Will Machine Learning Cause Demise of Contact Centre Managers?

Will Machine Learning cause Demise of Contact Centre Managers? Asks contactSPACE

robots.image.sep.2017contactSPACE performs millions of communications across voice, email and SMS with different customers each month, and each of these communications has a level of effort associated with it, which ultimately leads to an outcome. In a contact centre environment, these outcomes are then assessed by management by evaluating any number of reportable measures related to the performance of the data. Additionally, trends must be recognised and considered in the overall assessment of a campaign. From this assessment, the manager will then try to formulate and implement a strategy that delivers the greatest levels of efficiency and productivity from the agents using the data, to secure even better outcomes.

A strategic change with voice communications may, for example, include the reduction of wrap time or the number of times a record is attempted – the changes usually being specific in pursuit of certain outcomes. Following implementation of the changes, the manager will then monitor the effect of the changes over a period of time- and the cycle is then repeated. It’s a lot of information to gather, assess and act upon in a timely way, to develop and implement tactical change, and furthermore, managers will need to be able to recall previous strategies and their effects from memory and from some style of workforce planning tool- and sometimes from hand-written notes. Historically that has worked for businesses, however, times are changing and businesses processes are evolving. The subtle difference in efficiency between one business and another can mean greater profits or it can simply spell the difference between surviving or not.

We have two key issues to solve – What data we should be looking at – How long it will take us to action strategic changes.

Even someone that is highly skilled at generating and analysing reports will need time to develop a new strategy, implement it and then measure the outcome, and it’s here where machine learning can be used to profound effect.

We need to recognise that machine learning is very different to artificial intelligence (A.I.) as they are two separate and interesting subjects. The idea of artificial intelligence is to create machines that can eventually teach themselves to learn, make complicated decisions and consider the associated consequences; whereas machine learning aligns more to data mining in real time which provides an outcome based on a model that has been created.

There is, also a fundamental difference, however, between data mining and machine learning, in that data mining is reliant on a person, whereas machine learning is reliant on a set of mathematical models with prerequisites which is automated. You feed these models with information which is then processed and returned to you as insights and ultimately as recommendations on how to get the best outcomes for your unique situation.

Machine learning is already around us. It’s present in the things we use in our daily lives. If you’ve subscribed to a media streaming service like Netflix for example, you’ll receive movie recommendations which come from the Netflix Recommender System (RS) which is using machine learning. This platform is providing recommendations based on your viewing history compared to the viewing history of almost 100 million other subscribers. The premise is that users who are measurably similar in their historical preferences are likely to also share similar tastes in the future. Similarly, when you shop online with Kogan.com, you’ll often receive emails with a list of recommended products, and these have also been generated by a machine that is continuously learning about your shopping behaviour and that of the thousands of other shoppers who share similar purchasing behaviour to you. Scary stuff right!

All the big players such as Google, Amazon and Facebook are investing heavily in machine learning, and this type of outcome is becoming easier to achieve.

What could happen if we apply similar methods in a contact centre environment? What if machine learning could suggest to the contact centre manager the optimal dial rate or how many users to have on an initiative. What if machine learning could look at all the data quickly and suggest the optimum time to contact specific people. What if these decisions could be performed in real time based on a recommendation model? What if the changes were applied automatically in real time without any human intervention? The possibilities seem endless. It could potentially alleviate the contact centre manager from having to make these decisions and having to go through the lengthy processes outlined earlier, saving a lot of time and ultimately achieving far better outcomes.

Some experts are predicting that up to 49% of jobs in the financial services sector could be automated by 2025 using machine learning, particularly in those roles involved in international currency trading and calculations of insurance premiums.

The banks are investing heavily in “social robots” and “machine learning robotics” which is set on improving the human interaction from the delivery process and they’re using machine learning very specifically to deal with it.

If we examine the contact centre industry, we know that the same trend is likely to apply and as this scenario quickly takes hold, the pace with which it will arrive will shock people. The basic tools are already here.

Imagine if a machine is recommending to you the most efficient way to run your contact centre – the most efficient way to contact people and the most efficient way to get a better outcome.

Imagine if it increased the efficiency of your people by 100% – what’s going to happen? You could potentially perform the same amount of work with half the workforce!

contactSPACElogosmallAdditional Information

For additional information on contactSPACE visit their Website

error: Content Protected