How to Predict Why Customers Engage

Chris Caile of Nuance discusses how you can predict how customers will choose to engage with your organisation, through the use of an American football analogy.

Football is back and that means fantasy football leagues are in full swing. Every owner is searching for the winning combination. While each has their own strategy such as selecting players from their local team (Go Hawks!), every owner is counting on one thing – strong predictions.

Analysts and pundits all try to guess which players will do well this week or against a specific defence. Accurate predictions will make one owner a star, and poor predictions? Well, winning isn’t everything. And, hey, you get to pick first next year!

Football lives on stats and prognostication every year. And with the rise of Artificial Intelligence (AI) and data analytics, the power of prediction is moving into the mainstream for customer service.

Organisations seeking an edge to improve their customer service may want to investigate what prediction offers.

Prediction Playbook #1 –The Fundamentals

Using predictive capabilities starts, just like football, with the fundamentals and basic “blocking and tackling”. Most organisations already have the most fundamental element – customer transaction data.

It all revolves around the data. Your customers are calling your contact centre, engaging a live chat agent, or visiting your website to search for answers. With the right analysis, you can better determine why they are calling and if they will call again in the future.

Prediction Playbook #2 – The Coach

Data on its own doesn’t do any good if it’s sitting idle in a database. It would be like a bunch of players milling around on the field wondering what to do. Players and data need structure and someone to help them deliver their full potential. They need coaches.

Like a good football coach, when it comes to prediction, the “coach” is a set of machine learning models powered by AI.

Machine-learning models are sophisticated tools that aggregate the massive amounts of customer data and then conduct analysis on them to identify patterns and trends.

Over time, the models get smarter as they see more and more patterns and learn what doesn’t work. As the models improve, the service an organisation can offer its customers also improves.

Prediction Playbook #3 – Executing Plays

Once the fundamentals (data) and the coach (machine learning) are in place, organisations can set about deploying predictive capabilities using some of the most standard use cases. Think of these as the football team executing the plays they’ve worked hard to practise.

For most organisations there are three typical scenarios where prediction will serve them well:

Predictive intent – Predict why someone is contacting the organisation. This can be an incoming call or an engagement with a live chat agent.

If the company knows why someone is calling the IVR, for example, they can customise the menus and speed their resolution.

Predictive routing – Use prediction to proactively, and effectively, route an inbound caller to the best agent or resource on the web to service their needs.

Proactive notification – Machine-learning models can identify customer patterns and trends and proactively contact a customer in advance via SMS or email.

For example, a cellphone provider may spot patterns with certain customers calling about problems with a particular model.

Prediction allows the provider to proactively contact customers with the problem phone and address issues before a phone call.

In Summary

No matter how an organisation wants to improve customer service, the power of prediction is making it possible.

Organisations need to explore predictive capabilities to continue to innovate and stay ahead of customer needs. Besides, why should fantasy football owners have all the fun?