error-prone creatures — both customers and knowledge workers make them all the time. Mistakes happen, but they act as drags on processes. 

Errors by customers cause force them into exception processing queues, while errors by knowledge workers put both customers and the business at risk. Machine learning tools can act as a check on errors by guiding both customers and knowledge workers through difficult issues more easily. 

By distilling large amounts of data into human-sized bites, both customers and knowledge workers can move through a process more easily with fewer mistakes.

Where Machine Learning Impacts Customer Experience 

There are two types of customer process where machine learning may have the most impact on customer experience: intake processes and exception processing. 

Intake Processing

Intake processing is any process that begins a relationship with customers. It may be an application for a job, opening a checking account or admittance to a hospital. Intake processes have a series of steps that the customer and the company must move through to start a relationship.

For example, in order to open a brokerage account, the customer must provide accurate financial information to the company. Meanwhile, the brokerage firm must evaluate if the customer may be trying to commit fraud, may live in a country with financial sanctions, has money to invest and is doing so in an informed manner.

Machine learning could assist the customers in filling out the necessary information by walking them through a Q&A tailored toward them personally. At the same time, it can help the company in detecting possible anomalies that may be cause for concern.

Exception Processing

Another area where machine learning can help customer-facing processes, and hence the customer experience, is in exception processing. These are often the trickiest portions of standard processes because they are anything but standard. 

Whether it’s an error on the part of a customer, missing information or some external criteria against which the customer information must be checked — for example “does insurance cover this?” or “will the SEC allow this?” — exceptions have the ability to slow down or derail the customer experience.

Here again, machine learning has a part. By encapsulating the expertise of the company’s knowledge workers in a form that can be delivered digitally and directly to the consumer, customer errors can be reduced, leading to fewer exceptions. Machine learning can help detect exceptions more quickly and aid the knowledge worker in resolving them quickly.

An End to the Waiting Game?

By automating parts of processes that require human judgment, machine learning holds the promise of speeding up customer-focused processes and producing fewer irritating errors. Faster responses with less back and forth leads to a much-enhanced customer experience.

Despite their promise, the cost and complexity of machine learning systems, as well as the unfamiliarity most software developers have with their implementation, will inhibit deployment of these systems. Once machine learning becomes easier to implement, expect to see many complex customer-facing processes enhanced by its use.

Augmenting the human processes that currently cause customers so much stress will provide a boon for the overall customer experience — and that means a win for both the customer and the business.