28 June 2017

Guest Post: Artificial Intelligence On The Rise In Australian Contact Centers

Editor’s Note:This guest post, by James Witcombe of SMAART Recruitment, features an interview with PTP’s Erni Mededovic and highlights the rise of artificial intelligence in Australian contact centers, and it’s connection to customer experience.  

Australian contact centers are taking a steady and careful approach to introducing artificial intelligence (AI) into their contact centers. Around 7% of Australian contact centers currently use AI in their interactions with customers, with an additional 11% planning to introduce it in the coming 12 months. A high percentage of remaining contact centers are currently exploring the idea, however 44% of contact centers have no intentions to introduce AI.
The results have been released by SMAART Recruitment, who surveyed over 150 contact centre leaders and decision makers around Australia in its 2017 Contact Centre Leaders Survey.

The stigma associated with artificial intelligence being a slow to respond cold robotic process is wearing off as the technologies improve. However there is still largely an unknown ability of how and where AI can be utilized in the contact centre industry. With the industry seeking to become more educated we can expect to see an increase in the implementation both in back and front of house within Australian contact centers in the coming years.

AIEnhancing the Customer Experience

Erni Mededovic from PTP in California has a great passion for technical innovation through the use of new technologies aimed at enhancing customer experience.

“I think we will see much steeper progress in the world of AI over the next 2 years. A lot of companies are gearing up and testing different approaches to utilize this great technology segment, and we should probably see some great improvements, not only in consumer offerings, but in underlying methodologies and practices in how the AI agents are built currently. All this should lead to a better, faster and more efficient AI agent implementation that will allow us to handle bigger and more complex problems.”

According to the experts, baby steps are key when implementing any form of AI. Starting with internal processes rather than being customer facing is wise so that a contact centre can get exposure to how AI works before they start using it to deal with customers. Mededovic agrees “The best way to start would be to isolate a specific functionality that could benefit the call center from the utilization of the AI, and concentrate on it. Building such independent modules that rely on the common libraries could give an organization flexibility in scaling their AI system. “

Chatbots Are Here and Will Only get Better

Douglas Park, Customer Contact Architect from Aspect Software is seeing a real demand for AI in the contact centre industry. As an industry leader with the world’s first cloud based customer engagement software, Aspect Software has recognised the need to provide omni-channel platforms to contact centers because customers are demanding it.

“Calling a contact centre is a last resort for most customers”, Park said, “if you think about it, everyone tries to serve themselves first. You go to the website first. You look for FAQ’s or an instructional video. You download the app, you try yourself. People are almost desperate not to call organizations. The main place to do this is online, and often on the company website. So, given your customer is already on your website, why wouldn’t you offer a bot they can chat too?”

Park points out the reluctance of some businesses to embrace forms of AI often comes to a lack of understanding or an incorrect assumption. “A lot of people think a chatbot is some avatar which is a souped up interface for a FAQ. That’s not a chatbot. A chatbot attempts to better understand the customer intent and the question. Intent is a big part. Let’s look at IBM’s Watson – it wants to find better and more refined answers… however our chat software is interested in understanding the questions better.”

Park puts it simply, “chatbots need to be focused on being able to reduce customer effort.” If a bot can do this then why wouldn’t every contact centre – and business – provide it?

AI-DrawingWhere to Start With AI?

For chatbots, keep it simple. While they can be developed to become more intelligent over time, it’s important to start with common requests. “What are the top 5 things that are your customers are asking you?” suggest Park. “Focus on the simple things that are commonly asked. Don’t try to do everything at once.”  

Contact centers are cyclical and great recorders and keepers of information. Seasonal spikes in customer interaction are a great opportunity to be able to use AI to communicate to a customer in the way the customer prefers. “If an organization is about to send out a large run of bills they will generally know what increase this will have in calls and the common questions asked. Contact centers can now start asking, what problem can I solve most sufficiently with a bot?”

Gregor Harnell, Head of Insurance Direct Service at MLC Life Insurance agrees. “We are exploring chatbots and other forms of AI. We can see an opportunity to use chatbots for lot of basic enquiries – change of address, details or payment options. The bot replies with a link or an automated response. We think there may also be a place for it with quoting too. The big thing that hit me is that the chatbot answers when you are ready to ask the question. The bot is there when you need it. Go away, comeback and it’s still there”.

However, Mededovic has a timely reminder for contact centers racing to implement AI. “A lot of companies bet on using AI in a customer facing environment. In today’s world where customers are overloaded with offers from the competition they are too valuable to be exposed to any mistakes that could lead to customer disappointment. A small single mistake could cost you a valuable customer. AI is a great technology that is still in it’s infancy (due it’s complexity). Much better utilization of AI would be in a background processing (agent monitoring, intelligent IVR and skill transfer).”

Where are the Leaders Using AI?

Mededovic sees the leading organization using AI in a way that touches and improves many parts of the contact centre. “A great way to utilize AI in the call centre is the monitoring and connection of communication channels and social media. Building intelligent IVRs (interactive voice responses) which can predict customers IVR interaction patterns and product interests is the most likely next generation of the IVR. Using neural nets for data mining, performing intelligent agent skill transfers.” AI doesn’t have to be solely for customers, it can be used to gauge agent emotions as well as pitch and distress and really help to monitor and improve agent performance.

Park emphasizes the fact that we shouldn’t see AI as either a hindrance or a threat to employees. “Bots allow qualifying to the point of allowing the human to be able to be best at the human part.” Hartnell from MLC Life Insurance agrees, “There’s always the question “will automation reduce number of employees?” I actually think we will connect to more customers and the type of conversation will change. The more high valued calls will still be coming to the call centre. Online chat and chat bots will remove the simple calls – the failure calls, repeat calls, the really basic calls. The important calls we still really want to receive and focus on.”

AI presents contact centers with great opportunities and challenges in the coming years. Those that learn from it and understand it quickly will be the first to benefit.

Learn More

Read Erni Mededovic’s comprehensive post on how contact center’s can leverage AI to improve customer service

Authored bY

James Witcombe

James Witcombe is a recruiter based in Melbourne, Australia at SMAART Recruitment. James leads a team of talented recruiters that have a particular passion for recruiting for contact centers large and small.


Subscribe today to have our stories delivered directly to your inbox.

Invalid Email